Update #1 (October 7th, 2022): this thread has evolved as we all learned more about the issue. It turns out there appears to be a problem in Windows or the Nvidia drivers that is causing huge DPC spikes, and all of Google has no solutions, though thousands of people are experiencing the issue. It’s more than likely that you are too, if you have an Nvidia graphics card, most people just don’t realize what the issue is until they use LatencyMon to find the culprit (Nvidia driver nvlddmkm.sys). Meanwhile, people are having crashes, poor framerates, stuttering, audio dropouts, etcetera, and the drivers are likely the true cause, but the blame is being put elsewhere. My original post is below, but if you keep reading the replies from everyone it has grown into a lot of different troubleshooting, because it isn’t IRQ related as originally thought.
Update #2 (December 8th, 2022): a summary of the first 8 pages in this thread has been written up and posted as a reply here:
https://www.ntlite.com/community/in…lddmkm-sys-dpc-latency.2924/page-8#post-30724
Update #3 (December 12th, 2022): throwing in the towel for now and working on other research and optimizations instead. This is for sure a bug that Microsoft/Nvidia need to fix. At this point we just need people to file support tickets enmasse with Nvidia and Microsoft, and send letters to editors of gaming/benchmarking sites to bring news coverage to it, as this issue is affecting nearly every Nvidia user. It’s not some niche problem, there is clearly a serious issue here that is being swept under the rug:
https://www.ntlite.com/community/in…iver-nvlddmkm-sys-dpc-latency.2924/post-31026
Update #4 (December 17th, 2022): discussed virtualization, issues in 22H2, and pausing Windows Update:
https://www.ntlite.com/community/in…iver-nvlddmkm-sys-dpc-latency.2924/post-31207
Update #5 (December 24th, 2022): discussed differences between general DPC and Nvidia bug, made a DPC quick fix guide:
https://www.ntlite.com/community/in…iver-nvlddmkm-sys-dpc-latency.2924/post-31369
Update #6 (December 26th, 2022): updated rules for posting in this thread and tips on better testing methodology:
https://www.ntlite.com/community/in…iver-nvlddmkm-sys-dpc-latency.2924/post-31449
Update #7 (January 5th, 2023): update on works in progress and the future of this thread, please read this:
https://www.ntlite.com/community/in…iver-nvlddmkm-sys-dpc-latency.2924/post-31988
Update #8 (January 8th, 2023): posted a comprehensive DPC latency guide for fixing general DPC issues:
https://www.ntlite.com/community/in…iver-nvlddmkm-sys-dpc-latency.2924/post-32114
Update #9 (January 12th, 2023): must read information for people suffering from non-Nvidia DPC issues:
https://www.ntlite.com/community/in…iver-nvlddmkm-sys-dpc-latency.2924/post-32210
—————————————————
Original Post:
I don’t know much about IRQ and some of the related topics, and I’m hoping someone can please point me in the right direction, or maybe help me solve this issue since we have a lot of really advanced techies here.
Running LatencyMon reveals that nvlddmkm.sys (Nvidia Graphics driver kernel) keeps having DPC spikes up to 800 fairly frequently. This within itself is an extremely common issue that can be found all over Google, but I haven’t yet come across any solutions, except for 1, but it’s not really a sensible one.
According to one of the replies to a LinusTechTips forum (https://linustechtips.com/topic/1222687-dpc-latency-issues-with-nvlddmkm-and-dxgkrnlsys/) this is happening because the graphics card is sharing an IRQ with a problematic device. Indeed, it turns out that my card according to msinfo32.exe is in fact sharing IRQ 16 with one of my motherboard’s USB host controllers.
The «solution» given on that forum is to do the following:
1) Uninstall all PCI/PCIe cards using Device Manager. Select to remove drivers if asked. Remember to show «hidden devices» and clear those too. (Use DDU to clear the GFX drivers properly too)
2) Uninstall whatever device is sharing an IRQ with the GFX (but obviously not if it’s something vital)
3) Shut down the computer
4) Remove all power from your motherboard, and the battery — use the power switch to drain any caps
5) Physically remove the GFX card and other PCI/PCIe devices from the machine
6) Boot to BIOS and load defaults
7) Boot to Windows
Shut down again and power off normally to re-install the GFX card
9) Boot to Windows and install the basic GFX drivers, without GFExperience. Just GFX, PhysX, HDAudio, and then test it out
10) If that is all good, add the remaining devices back, if any
Note: if you have no onboard GFX/Video, then you have to have it installed for step 7 onwards… In this case, make a note of the key-presses needed to load BIOS defaults, and do that step blind (F10 and enter to save, when it starts rebooting again, hit the power button to turn off before it loads Windows). However you do it, you need at least one power-on without the GFX card before re-adding it
There just has to be a better way than this? I already have an idea in mind that I’m going to test tomorrow. I will go into my bios and just randomly disable USB ports until I find the specific ones that are tied to this host controller (usually motherboards split the USB slots into 2 or 3 sets, meaning you’ll have 2 or 3 IRQ assignments in total just for USB host controllers). This seems much more simple than doing all the steps listed in the solution. Though I don’t know if it will solve the problem.
Куратор(ы):
eLfiK
| Автор | Сообщение | |||
|---|---|---|---|---|
|
||||
Member Статус: Не в сети |
Обсуждение проблем ОС и оборудования: задержка реакции системы (latency), микроcтаттер, инпутлаг, фризы. Осуществлять мониторинг программой Latency Monitor нужно в течение 1 минуты, в состоянии простоя системы т.е. без дисковой, сетевой активности, и любой другой, с выключенным ав и приложениями в трее и автозагрузке, Презентация NVIDIA (на англ.) о проблемах статтеров, фризов и лагов (терминология, описание и причины возникновения) Последний раз редактировалось iG0Lka 07.02.2018 15:48, всего редактировалось 8 раз(а). |
Добавлено: 12.10.2016 12:14 | Реклама | |
|
Партнер |
|
Phenomenum |
|
|
Member Статус: Не в сети |
yylseijfvlu0 Железо напиши, пожалуйста. Все, что есть. И ответь на все мои вопросы. |
|
Y0! |
|
|
Member Статус: Не в сети |
Привет. BIOS сброшен в дефолт, вручную не изменено ничего, версия 5851 (последняя от ASUS). Винда — десятка корпоративная 21H1 (не сборка) из образа с обновлениями от июня 2021, не изменены никакие настройки (разве что убрана галочка с повышенной точности указателя мыши), не удалено ничего, из драйверов установлен только графический 471.22 без geforce experience (чистая установка), никакие другие драйверы не ставил, так что стоит только то, что было встроено в винду. В предыдущий раз пробовал ставить все самые новые драйверы, но ничего это не исправило, потому сейчас поставил только видео. Показатели LatencyMon (снято во время работы CS:GO) какие-то слишком пугающие. Interrupt to user process latency Код: _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ Note: reported execution times may be calculated based on a fixed reported CPU speed. Disable variable speed settings like Intel Speed Step and AMD Cool N Quiet in the BIOS setup for more accurate results. _________________________________________________________________________________________________________ Highest measured interrupt to process latency (µs): 2988,40 Highest measured interrupt to DPC latency (µs): 2985,0 _________________________________________________________________________________________________________ Highest ISR routine execution time (µs): 259,614807 Highest reported total ISR routine time (%): 0,266056 Total time spent in ISRs (%) 0,269609 ISR count (execution time <250 µs): 760454 _________________________________________________________________________________________________________ Highest DPC routine execution time (µs): 6000,838297 Highest reported total DPC routine time (%): 0,122476 Total time spent in DPCs (%) 0,189475 DPC count (execution time <250 µs): 1577122 _________________________________________________________________________________________________________ NOTE: some processes were hit by hard pagefaults. If these were programs producing audio, they are likely to interrupt the audio stream resulting in dropouts, clicks and pops. Check the Processes tab to see which programs were hit. Process with highest pagefault count: steamservice.exe Total number of hard pagefaults 877 _________________________________________________________________________________________________________ Interrupt to DPC latency Код: _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ Note: reported execution times may be calculated based on a fixed reported CPU speed. Disable variable speed settings like Intel Speed Step and AMD Cool N Quiet in the BIOS setup for more accurate results. _________________________________________________________________________________________________________ Highest measured interrupt to DPC latency (µs): 4408,0 _________________________________________________________________________________________________________ Highest ISR routine execution time (µs): 263,462288 Highest reported total ISR routine time (%): 0,360598 Total time spent in ISRs (%) 0,365003 ISR count (execution time <250 µs): 707634 _________________________________________________________________________________________________________ Highest DPC routine execution time (µs): 6007,821876 Highest reported total DPC routine time (%): 0,170995 Total time spent in DPCs (%) 0,254326 DPC count (execution time <250 µs): 1458428 _________________________________________________________________________________________________________ NOTE: some processes were hit by hard pagefaults. If these were programs producing audio, they are likely to interrupt the audio stream resulting in dropouts, clicks and pops. Check the Processes tab to see which programs were hit. Process with highest pagefault count: steam.exe Total number of hard pagefaults 41 _________________________________________________________________________________________________________ Kernel timer latency Код: _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ Note: reported execution times may be calculated based on a fixed reported CPU speed. Disable variable speed settings like Intel Speed Step and AMD Cool N Quiet in the BIOS setup for more accurate results. _________________________________________________________________________________________________________ Highest measured kernel timer latency (µs): 26668,10 _________________________________________________________________________________________________________ Highest ISR routine execution time (µs): 291,035903 Highest reported total ISR routine time (%): 0,363446 Total time spent in ISRs (%) 0,367381 ISR count (execution time <250 µs): 1001708 _________________________________________________________________________________________________________ Highest DPC routine execution time (µs): 5995,026997 Highest reported total DPC routine time (%): 0,174894 Total time spent in DPCs (%) 0,263101 DPC count (execution time <250 µs): 1984892 _________________________________________________________________________________________________________ NOTE: some processes were hit by hard pagefaults. If these were programs producing audio, they are likely to interrupt the audio stream resulting in dropouts, clicks and pops. Check the Processes tab to see which programs were hit. Process with highest pagefault count: steam.exe Total number of hard pagefaults 50 _________________________________________________________________________________________________________ |
|
Dimonas |
|
|
Member Статус: Не в сети |
Y0! писал(а): оказатели LatencyMon (снято во время работы CS:GO) какие-то слишком пугающие. Показатели надо снимать в простое. FAQ читал? |
|
Y0! |
|
|
Member Статус: Не в сети |
Dimonas так лучше?.. Вложение:
Вложение:
В msconfig была снята галочка с элементов автозагрузки, скрыты службы Майкрософт и отключены все прочие службы, сразу после загрузки рабочего стола был запущен последний LatencyMon. Проблема проявляется в микрофризах в CS:GO, BFV, некоторые из них, возможно, появляются перед проигрыванием любых внутриигровых звуков. Тестирование проводилось как с разнообразными твиками (которые могут помочь при первом запуске после их применения, но потом всё возвращается), так и на девственно чистой корпоративной десятке, драйверы — самые-самые новые, пробовал не ставить драйвер аудио, мыши, видеодрайвер всегда ставится без GeForce Experience, как в одиночку, так и в сочетании с комплектным HD аудиодрайвером, USBC и PhysX. Пробовал разные USB порты. Около года назад значительно уменьшить количество микрофризов помогала неустановка драйвера мыши, но сейчас и это ничего не меняет. |
|
Dimonas |
|
|
Member Статус: Не в сети |
Y0! писал(а): Проблема проявляется в микрофризах в CS:GO, BFV, некоторые из них, возможно, появляются перед проигрыванием любых внутриигровых звуков. По скринам трудно сказать в чем у тебя проблема. Драйверы мыши? Без них никак? И что за драйверы VMware в системе? Винда точно свежеустановлена? |
|
Y0! |
|
|
Member Статус: Не в сети |
Dimonas можно и без драйвера мыши, вот сейчас его нет, но это ничего к лучшему не меняет. Возможно, это аппаратная проблема — винда, на которую из дров я поставил только видео, а также Steam (для тестирования фризов), умудрилась упасть в BSOD 1000007e (ntfs.sys), когда Steam скачивал CS:GO. В системе два диска — 860 evo (винда и всё, что на неё установлено) и обычный винчестер 7200 об., первый вчера сканировал samsung magician’ом, второй — chkdsk /f /r, ошибок не найдено. |
|
Y0! |
|
|
Member Статус: Не в сети |
Вложение:
|
|
BOBKOC |
|
|
Member Статус: В сети |
так то криминала не видно, но можешь поэкспериментить с переводом в msi режим видюхи и аудио. ребутаться ток надо при каждом применении + проверять применилось ли |
|
Y0! |
|
|
Member Статус: Не в сети |
BOBKOC один из твиков заключался в том, чтобы всё перевести в msi и назначить высокий приоритет всему, кроме AMD PSP и сетевухи. Также прерывания видеокарты вешал на последние два потока. Также менял настройки HKLMSOFTWAREMicrosoftWindows NTCurrentVersionMultimediaSystemProfile (SystemResponsiveness dword:00000000, Scheduling Category & SFIO Priority — high, Priority dword:00000006) и HKLMSYSTEMControlSet001ControlPriorityControl (Win32PrioritySeparation dword:00000006). Также ставил системный таймер в 0,5 мс и ставил автоочищение standby list (программой ISLC). Всё это вместе могло добавить до 9% средних fps в CS:GO, но фризы полностью не исчезают. После этих твиков может при первом запуске чувствоваться очень низкая задержка мыши, но это почему-то быстро пропадает и потом может случайным образом снова проявляться, закономерности не замечено. |
|
Dimonas |
|
|
Member Статус: Не в сети |
Y0! Таймеры и приоритеты прерываний можно не крутить было. По моим наблюдениям это мало что дает в играх. А вот с фризами системы в первую очередь надо проверять связку процессор-память. Попробовать обновить биос материнки до последней версии с новой AGESA, поставить XMP тайминги и CMD 2T для стабильности. Я лично наблюдал ситуацию, когда в CMD 1T все тесты памяти без ошибок, а в PUBG фризы были. Поставил CMD 2T и проблема ушла. |
|
Y0! |
|
|
Member Статус: Не в сети |
Dimonas BIOS, как указывал, самый новый. Пробовал ставить всё в авто и тогда ставится режим 2400 CL19, попробовал прописать то же, что прописывается в D.O.C.P., но с 2T вместо 1T. 2T применилось, но фризит по-прежнему. Сейчас прерывания не изменены, я просто написал то, что тестировал в недалёком прошлом. Интересно, как всё же выявить корень этой проблемы? |
|
OLD Hunter |
|
|
Member Статус: Не в сети |
А что за сборка ПК, БП вот это все, где? |
|
Y0! |
|
|
Member Статус: Не в сети |
Применил частично твики по мануалу — фиг что изменилось… OLD Hunter Привет, снова заполнил. Фиксы отключены. Фризы были с самого начала этой сборки (всё куплено с нуля, кроме мыши, это и помогло выявить первичную проблему, как тогда казалось — неустановка Synapse 2 сводила фризы до самого минимума, но непроизвольное нажатие боковых кнопок всякий раз заставляло вернуться к Synapse и терпеть; потом были попытки разогнать мой «отборный» B-die, о них я писал тут в соответствующей теме, фризов от такого разгона становилось только больше; какое-то время назад фризы мне надоели и я начал писать в техподдержку NVIDIA, на тот момент винда уже была обновлена до 21H1). Техподдержка NVIDIA расписалась в собственном бессилии что-либо продиагностировать — сказали проверить видеокарту в другом компьютере, посоветовали обратиться к разработчику игры (и не важно, что игр две и они вообще разные, как и их разработчики). Что ж, когда выброшу этот комп на помойку и куплю новый — может, исправлю фризы. |
|
OLD Hunter |
|
|
Member Статус: Не в сети |
Y0! Хм, тогда остается только в СЦ обратиться, железо же на гарантии еще? |
|
Y0! |
|
|
Member Статус: Не в сети |
OLD Hunter эта серия БП была в списке рекомендуемых тут в соответствующей теме, возможно, она и сейчас там. |
|
BOBKOC |
|
|
Member Статус: В сети |
Y0! писал(а): сказали & посоветовали а может те win11 установить ради эксперимента |
|
Dimonas |
|
|
Member Статус: Не в сети |
Y0! |
|
OLD Hunter |
|
|
Member Статус: Не в сети |
Видео бы еще увидеть этих фризов, ну т.е с графиком фреймтайма. Ну у странно что они не во всех играх. На движке фростайбта они есть? Можно скачать что-то по типу бф3 или 4 , да любой бф, можно и более поздние, может и лиц есть. Там есть встроенный график производительности, там график и цпу и ГП, т.е можно точно определить откуда фриз — ЦП или ГП. |
|
Y0! |
|
|
Member Статус: Не в сети |
Вложение: Комментарий к файлу: Вот скриншоты с моим комментарием на ломаном английском (в архиве).
Я упомянул две игры, в которые я играл — CS:GO и BF5. В обеих — фризы. |
—
Кто сейчас на конференции |
|
Сейчас этот форум просматривают: нет зарегистрированных пользователей и гости: 4 |
| Вы не можете начинать темы Вы не можете отвечать на сообщения Вы не можете редактировать свои сообщения Вы не можете удалять свои сообщения Вы не можете добавлять вложения |
Лаборатория
Новости
Содержание
- Высокие задержки DPC в видеокартах NVIDIA GTX 10-х серий и метод решения этой проблемы
- реклама
- реклама
- реклама
- реклама
- Nvidia windows kernel mode driver dpc latency
- DPC Latency Caused by Nvidia Drivers
- NVIDIA Windows Kernel Module Driver high latency causing in game microstutter.
- nevermore15
- High DPC Latency (nvlddmkm.sys, dxgkrnl.sys, Wdf01000.sys, ACPI.sys)
- levider
Высокие задержки DPC в видеокартах NVIDIA GTX 10-х серий и метод решения этой проблемы
До сих пор еще, очень много пользователей имеют в своих компьютерах видеокарты Nvidia 10-ой серии GTX 1060, GTX 1070, GTX 1080. Их владельцы по причине массового безумства, называемого майнингом, так и не смогли их поменять на видеокарты 20 и 30 серии по известным причинам. Потому данная серия видеокарт на сегодняшний день является актуальной для пользователей. И я не являюсь исключением, у меня есть компьютер с видеокартой PALIT JETSTREAM GTX 1070.
реклама
Майнинг нас всех вынудил сделать шаг назад по поколениям комплектующих. В начале 2020 года я думал, что моя видеокарта доживает уже последние дни своей актуальности. А сейчас понимаю, что она полностью актуальна на сегодняшний день, в связи со сложившейся ситуацией на рынке видеокарт.
В этой статье хочу поделиться с читателями о том, как компания NVIDIA в очередной раз «села в лужу» с видеокартами 10-й серии, причем конкретно так села, и до сих пор так и не смогла устранить имеющиеся с ними проблемы. Видеокарты Nvidia 10 серии прославились фризами и подвисаниями в играх, несмотря на высокие значения FPS при этом. Разбираться с этим вопросом я стал после того, как сам столкнулся с такими проблемами на своей видеокарте GTX 1070. Причем изменение настроек в любых их вариантах проблему не решило, фризы никуда не исчезли. Единственная настройка в драйвере видеокарты (в панели управления NVIDIA), которая хотя бы несколько уменьшила фризы, это включение на «ультра» режима низкой задержки.
реклама
Ответ компании NVIDIA, что проблема решена..
реклама
Последующая жалоба в компанию NVIDIA о нерешенной проблеме.
Никакие программные ухищрения к устранению данной проблемы не привели. А драйвера начиная с версии 386.95 с хост-фиксом, это такая себе полумера, для отвода глаз, которая проблему окончательно так и не решила, и после выпуска этого драйвера жалобы так и не прекратились. У меня вообще сложилось впечатление, что компания NVIDIA специально, осознанно «завалила» этот параметр задержек, чтобы за счет его ухудшения улучшить другие, по которым оценивается производительность в синтетических тестах, таким образом, для значительного повышения FPS они и создали эту неправильную работу системных таймеров прерываний. И получается, что в тестах все прекрасно, а играть невозможно.
Данную проблему еще усугубляет «количество подготовленных кадров», опция этого параметра так же находится в панели управления NVIDIA. Подготовленные кадры, так сказать кадры, хранящиеся в запасе, необходимы для того, чтобы поступление их на рендеринг в видеокарту от центрального процессора было непрерывным, даже если центральный процессор «затупит», и не успеет подготовить очередной кадр в срок по какой-либо причине. Эта функция не позволяет видеокарте простаивать в момент отсутствия очередного кадра от центрального процессора, кадр берется из запаса. Так же эта функция повышает FPS (количество кадров в секунду), но она значительно увеличивает задержки DPC, и для уменьшения задержек необходимо эту функцию отключить, или свести к минимуму количество подготовленных кадров.
Приведу для примера измерение задержек в своем компьютере с помощью программы «LatencyMon». Она в режиме реального времени проводит измерение задержек, и в случае превышения задержек, в скобках указывает виновника.
реклама
В первом случае отключаю драйвер видеокарты NVIDIA путем отключения устройства «NVIDIA GeForce GTX 1070» в диспетчере устройств, видеокарта в этом случае будет работать без драйвера, и провожу измерение задержек.
И по результату измерений видно, что программа «LatencyMon» ничего касающегося видеокарты, как виновника больших задержек не указала. А указала сетевой драйвер (network driver interface specification).
Теперь включаю драйвер видеокарты NVIDIA путем включения устройства «NVIDIA GeForce GTX 1070» в диспетчере устройств, и провожу повторно измерение.
И теперь видно, что программа «LatencyMon» указала виновника в больших задержках, которые составляют 551 мкс. И это драйвер видеокарты NVIDIA версии 465.89.
Что бы хотелось посоветовать для уменьшения эффекта фризов в видеокартах NVIDIA GTX 10-х серий:
1. В настройках панели управления NVIDIA, в режиме управления электропитанием установить режим максимальной производительности.
2. В настройках панели управления NVIDIA, режим низкой задержки обязательно включить на «ультра».
3. В настройках панели управления NVIDIA, максимальное количество заранее подготовленных кадров установить на минимум.
Сталкивались ли вы с такой проблемой в видеокартах NVIDIA GTX 10-х серий? И как вы их решили? Пишите в комментариях.
Источник
Nvidia windows kernel mode driver dpc latency
Как уже не раз говорилось – не у всех и не всегда настройки из данного гайда дадут хорошие результаты в плане FPS, но в первую очередь он предназначен для уменьшения задержек с чем он должен справится у большинства.
Все тесты сделаны в чистой Windows (всё настроено и используется по-умолчанию), но с обновлёнными драйверами nVidia, а в качестве оппонента выступает та же Windows, но уже настроенная по гайду. Следует учитывать, что как и все тесты они не показывают полноты всей картины, разбежки между запусками это нормально, поэтому приведены средние результаты.
Версии используемые в тестах:
Благодаря nVidia тут замечательно видно, что основу моих задержек составляют именно её драйверы (к слову, на предыдущей версии результаты были получше), а борьба с этим отдельный вид искусства. Так же эти результаты не показывают идеальный вариант, а весьма наглядно демонстрируют что улучшилось и с чем еще предстоит бороться.
‼️ Рекомендуется после обновления драйверов nVidia смотреть с помощью LatencyMon не стало ли хуже и, весьма вероятно, имеет смысл откатить драйвер на предыдущую версию.
Результаты LatencyMon после настройки:
Результаты LatencyMon до настройки:
Результаты LatencyMon после настройки:
Результаты LatencyMon до настройки:
Для оценки можно придерживаться примерно таких оптимальных цифр в результатах статистики:
Если ваши значения не больше чем в два раза, то есть смысл для более детального решения проблемы – поиска проблемных устройств или драйверов (о чём будет ниже). Если же ваши значения больше на несколько порядков, то что-то в вашей системе явно не так.
Результаты LatencyMon после настройки:
Результаты LatencyMon до настройки:
На этой вкладке можно увидеть какие именно драйверы влияют на задержки и исходя из показателей искать решение.
Основной список системных драйверов:
Результаты LatencyMon после настройки:
Результаты LatencyMon до настройки:
Здесь можно более подробно посмотреть как распараллелены драйверы по ядрам и их загрузка.
В качестве дополнительной оптимизации так же можно попробовать перенести драйвер сетевого адаптера на другое ядро, но по-хорошему, кроме печальной или не очень статистики, с этим мы вряд ли можем что-то сделать не угробив в дальнейшем Windows.
Для облизывания на FPS мы будем использовать популярный синтетический стресс-тест FurMark [скачать] для оценки производительности в Open GL.
Результаты FurMark после настройки:
Результаты FurMark до настройки:
✨ Кроме того, не стоит пренебрегать такими решениями как MSI Afterburner и настройками в BIOS для увеличение производительности GPU и CPU. Поэтому результаты в играх при правильном подходе будут показывать еще больший прирост в производительности.
Источник
DPC Latency Caused by Nvidia Drivers
| Processor | AMD Ryzen 3700X |
|---|---|
| Motherboard | Asus X570 TUF Gaming Plus |
| Cooling | NZXT Kraken X62 |
| Memory | G.Skill 2x8GB 3600CL16 |
| Video Card(s) | Asus Strix RTX 2070 |
| Storage | Samsung 850 SSD 500GB |
| Display(s) | Acer Predator XB271HU |
| Case | NZXT S340 Elite |
I dont edit audio, but I am after tweaking my system. I’ve used DPC Latency Checker to measure the systems latency and found it is most likely caused by the Nvidia drivers. Nvidia has addressed this in the past and claims it was fixed.
Setup: AMD 3700X set to 4.0Ghz constant, with C-States disabled. GPU RTX 2070 no OC. Fresh install o Windows 10 2004 build 19041.421, only Windows10 default drivers and Nvidia Driver 451.85 is installed. Tried also moded W10 with removed bloatware (Superlite edition). Also removed all possibly AMD, Intel, Nvidia by Windows installed drivers with Display Driver Uninstaller and installed Nvidia Driver only, without PhysX, HD Audio, Experience, USB-C.
When I disable the Nvidia Driver in Device Manager
The latency is a lot lower
I did not try older Nvidia drivers as I like the Max Frame Rate option in the newer driver versions. As I dont have to run Riva Tuner or any additional software to have a fixed max frame rate.
I’ve ruled out HPET On/Off tweaking as the latency result was reproducible when disabling the Nvidia driver.
Источник
NVIDIA Windows Kernel Module Driver high latency causing in game microstutter.
nevermore15
Active member
Hi there, I’ve been having issues for around 4 months where all my videogames stutter and micro stutter to some degree. It’s been really frustrating and I hope that one of y’all can help. I’ve tried countless guides and I’ve even gone to Microcenter to have the techs look at it just for them to tell me that there’s nothing that they can do. I used latency mon and I’ve attatched the results. Sorry if something in my post is wrong. I’m not too experienced with this stuff. It seems that its mostly the NVIDIA Kernel Driver during gaming, and the Network driver interface specification during normal use. Please let me know if I need to post more information I just want to get this over with.
Specs: ASUS ROG B450F Gaming II
Zotac RTX 3070
Ryzen 3600x
16gb corsair vengeance white 3200mhz
500gb Nvme m.2 WD black
2tb WD Hard Drive
750 watt PSU
_________________________________________________________________________________________________________
CONCLUSION
_________________________________________________________________________________________________________
Your system appears to be suitable for handling real-time audio and other tasks without dropouts.
LatencyMon has been analyzing your system for 0:02:14 (h:mm:ss) on all processors.
_________________________________________________________________________________________________________
SYSTEM INFORMATION
_________________________________________________________________________________________________________
Computer name: DESKTOP-ITVKR7D
OS version: Windows 10, 10.0, version 2009, build: 19042 (x64)
Hardware: System Product Name, System manufacturer
CPU: AuthenticAMD AMD Ryzen 5 3600X 6-Core Processor
Logical processors: 12
Processor groups: 1
RAM: 16294 MB total
_________________________________________________________________________________________________________
CPU SPEED
_________________________________________________________________________________________________________
Reported CPU speed: 380 MHz
Note: reported execution times may be calculated based on a fixed reported CPU speed. Disable variable speed settings like Intel Speed Step and AMD Cool N Quiet in the BIOS setup for more accurate results.
_________________________________________________________________________________________________________
MEASURED INTERRUPT TO USER PROCESS LATENCIES
_________________________________________________________________________________________________________
The interrupt to process latency reflects the measured interval that a usermode process needed to respond to a hardware request from the moment the interrupt service routine started execution. This includes
the scheduling and execution of a DPC routine, the signaling of an event and the waking up of a usermode thread from an idle wait state in response to that event.
Highest measured interrupt to process latency (µs): 796.10
Average measured interrupt to process latency (µs): 7.373617
Highest measured interrupt to DPC latency (µs): 791.90
Average measured interrupt to DPC latency (µs): 3.786315
_________________________________________________________________________________________________________
REPORTED ISRs
_________________________________________________________________________________________________________
Interrupt service routines are routines installed by the OS and device drivers that execute in response to a hardware interrupt signal.
Total time spent in ISRs (%) 0.007531
ISR count (execution time =4000 µs): 0
_________________________________________________________________________________________________________
REPORTED DPCs
_________________________________________________________________________________________________________
DPC routines are part of the interrupt servicing dispatch mechanism and disable the possibility for a process to utilize the CPU while it is interrupted until the DPC has finished execution.
Total time spent in DPCs (%) 0.848527
DPC count (execution time =4000 µs): 0
_________________________________________________________________________________________________________
REPORTED HARD PAGEFAULTS
_________________________________________________________________________________________________________
Hard pagefaults are events that get triggered by making use of virtual memory that is not resident in RAM but backed by a memory mapped file on disk. The process of resolving the hard pagefault requires
reading in the memory from disk while the process is interrupted and blocked from execution.
NOTE: some processes were hit by hard pagefaults. If these were programs producing audio, they are likely to interrupt the audio stream resulting in dropouts, clicks and pops. Check the Processes tab to see
which programs were hit.
Process with highest pagefault count: msmpeng.exe
Total number of hard pagefaults 3006
Hard pagefault count of hardest hit process: 1077
Number of processes hit: 58
_________________________________________________________________________________________________________
PER CPU DATA
_________________________________________________________________________________________________________
CPU 0 Interrupt cycle time (s): 13.877046
CPU 0 ISR highest execution time (µs): 94.790
CPU 0 ISR total execution time (s): 0.001418
CPU 0 ISR count: 439
CPU 0 DPC highest execution time (µs): 947.220
CPU 0 DPC total execution time (s): 5.532492
CPU 0 DPC count: 155462
_________________________________________________________________________________________________________
CPU 1 Interrupt cycle time (s): 10.036351
CPU 1 ISR highest execution time (µs): 80.410
CPU 1 ISR total execution time (s): 0.001399
CPU 1 ISR count: 1022
CPU 1 DPC highest execution time (µs): 883.020
CPU 1 DPC total execution time (s): 3.329040
CPU 1 DPC count: 112728
_________________________________________________________________________________________________________
CPU 2 Interrupt cycle time (s): 8.029588
CPU 2 ISR highest execution time (µs): 85.170
CPU 2 ISR total execution time (s): 0.008390
CPU 2 ISR count: 2299
CPU 2 DPC highest execution time (µs): 914.610
CPU 2 DPC total execution time (s): 2.778771
CPU 2 DPC count: 67206
_________________________________________________________________________________________________________
CPU 3 Interrupt cycle time (s): 2.713610
CPU 3 ISR highest execution time (µs): 103.330
CPU 3 ISR total execution time (s): 0.018220
CPU 3 ISR count: 17726
CPU 3 DPC highest execution time (µs): 572.630
CPU 3 DPC total execution time (s): 0.736754
CPU 3 DPC count: 32188
_________________________________________________________________________________________________________
CPU 4 Interrupt cycle time (s): 2.126310
CPU 4 ISR highest execution time (µs): 106.160
CPU 4 ISR total execution time (s): 0.030711
CPU 4 ISR count: 17578
CPU 4 DPC highest execution time (µs): 490.790
CPU 4 DPC total execution time (s): 0.564582
CPU 4 DPC count: 28795
_________________________________________________________________________________________________________
CPU 5 Interrupt cycle time (s): 1.290346
CPU 5 ISR highest execution time (µs): 117.050
CPU 5 ISR total execution time (s): 0.019130
CPU 5 ISR count: 10471
CPU 5 DPC highest execution time (µs): 196.340
CPU 5 DPC total execution time (s): 0.268027
CPU 5 DPC count: 16862
_________________________________________________________________________________________________________
CPU 6 Interrupt cycle time (s): 0.817404
CPU 6 ISR highest execution time (µs): 108.40
CPU 6 ISR total execution time (s): 0.009544
CPU 6 ISR count: 2585
CPU 6 DPC highest execution time (µs): 164.870
CPU 6 DPC total execution time (s): 0.087005
CPU 6 DPC count: 12797
_________________________________________________________________________________________________________
CPU 7 Interrupt cycle time (s): 0.760451
CPU 7 ISR highest execution time (µs): 8.870
CPU 7 ISR total execution time (s): 0.000302
CPU 7 ISR count: 52
CPU 7 DPC highest execution time (µs): 110.340
CPU 7 DPC total execution time (s): 0.026813
CPU 7 DPC count: 6535
_________________________________________________________________________________________________________
CPU 8 Interrupt cycle time (s): 0.979627
CPU 8 ISR highest execution time (µs): 102.610
CPU 8 ISR total execution time (s): 0.020786
CPU 8 ISR count: 37079
CPU 8 DPC highest execution time (µs): 123.650
CPU 8 DPC total execution time (s): 0.064216
CPU 8 DPC count: 14415
_________________________________________________________________________________________________________
CPU 9 Interrupt cycle time (s): 0.803844
CPU 9 ISR highest execution time (µs): 5.090
CPU 9 ISR total execution time (s): 0.001536
CPU 9 ISR count: 2253
CPU 9 DPC highest execution time (µs): 162.520
CPU 9 DPC total execution time (s): 0.049710
CPU 9 DPC count: 9496
_________________________________________________________________________________________________________
CPU 10 Interrupt cycle time (s): 0.950035
CPU 10 ISR highest execution time (µs): 93.940
CPU 10 ISR total execution time (s): 0.003605
CPU 10 ISR count: 4844
CPU 10 DPC highest execution time (µs): 128.990
CPU 10 DPC total execution time (s): 0.091793
CPU 10 DPC count: 18429
_________________________________________________________________________________________________________
CPU 11 Interrupt cycle time (s): 1.026593
CPU 11 ISR highest execution time (µs): 108.970
CPU 11 ISR total execution time (s): 0.006137
CPU 11 ISR count: 8964
CPU 11 DPC highest execution time (µs): 132.330
CPU 11 DPC total execution time (s): 0.124589
CPU 11 DPC count: 18301
_________________________________________________________________________________________________________
Источник
High DPC Latency (nvlddmkm.sys, dxgkrnl.sys, Wdf01000.sys, ACPI.sys)
levider
Member
Hello guys
I have huge problem with high DPC latency on my Omen laptop
1. I reinstalled Nvidia drivers (now without PhysX, only main driver)
2. I tried to disable Device Manager > Battery > something called battery control to solve ACPI.sys problem. And it worked, but my second monitor doesn’t work :/
_________________________________________________________________________________________________________
CONCLUSION
_________________________________________________________________________________________________________
Your system appears to be having trouble handling real-time audio and other tasks. You are likely to experience buffer underruns appearing as drop outs, clicks or pops. One or more DPC routines that belong to a driver running in your system appear to be executing for too long. One problem may be related to power management, disable CPU throttling settings in Control Panel and BIOS setup. Check for BIOS updates.
LatencyMon has been analyzing your system for 0:02:42 (h:mm:ss) on all processors.
_________________________________________________________________________________________________________
CPU SPEED
_________________________________________________________________________________________________________
Reported CPU speed: 2496 MHz
Note: reported execution times may be calculated based on a fixed reported CPU speed. Disable variable speed settings like Intel Speed Step and AMD Cool N Quiet in the BIOS setup for more accurate results.
_________________________________________________________________________________________________________
MEASURED INTERRUPT TO USER PROCESS LATENCIES
_________________________________________________________________________________________________________
The interrupt to process latency reflects the measured interval that a usermode process needed to respond to a hardware request from the moment the interrupt service routine started execution. This includes the scheduling and execution of a DPC routine, the signaling of an event and the waking up of a usermode thread from an idle wait state in response to that event.
Highest measured interrupt to process latency (µs): 3244,301037
Average measured interrupt to process latency (µs): 13,158952
Highest measured interrupt to DPC latency (µs): 3236,506182
Average measured interrupt to DPC latency (µs): 3,636769
_________________________________________________________________________________________________________
REPORTED ISRs
_________________________________________________________________________________________________________
Interrupt service routines are routines installed by the OS and device drivers that execute in response to a hardware interrupt signal.
Total time spent in ISRs (%) 0,012927
ISR count (execution time =4000 µs): 0
_________________________________________________________________________________________________________
REPORTED DPCs
_________________________________________________________________________________________________________
DPC routines are part of the interrupt servicing dispatch mechanism and disable the possibility for a process to utilize the CPU while it is interrupted until the DPC has finished execution.
Total time spent in DPCs (%) 0,422731
DPC count (execution time =4000 µs): 0
_________________________________________________________________________________________________________
REPORTED HARD PAGEFAULTS
_________________________________________________________________________________________________________
Hard pagefaults are events that get triggered by making use of virtual memory that is not resident in RAM but backed by a memory mapped file on disk. The process of resolving the hard pagefault requires reading in the memory from disk while the process is interrupted and blocked from execution.
NOTE: some processes were hit by hard pagefaults. If these were programs producing audio, they are likely to interrupt the audio stream resulting in dropouts, clicks and pops. Check the Processes tab to see which programs were hit.
Process with highest pagefault count: latmon.exe
Total number of hard pagefaults 598
Hard pagefault count of hardest hit process: 422
Number of processes hit: 15
_________________________________________________________________________________________________________
PER CPU DATA
_________________________________________________________________________________________________________
CPU 0 Interrupt cycle time (s): 10,463318
CPU 0 ISR highest execution time (µs): 491,963942
CPU 0 ISR total execution time (s): 0,078858
CPU 0 ISR count: 6069
CPU 0 DPC highest execution time (µs): 2323,189904
CPU 0 DPC total execution time (s): 2,611212
CPU 0 DPC count: 278151
_________________________________________________________________________________________________________
CPU 1 Interrupt cycle time (s): 2,031970
CPU 1 ISR highest execution time (µs): 90,870994
CPU 1 ISR total execution time (s): 0,003447
CPU 1 ISR count: 637
CPU 1 DPC highest execution time (µs): 413,272436
CPU 1 DPC total execution time (s): 0,070354
CPU 1 DPC count: 3115
_________________________________________________________________________________________________________
CPU 2 Interrupt cycle time (s): 2,661673
CPU 2 ISR highest execution time (µs): 22,421474
CPU 2 ISR total execution time (s): 0,001042
CPU 2 ISR count: 242
CPU 2 DPC highest execution time (µs): 279,509615
CPU 2 DPC total execution time (s): 0,021926
CPU 2 DPC count: 1688
_________________________________________________________________________________________________________
CPU 3 Interrupt cycle time (s): 2,487784
CPU 3 ISR highest execution time (µs): 3,746795
CPU 3 ISR total execution time (s): 0,000444
CPU 3 ISR count: 198
CPU 3 DPC highest execution time (µs): 409,529647
CPU 3 DPC total execution time (s): 0,036582
CPU 3 DPC count: 2471
_________________________________________________________________________________________________________
Driver file Description ISR count DPC count Highest execution (ms) Total execution (ms) Image base Image size Company Product Version Path
————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————
Источник
Is there any way i can reduce the latency ? Is this the normal latency ? Since my cpu doesn’t have integrated igpu. My gpu driver version 456.55 tried the latest driver but it cause it even more up to 10. Thanks
My system:
Ryzen 5 3600
Rtx 2060
B450 tomhawk max
512 gb transcend ssd
corsair cx650 bronze psu
16 gigs ram 3200 mhz
Joined Aug 28, 2012
·
86 Posts
This spike is happening when the p-state changes. I wouldn’t worry about it as the same thing happens with my RTX 3090. P-state changes always have a DPC spike.
Joined Jul 10, 2011
·
8,712 Posts
This spike is happening when the p-state changes. I wouldn’t worry about it as the same thing happens with my RTX 3090. P-state changes always have a DPC spike.
Yeah, this is only a problem if you want it to be one. However, if this is related to p-states, setting power management to prefer maximum performance should solve it, right?
Joined Oct 18, 2020
·
84 Posts
Discussion Starter
·
#4
·
Oct 26, 2020
This spike is happening when the p-state changes. I wouldn’t worry about it as the same thing happens with my RTX 3090. P-state changes always have a DPC spike.
I am not getting spikes tho. Like when i am using chrome playing video or playing cod mw my latency stays 1.2+
Joined Jun 10, 2014
·
11,665 Posts
Nvidia has a problem with DPC latency every single launch of new GPUs as they update their drivers, break crap ton of things in process and fail to realize that driver latency matters and they bloat it up.
And yes alt+tab etc. makes this tools fairly useless as you need to see live values not just max, max often happens during changes of window focus, change of state, …
Joined Nov 24, 2011
·
3,764 Posts
Thats normal. Been a issue since Maxwell drivers and they just dont care.
Joined Dec 28, 2007
·
11,681 Posts
Is this an AMD only issue?
No issues on my Maximus 12 extreme and 3090.
Joined Jan 25, 2010
·
6,662 Posts
Is this an AMD only issue?
No issues on my Maximus 12 extreme and 3090.View attachment 2463537
its probably more a per system scenario, i wouldnt doubt it. either that or we all put our tinfoil hats on, and say intel is paying off nvidia
also, to OP: i am doubting, in any real world scenario, that your going to notice that 1ms delay, if it was 4 or 5 you might.
Joined Aug 28, 2012
·
86 Posts
Is this an AMD only issue?
No issues on my Maximus 12 extreme and 3090.View attachment 2463537
You ran it for a minute and did nothing. If he did the same thing his would look like yours. Now with LatencyMon running go open a Twitch stream in Google Chrome or a game and watch the spike happen. Post what yours looks like and I bet you it looks the same. P-state changes cause the spike. It is normal.
Joined Apr 3, 2016
·
703 Posts
Bullshit.
I was running Heaven benchmark at the same time. I do not like nor appreciate when people accuse me of something I didn’t do.
Here. Take this and shove it. Sorry for knowing how to maintain my system correctly.View attachment 2463614
I have no trouble achieving low DPC latency on nvlddmkm either.
My power management is fully enabled in BIOS. These are 30+ minute runs. Even on balanced the numbers are not terrible either.
On balance mode
On Ultimate Performance
-
2.5 MB
Views: 101 -
2.5 MB
Views: 137
Joined Nov 3, 2017
·
20 Posts
Disable sysmain service.. restart windows and try latencymon again
Joined Dec 28, 2007
·
11,681 Posts
I have no trouble achieving low DPC latency on nvlddmkm either.
My power management is fully enabled in BIOS. These are 30+ minute runs. Even on balanced the numbers are not terrible either.On balance mode
View attachment 2463627
On Ultimate Performance
View attachment 2463628
Thank you.
It’s sad that some people think that just because someone has a problem, that everyone else must have a problem, and you literally have to record 30 minute videos of multiple games and benchmarks running at the same time to convince them it’s their system that’s at fault, not ours
Joined Aug 28, 2012
·
86 Posts
Thank you.
It’s sad that some people think that just because someone has a problem, that everyone else must have a problem, and you literally have to record 30 minute videos of multiple games and benchmarks running at the same time to convince them it’s their system that’s at fault, not ours![:)]( "Smile :)")
You are a clown you still haven’t posted what I asked since I mentioned it was specific to the RTX 3090. You are confusing me with the OP. I do not have any latency issues as well.
Joined Nov 3, 2017
·
20 Posts
Is there any way i can reduce the latency ? Is this the normal latency ? Since my cpu doesn’t have integrated igpu. My gpu driver version 456.55 tried the latest driver but it cause it even more up to 10. Thanks
My system:
Ryzen 5 3600
Rtx 2060
B450 tomhawk max
512 gb transcend ssd
corsair cx650 bronze psu
16 gigs ram 3200 mhzView attachment 2463490
View attachment 2463493
Disable sysmain service.. restart windows and try latencymon again
Joined Jun 10, 2014
·
11,665 Posts
The DPC latency as far as I remember is a major problem for «real time» users such as audio engineers and why they may avoid Nvidia GPUs entirely, cherry pick an OS or entire platform.
Sometimes the DPC latency was so bad it can cause problems even for regular users.
All you can do is submit bug reports directly at Nvidia’s site and complain to the internet, trying to get attention and Nvidia noticing that it’s being talked negatively about for a reason and fix that reason.
Back when Pascal launched, I’ve had the high DPC too, everyone pretty much suffers from it with new NV GPUs or even just new generation/major update of NV drivers.
Joined Jul 5, 2020
·
31 Posts
M12 extreme + 3090 strix oc and have this problem big time.
first it was acpi.sys breaking havoc and after stopping monitoring the EC sensors on the mb it resolved but nvlddmkm.sys dpc is still crazy high.
Joined Apr 24, 2012
·
4,359 Posts
The DPC latency as far as I remember is a major problem for «real time» users such as audio engineers and why they may avoid Nvidia GPUs entirely, cherry pick an OS or entire platform.
Sometimes the DPC latency was so bad it can cause problems even for regular users.All you can do is submit bug reports directly at Nvidia’s site and complain to the internet, trying to get attention and Nvidia noticing that it’s being talked negatively about for a reason and fix that reason.
Back when Pascal launched, I’ve had the high DPC too, everyone pretty much suffers from it with new NV GPUs or even just new generation/major update of NV drivers.
it also affects frametimes ALOT…
- This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Typically my computer averages somewhere in the low teen’s to low thirties according to Latencymon. However, I noticed recently that I’ve been getting massive spikes in latency related to dxgkrnl.sys and nvlddmkm.sys. My BIOS is updated to the most recent release and I’ve uninstalled all the useless bloatware that comes with Win 10 (Cortana, Game Bar etc..) Can anyone help me figure out what the issue is? Here’s the Latencymon breakdown from earlier today:
_________________________________________________________________________________________________________
CONCLUSION
_________________________________________________________________________________________________________
Your system appears to be having trouble handling real-time audio and other tasks. You are likely to experience buffer underruns appearing as drop outs, clicks or pops. One or more DPC routines that belong to a driver running in your system appear to be executing for too long. Also one or more ISR routines that belong to a driver running in your system appear to be executing for too long. One problem may be related to power management, disable CPU throttling settings in Control Panel and BIOS setup. Check for BIOS updates.
LatencyMon has been analyzing your system for 5:07:39 (h:mm:ss) on all processors.
_________________________________________________________________________________________________________
SYSTEM INFORMATION
_________________________________________________________________________________________________________
Computer name: MY-PC
OS version: Windows 10 , 10.0, build: 18363 (x64)
Hardware: ASUSTeK COMPUTER INC., Z170-A
CPU: GenuineIntel Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz
Logical processors: 8
Processor groups: 1
RAM: 16322 MB total
_________________________________________________________________________________________________________
CPU SPEED
_________________________________________________________________________________________________________
Reported CPU speed: 4008 MHz
Note: reported execution times may be calculated based on a fixed reported CPU speed. Disable variable speed settings like Intel Speed Step and AMD Cool N Quiet in the BIOS setup for more accurate results.
WARNING: the CPU speed that was measured is only a fraction of the CPU speed reported. Your CPUs may be throttled back due to variable speed settings and thermal issues. It is suggested that you run a utility which reports your actual CPU frequency and temperature.
_________________________________________________________________________________________________________
MEASURED INTERRUPT TO USER PROCESS LATENCIES
_________________________________________________________________________________________________________
The interrupt to process latency reflects the measured interval that a usermode process needed to respond to a hardware request from the moment the interrupt service routine started execution. This includes the scheduling and execution of a DPC routine, the signaling of an event and the waking up of a usermode thread from an idle wait state in response to that event.
Highest measured interrupt to process latency (µs): 2679.50
Average measured interrupt to process latency (µs): 2.151481
Highest measured interrupt to DPC latency (µs): 2675.0
Average measured interrupt to DPC latency (µs): 0.724937
_________________________________________________________________________________________________________
REPORTED ISRs
_________________________________________________________________________________________________________
Interrupt service routines are routines installed by the OS and device drivers that execute in response to a hardware interrupt signal.
Highest ISR routine execution time (µs): 1241.473802
Driver with highest ISR routine execution time: dxgkrnl.sys — DirectX Graphics Kernel, Microsoft Corporation
Highest reported total ISR routine time (%): 0.001431
Driver with highest ISR total time: dxgkrnl.sys — DirectX Graphics Kernel, Microsoft Corporation
Total time spent in ISRs (%) 0.001628
ISR count (execution time <250 µs): 111676
ISR count (execution time 250-500 µs): 0
ISR count (execution time 500-999 µs): 2824
ISR count (execution time 1000-1999 µs): 10
ISR count (execution time 2000-3999 µs): 0
ISR count (execution time >=4000 µs): 0
_________________________________________________________________________________________________________
REPORTED DPCs
_________________________________________________________________________________________________________
DPC routines are part of the interrupt servicing dispatch mechanism and disable the possibility for a process to utilize the CPU while it is interrupted until the DPC has finished execution.
Highest DPC routine execution time (µs): 3467.046906
Driver with highest DPC routine execution time: nvlddmkm.sys — NVIDIA Windows Kernel Mode Driver, Version 445.87 , NVIDIA Corporation
Highest reported total DPC routine time (%): 0.001196
Driver with highest DPC total execution time: iaStorA.sys — Intel(R) Rapid Storage Technology driver — x64, Intel Corporation
Total time spent in DPCs (%) 0.005295
DPC count (execution time <250 µs): 1519989
DPC count (execution time 250-500 µs): 0
DPC count (execution time 500-999 µs): 8426
DPC count (execution time 1000-1999 µs): 166
DPC count (execution time 2000-3999 µs): 37
DPC count (execution time >=4000 µs): 0
_________________________________________________________________________________________________________
REPORTED HARD PAGEFAULTS
_________________________________________________________________________________________________________
Hard pagefaults are events that get triggered by making use of virtual memory that is not resident in RAM but backed by a memory mapped file on disk. The process of resolving the hard pagefault requires reading in the memory from disk while the process is interrupted and blocked from execution.
NOTE: some processes were hit by hard pagefaults. If these were programs producing audio, they are likely to interrupt the audio stream resulting in dropouts, clicks and pops. Check the Processes tab to see which programs were hit.
Process with highest pagefault count: msmpeng.exe
Total number of hard pagefaults 36177
Hard pagefault count of hardest hit process: 20966
Number of processes hit: 28
_________________________________________________________________________________________________________
PER CPU DATA
_________________________________________________________________________________________________________
CPU 0 Interrupt cycle time (s): 226.784686
CPU 0 ISR highest execution time (µs): 1241.473802
CPU 0 ISR total execution time (s): 2.179280
CPU 0 ISR count: 113478
CPU 0 DPC highest execution time (µs): 3076.739521
CPU 0 DPC total execution time (s): 6.233969
CPU 0 DPC count: 1249865
_________________________________________________________________________________________________________
CPU 1 Interrupt cycle time (s): 240.373923
CPU 1 ISR highest execution time (µs): 882.187375
CPU 1 ISR total execution time (s): 0.224942
CPU 1 ISR count: 1032
CPU 1 DPC highest execution time (µs): 3467.046906
CPU 1 DPC total execution time (s): 1.214712
CPU 1 DPC count: 180483
_________________________________________________________________________________________________________
CPU 2 Interrupt cycle time (s): 200.412169
CPU 2 ISR highest execution time (µs): 0.0
CPU 2 ISR total execution time (s): 0.0
CPU 2 ISR count: 0
CPU 2 DPC highest execution time (µs): 624.690619
CPU 2 DPC total execution time (s): 0.118454
CPU 2 DPC count: 48920
_________________________________________________________________________________________________________
CPU 3 Interrupt cycle time (s): 216.404346
CPU 3 ISR highest execution time (µs): 0.0
CPU 3 ISR total execution time (s): 0.0
CPU 3 ISR count: 0
CPU 3 DPC highest execution time (µs): 606.038174
CPU 3 DPC total execution time (s): 0.032019
CPU 3 DPC count: 2455
_________________________________________________________________________________________________________
CPU 4 Interrupt cycle time (s): 200.096209
CPU 4 ISR highest execution time (µs): 0.0
CPU 4 ISR total execution time (s): 0.0
CPU 4 ISR count: 0
CPU 4 DPC highest execution time (µs): 419.529691
CPU 4 DPC total execution time (s): 0.061491
CPU 4 DPC count: 18421
_________________________________________________________________________________________________________
CPU 5 Interrupt cycle time (s): 242.843178
CPU 5 ISR highest execution time (µs): 0.0
CPU 5 ISR total execution time (s): 0.0
CPU 5 ISR count: 0
CPU 5 DPC highest execution time (µs): 454.436627
CPU 5 DPC total execution time (s): 0.027757
CPU 5 DPC count: 4247
_________________________________________________________________________________________________________
CPU 6 Interrupt cycle time (s): 210.318140
CPU 6 ISR highest execution time (µs): 0.0
CPU 6 ISR total execution time (s): 0.0
CPU 6 ISR count: 0
CPU 6 DPC highest execution time (µs): 796.492016
CPU 6 DPC total execution time (s): 0.066206
CPU 6 DPC count: 20103
_________________________________________________________________________________________________________
CPU 7 Interrupt cycle time (s): 236.110250
CPU 7 ISR highest execution time (µs): 0.0
CPU 7 ISR total execution time (s): 0.0
CPU 7 ISR count: 0
CPU 7 DPC highest execution time (µs): 620.879990
CPU 7 DPC total execution time (s): 0.065337
CPU 7 DPC count: 4124
_________________________________________________________________________________________________________
Question ndis.sys & nvlddmkm.sys — DPC exception causing severe latency
-
Thread starterMazzimo
-
Start dateJun 17, 2022
-
- Nov 21, 2013
-
- 12
-
- 0
-
- 18,510
- 0
-
#1
Below is info from LatencyMon;
Conclusion: Your system seems to be having difficulty handling real-time audio and other tasks. You may experience drop outs, clicks or pops due to buffer underruns. One or more DPC routines that belong to a driver running in your system appear to be executing for too long. At least one detected problem appears to be network related. In case you are using a WLAN adapter, try disabling it to get better results. One problem may be related to power management, disable CPU throttling settings in Control Panel and BIOS setup. Check for BIOS updates.
https://ibb.co/RQz7D2N
Last edited: Jun 23, 2022
-
- Jun 12, 2015
-
- 61,162
-
- 5,187
-
- 166,290
- 10,454
-
#2
- right click desktop
- choose show more options
- Nvidia control panel
- Manage 3d settings
- Scroll down right hand screen to Power Management Mode
- Set it to Prefer Maximum Power
- save changes
video might help
-
- Nov 21, 2013
-
- 12
-
- 0
-
- 18,510
- 0
-
#3
see if this helps
- right click desktop
- choose show more options
- Nvidia control panel
- Manage 3d settings
- Scroll down right hand screen to Power Management Mode
- Set it to Prefer Maximum Power
- save changes
video might help
Thank you for the reply!
I’ve done so and haven’t seen changes yet. I’ll check the video though.
-
- Nov 21, 2013
-
- 12
-
- 0
-
- 18,510
- 0
-
#4
Conclusion: Your system seems to be having difficulty handling real-time audio and other tasks. You may experience drop outs, clicks or pops due to buffer underruns. One or more DPC routines that belong to a driver running in your system appear to be executing for too long. At least one detected problem appears to be network related. In case you are using a WLAN adapter, try disabling it to get better results. One problem may be related to power management, disable CPU throttling settings in Control Panel and BIOS setup. Check for BIOS updates.
https://ibb.co/RQz7D2N
Last edited: Jun 17, 2022
-
- Jun 12, 2015
-
- 61,162
-
- 5,187
-
- 166,290
- 10,454
-
#5
Did you post about the problem on the insiders forum as things like this are why Insiders exists, so problems can be identified before release.
Its likely others have same problem.
which build of win 11
- right click start
- choose run…
- type winver and press enter
- My current build is 22000.739 but I know Insiders is different?
-
- Nov 21, 2013
-
- 12
-
- 0
-
- 18,510
- 0
-
#6
Issue has gotten more difficult to navigate through Windows. Gaming and productivity nearly impossible
[URL unfurl=»true»]https://ibb.co/Rb8M6h5[/URL]
Last edited: Jun 23, 2022
-
- Nov 21, 2013
-
- 12
-
- 0
-
- 18,510
- 0
-
- Jun 12, 2015
-
- 61,162
-
- 5,187
-
- 166,290
- 10,454
-
#8
what are specs of the PC?
latest chipset drivers?
Last edited: Jun 24, 2022
-
#9
This is a fresh installation of Windows 11, build 22000.856
I’ve tried the following:
- Set power to high performance mode
- Disabled onboard audio & wifi in the bios
- Reinstalling NVIDIA drivers through the geforce experience, both game-ready and studio
CONCLUSION
_
Your system appears to be having trouble handling real-time audio and other tasks. You are likely to experience buffer underruns appearing as drop outs, clicks or pops. One or more DPC routines that belong to a driver running in your system appear to be executing for too long. One problem may be related to power management, disable CPU throttling settings in Control Panel and BIOS setup. Check for BIOS updates.
LatencyMon has been analyzing your system for 0:04:02 (h🇲🇲ss) on all processors.
_
SYSTEM INFORMATION
_
Computer name: LUKE-PC
OS version: Windows 11, 10.0, version 2009, build: 22000 (x64)
Hardware: Z390 DESIGNARE, Gigabyte Technology Co., Ltd.
BIOS: F9
CPU: GenuineIntel Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz
Logical processors: 16
Processor groups: 1
Processor group size: 16
RAM: 32630 MB total
_
CPU SPEED
_
Reported CPU speed (WMI): 360 MHz
Reported CPU speed (registry): 360 MHz
Note: reported execution times may be calculated based on a fixed reported CPU speed. Disable variable speed settings like Intel Speed Step and AMD Cool N Quiet in the BIOS setup for more accurate results.
_
MEASURED INTERRUPT TO USER PROCESS LATENCIES
_
The interrupt to process latency reflects the measured interval that a usermode process needed to respond to a hardware request from the moment the interrupt service routine started execution. This includes the scheduling and execution of a DPC routine, the signaling of an event and the waking up of a usermode thread from an idle wait state in response to that event.
Highest measured interrupt to process latency (µs): 14607.0
Average measured interrupt to process latency (µs): 15.584596
Highest measured interrupt to DPC latency (µs): 14604.30
Average measured interrupt to DPC latency (µs): 12.352655
_
REPORTED ISRs
_
Interrupt service routines are routines installed by the OS and device drivers that execute in response to a hardware interrupt signal.
Highest ISR routine execution time (µs): 118.223333
Driver with highest ISR routine execution time: Wdf01000.sys — Kernel Mode Driver Framework Runtime, Microsoft Corporation
Highest reported total ISR routine time (%): 0.001713
Driver with highest ISR total time: Wdf01000.sys — Kernel Mode Driver Framework Runtime, Microsoft Corporation
Total time spent in ISRs (%) 0.001713
ISR count (execution time <250 µs): 66112
ISR count (execution time 250-500 µs): 0
ISR count (execution time 500-1000 µs): 0
ISR count (execution time 1000-2000 µs): 0
ISR count (execution time 2000-4000 µs): 0
ISR count (execution time >=4000 µs): 0
_
REPORTED DPCs
_
DPC routines are part of the interrupt servicing dispatch mechanism and disable the possibility for a process to utilize the CPU while it is interrupted until the DPC has finished execution.
Highest DPC routine execution time (µs): 1942.404444
Driver with highest DPC routine execution time: nvlddmkm.sys — NVIDIA Windows Kernel Mode Driver, Version 516.94 , NVIDIA Corporation
Highest reported total DPC routine time (%): 0.020555
Driver with highest DPC total execution time: Wdf01000.sys — Kernel Mode Driver Framework Runtime, Microsoft Corporation
Total time spent in DPCs (%) 0.061993
DPC count (execution time <250 µs): 225223
DPC count (execution time 250-500 µs): 0
DPC count (execution time 500-10000 µs): 152
DPC count (execution time 1000-2000 µs): 7
DPC count (execution time 2000-4000 µs): 0
DPC count (execution time >=4000 µs): 0
_
REPORTED HARD PAGEFAULTS
_
Hard pagefaults are events that get triggered by making use of virtual memory that is not resident in RAM but backed by a memory mapped file on disk. The process of resolving the hard pagefault requires reading in the memory from disk while the process is interrupted and blocked from execution.
NOTE: some processes were hit by hard pagefaults. If these were programs producing audio, they are likely to interrupt the audio stream resulting in dropouts, clicks and pops. Check the Processes tab to see which programs were hit.
Process with highest pagefault count: msmpeng.exe
Total number of hard pagefaults 622
Hard pagefault count of hardest hit process: 125
Number of processes hit: 29
_
PER CPU DATA
_
CPU 0 Interrupt cycle time (s): 7.205096
CPU 0 ISR highest execution time (µs): 118.223333
CPU 0 ISR total execution time (s): 0.066352
CPU 0 ISR count: 66112
CPU 0 DPC highest execution time (µs): 1942.404444
CPU 0 DPC total execution time (s): 2.251743
CPU 0 DPC count: 200802
_
CPU 1 Interrupt cycle time (s): 0.286504
CPU 1 ISR highest execution time (µs): 0.0
CPU 1 ISR total execution time (s): 0.0
CPU 1 ISR count: 0
CPU 1 DPC highest execution time (µs): 63.262778
CPU 1 DPC total execution time (s): 0.003358
CPU 1 DPC count: 832
_
CPU 2 Interrupt cycle time (s): 0.722162
CPU 2 ISR highest execution time (µs): 0.0
CPU 2 ISR total execution time (s): 0.0
CPU 2 ISR count: 0
CPU 2 DPC highest execution time (µs): 376.265556
CPU 2 DPC total execution time (s): 0.073818
CPU 2 DPC count: 10000
_
CPU 3 Interrupt cycle time (s): 0.290657
CPU 3 ISR highest execution time (µs): 0.0
CPU 3 ISR total execution time (s): 0.0
CPU 3 ISR count: 0
CPU 3 DPC highest execution time (µs): 45.470
CPU 3 DPC total execution time (s): 0.003508
CPU 3 DPC count: 703
_
CPU 4 Interrupt cycle time (s): 0.503655
CPU 4 ISR highest execution time (µs): 0.0
CPU 4 ISR total execution time (s): 0.0
CPU 4 ISR count: 0
CPU 4 DPC highest execution time (µs): 88.396111
CPU 4 DPC total execution time (s): 0.017133
CPU 4 DPC count: 3341
_
CPU 5 Interrupt cycle time (s): 0.272908
CPU 5 ISR highest execution time (µs): 0.0
CPU 5 ISR total execution time (s): 0.0
CPU 5 ISR count: 0
CPU 5 DPC highest execution time (µs): 41.743889
CPU 5 DPC total execution time (s): 0.001701
CPU 5 DPC count: 304
_
CPU 6 Interrupt cycle time (s): 0.414402
CPU 6 ISR highest execution time (µs): 0.0
CPU 6 ISR total execution time (s): 0.0
CPU 6 ISR count: 0
CPU 6 DPC highest execution time (µs): 376.646111
CPU 6 DPC total execution time (s): 0.010197
CPU 6 DPC count: 1965
_
CPU 7 Interrupt cycle time (s): 0.354692
CPU 7 ISR highest execution time (µs): 0.0
CPU 7 ISR total execution time (s): 0.0
CPU 7 ISR count: 0
CPU 7 DPC highest execution time (µs): 62.744444
CPU 7 DPC total execution time (s): 0.002937
CPU 7 DPC count: 471
_
CPU 8 Interrupt cycle time (s): 0.406060
CPU 8 ISR highest execution time (µs): 0.0
CPU 8 ISR total execution time (s): 0.0
CPU 8 ISR count: 0
CPU 8 DPC highest execution time (µs): 84.4250
CPU 8 DPC total execution time (s): 0.006884
CPU 8 DPC count: 1479
_
CPU 9 Interrupt cycle time (s): 0.292163
CPU 9 ISR highest execution time (µs): 0.0
CPU 9 ISR total execution time (s): 0.0
CPU 9 ISR count: 0
CPU 9 DPC highest execution time (µs): 53.614444
CPU 9 DPC total execution time (s): 0.001081
CPU 9 DPC count: 231
_
CPU 10 Interrupt cycle time (s): 0.375734
CPU 10 ISR highest execution time (µs): 0.0
CPU 10 ISR total execution time (s): 0.0
CPU 10 ISR count: 0
CPU 10 DPC highest execution time (µs): 93.336667
CPU 10 DPC total execution time (s): 0.004743
CPU 10 DPC count: 997
_
CPU 11 Interrupt cycle time (s): 0.218088
CPU 11 ISR highest execution time (µs): 0.0
CPU 11 ISR total execution time (s): 0.0
CPU 11 ISR count: 0
CPU 11 DPC highest execution time (µs): 345.411111
CPU 11 DPC total execution time (s): 0.001883
CPU 11 DPC count: 261
_
CPU 12 Interrupt cycle time (s): 0.366573
CPU 12 ISR highest execution time (µs): 0.0
CPU 12 ISR total execution time (s): 0.0
CPU 12 ISR count: 0
CPU 12 DPC highest execution time (µs): 83.841111
CPU 12 DPC total execution time (s): 0.014868
CPU 12 DPC count: 2680
_
CPU 13 Interrupt cycle time (s): 0.260754
CPU 13 ISR highest execution time (µs): 0.0
CPU 13 ISR total execution time (s): 0.0
CPU 13 ISR count: 0
CPU 13 DPC highest execution time (µs): 44.847222
CPU 13 DPC total execution time (s): 0.000708
CPU 13 DPC count: 112
_
CPU 14 Interrupt cycle time (s): 0.365882
CPU 14 ISR highest execution time (µs): 0.0
CPU 14 ISR total execution time (s): 0.0
CPU 14 ISR count: 0
CPU 14 DPC highest execution time (µs): 73.653889
CPU 14 DPC total execution time (s): 0.005681
CPU 14 DPC count: 1034
_
CPU 15 Interrupt cycle time (s): 0.214438
CPU 15 ISR highest execution time (µs): 0.0
CPU 15 ISR total execution time (s): 0.0
CPU 15 ISR count: 0
CPU 15 DPC highest execution time (µs): 78.587222
CPU 15 DPC total execution time (s): 0.001194
CPU 15 DPC count: 170
_
-
- Jun 12, 2015
-
- 61,162
-
- 5,187
-
- 166,290
- 10,454
-
#10
Often updating sound, lan, or GPU drivers can help.
Updating BIOS as well.
I don’t like commenting on these posts as I had/have a latency issue and everytime I think its fixed I comment on a thread like this and it happens again (fingers crossed).
- Advertising
- Cookies Policies
- Privacy
- Term & Conditions
- Topics
Page 6 of 7
-
Smough
Master Guru- Messages:
- 936
- Likes Received:
- 281
- GPU:
Well then, then just don’t come up and reply with «nonsense» and «beliefs» when someone tries to link stuff so you can see their point of view and why they sustain this or that argument. You have given me zero sources or videos so I can say «oh snap, he is right about this particular thing». The only ones who say you should disable HPET on the BIOS are «guides» made by random people who guess its ok to disable it. That’s it. I showed you a segment of a video where an user actually explains why disabling HPET on the BIOS isn’t a good idea. And what do you do? Saying that I should be banned. Just lol. And as if you haven’t offended me indirectly, because of course saying what I posted its a «superstitious belief in magic numbers» is so very nice of you. Yet disabling HPET on BIOS isn’t. And without explaining any of the two.
Last edited: Feb 16, 2020
-
It disappears from device manager and Linux doesn’t list it as available clock source anymore. Simple.
Other simple thing: I have a flat frame time graph in lots of demanding games like BFV or Hitman 2. What else to prove? Do you also expect somebody to fly you into space so you can start accepting that earth isn’t flat? -
Smough
Master Guru- Messages:
- 936
- Likes Received:
- 281
- GPU:
You have a severe attitude problem. And your frametime with v-sync off wasn’t flat whatsoever; it had fluctuations. Use RTSS Framerate limiter, try it for once if you never have and you’ll see a real flat frametime line graph. Also, turn on v-sync and disable any framerate limiter and go to Exclusive Fullscreen mode. Look at the frametime disaster of never being 16.6 ms. This is why you actually need RTSS, nVidia Framerate limiter or FSO at games that work properly with them active.
-
What you need to prove isn’t that turning it off doesn’t hurt anything. Relax.
You need to prove turning it off does anything at all. Other than cure your OCD. Otherwise, you’re also doing things for no reason. -
As you can see, it’s not there:
https://abload.de/img/hpetmissingm9kon.png
Running without it for probably almost two years now. No weird crashes, no weird stutter etc. -
I’m not sure how i can be more clear.
Ok, turn hpet back on and tell me if you notice any difference.
If there is no difference, then you’re changing something for no reason, right? Potentially, just like Smough and his cmd commands.
-
Wrong. hwinfo uses HPET by default. It likely can do so via kernel driver, despite of not being the kernel’s default clock source. If hwinfo driver can do it, other drivers can do it too.
Just because there are no problems, it doesn’t mean there will never be any. It has been proven (even the anandtech link mentions it) that HPET causes harm these days. And I’ve already pointed out several times that there are no advantages over TSC in reality (or someone proves otherwise with a real example than can be reproduced, lengthy YT channels of «gamers» talking clever are not my source for education), so it’s perfectly rational to turn HPET off in bios.Edit: The alternative would be that you are smarter than Linux kernel devs, which I doubt.
-
Smough
Master Guru- Messages:
- 936
- Likes Received:
- 281
- GPU:
-
Lol no offense. There’s no reason to have a horse in this race.
It’s just, neither you or @aufkrawall2 have proven taking any action is beneficial.
Your evidence is anecdotal (it feels so much smoother), his is paranoia (just because there are no problems, it doesn’t mean there will never be any).Oh, and hwinfo using hpet is useful. Doesn’t hurt anything. In fact, BenchMate uses hwinfo’s hpet implementation to verify overclocks. Which, like aufkrawall pointed out, isn’t a mainstream thing to use, but it certainly doesn’t hurt anything when it does it.
Last edited: Feb 17, 2020
-
Martigen
Master Guru- Messages:
- 517
- Likes Received:
- 241
- GPU:
While I haven’t investigated it, I really don’t think that’s the case. I’m pretty sure Windows does set this on demand, -and- applications can request a higher rate as well. There is no way every developer for every program would even know about setting a higher resolution timer, and even then they cannot know what else is running on your system and what impact this would have. That would likely go against recommended guidelines. Windows itself would have a much better of idea based on types of processes and load, even if it can be forced on an individual basis.
-
Smough
Master Guru- Messages:
- 936
- Likes Received:
- 281
- GPU:
Its all good, I get what you are saying. Maybe I shouldn’t even disable them, but I keep doing it. I will re-enable them and I will do some tests at some point.
-
@Martigen That’s what dynamic tick is for but even fully dynticks / no_hz_full systems have a tickratre (I think). Something like a «base» rate or a maximum rate…
-
AsiJu
Ancient Guru- Messages:
- 8,234
- Likes Received:
- 2,952
- GPU:
Guys could you offer some insight on this, ie:
— no latency spikes when idle, notable spikes when gaming, acceptable or not and indeed can some other «stall» like loading assets be visible in latencymon?
EDIT: happpens when eg. running Media Player and browsing (Firefox) as well. Frack…
(high DPC latency pointing to NVIDIA Windows Kernel Mode driver).
Are NV drivers again having latency issues? Using 436.30 now but I had this with 442.01 as well and worse.
I tried disabling interrupt moderation completely now for network adapter. As of typing this:so ISR/DPC atm somewhat fine but high interrupt to process latency recorded.
Last edited: Feb 17, 2020
-
mbk1969
Ancient Guru- Messages:
- 13,450
- Likes Received:
- 11,662
- GPU:
You can ignore it. Do not forget that Latencymon is aimed at audio processing/recording people where almost real-time efficiency is needed.
-
AsiJu
Ancient Guru- Messages:
- 8,234
- Likes Received:
- 2,952
- GPU:
Really? Thanks.
As it happens, since posting that image, I wasn’t able to reproduce that spike even when doing exact same things (Media Player, Firefox, UPlay, Task Manager open).
I’m not one to be obsessed with Latencymon but like said lately I’ve had clear stalling in Division 2 that I thought was remediated by tweaking interrupt moderation for network adapter (it’s a constantly online game so makes sense in that regard).
However for curiosity left Latencymon running while playing RAGE 2 and there were again notable «drop-outs» in places, not nearly as severe as in Division 2 but notable.
I wasnt sure if that was just asset streaming or something else, hence these questions here.I suppose any glitch or stall anywhere can cause a high reading in Latencymon? Meaning even if the said stall does not originate from ISR/DPC execution time or interrupt to process latency?
(Example: UAC prompts create a reproducible spike in Latencymon. Guess it again makes sense as the system is waiting for user input).
-
AsiJu
Ancient Guru- Messages:
- 8,234
- Likes Received:
- 2,952
- GPU:
Follow up post:
running Ryzen High Performance power plan (of course), installed latest chipset drivers and using latest BIOS (w/ AGESA 1004b).
I just reinstalled chipset drivers and that seems to have properly disabled core parking now w/ High perf. power plan according to Ryzen Master.
Previously, even with High perf. power plan, unused cores went to «Sleep». Which was kind of strange to me anyway.I’m not sure if 1909 update (applied yesterday) + reinstalling chipset drivers fixed this or what happened exactly. Perhaps this was the source of the issues (doubt it a bit but who knows).
If I still get lag / stalls in games I’ll clean install Windows, been too long anyway.
-
If you check latencyMon site it says to test with nothing running, not with a load.
-
I once benchmarked Crysis 3 (already on Windows 10) when CPU bound and if I recall correctly, it really stayed at 15ms without Origin running and CPU performance becoming very crippled. I might try again at some point so we can know for sure about whether it really stays at 15 or just something slightly better.
-
AsiJu
Ancient Guru- Messages:
- 8,234
- Likes Received:
- 2,952
- GPU:
Yeah, should have checked that site… RTFM
thanks a lot!
EDIT: correction about Ryzen Master / core Sleep: still happens when just idling at desktop.
(Prior I had at least Media Player running which apparently created enough load to keep all cores active).Surely by design it’s like that, but to avoid spreading misinformation wanted to clarify this.
Last edited: Feb 17, 2020
-
X7007
Ancient Guru- Messages:
- 1,816
- Likes Received:
- 64
- GPU:
I learned that the computer time is drifting and also windows somehow checking your timing and timer support every time you reboot the computer and system load up.
We are talking about the Newest CPU’s with TSC-InvariantWe need to fix 3 issues.
CPU time Drift — If you don’t restart your computer every day you are going to have drifted, I saw computers at work drifting 59 Seconds after a month or so! — The fix is installing NTP software to keep timing — Works fine I tested it
https://time.is/
https://www.meinbergglobal.com/english/sw/ntp.htmThe program will use its own service instead of windows time, upkeeping the time all the time, fixing the drift at any giving moment or not letting it to happen.
RTC is important timer that needs to be sync with the OS because the CPU doesn’t know what is real until it is SYNC, as we know power adjustments and drifting will occur no matter what.
1 second in real life needs to be 1 second in the CPU, or else weird things will happen. TSC does a good job but can’t do it 100% long enough.Windows Power Saving/Bios Power Saving — Only use specific C State, for example, AMD it’s called Global State State, for me it’s automatic, it doesn’t interrupt performance or cause issue. Disable EIST and Cool & Quiet though. and use Maximum power Management profile with minimum CPU 100% and disable Use suspend
Making sure we are using the best for a specific CPU. what I mean it’s not performance, it’s about precision. for example, I can’t be sure if for AMD 1950x because it has NUMA or UMA which has 2 separate CPU’s inside 1 means they need to SYNC. now Tscsyncpolicy /Enhanced /Legacy /Default I want to know what it is actually using. because I can’t be sure if windows really really know which one to choose. maybe bug? maybe something else… we know that sometimes something doesn’t work. And we already know that we do need to add BCDEDIT command no matter what, DisabledynamicTick Yes
On gigabyte motherboard, I do have SpreadSpecturm on CPU that I can’t disable. look as follow V Vhttps://www.kernel.org/doc/Documentation/virtual/kvm/timekeeping.txt
3.3) TSC and multi-socket / NUMAMulti-socket systems, especially large multi-socket systems are likely to have
individual clocksources rather than a single, universally distributed clock.
Since these clocks are driven by different crystals, they will not have
perfectly matched frequency, and temperature and electrical variations will
cause the CPU clocks, and thus the TSCs to drift over time. Depending on the
exact clock and bus design, the drift may or may not be fixed in absolute
error, and may accumulate over time.In addition, very large systems may deliberately slew the clocks of individual
cores. This technique, known as spread-spectrum clocking, reduces EMI at the
clock frequency and harmonics of it, which may be required to pass FCC
standards for telecommunications and computer equipment.It is recommended not to trust the TSCs to remain synchronized on NUMA or
multiple socket systems for these reasons.I fixed many weird issues I had. including a weird randomly stuttering lag in COD MW 2019 which randomly means if I ran the game it could work fine, but then if I exit and ran again it would start stuttering Heavy, like HDD 5400 rpm run 100% and CPU 100% usage for 5 seconds which what actually happened. investigating with RTSS I saw that when the freeze happens I saw that CPU usage from NODE0 1-15 with SMT goes to 100% between all cores, not all, like 5-6 cores, happens on intervals of 10-20 and sometimes minutes. and randomly when I ran it the second time it would run on NODE1 which is 16-31 and didn’t have any stuttering at all!. After checking again on NODE0 and disabling SMT, the game worked perfectly.
Another issue is Latency NVIDIA when using Power Save or Adaptive, whenever the GPU doesn’t use more than 139mhz or ram there will be heavy latency on the computer because windows use WDDM and need to refresh what you see on screen, 139mhz is not enough. Fixing it is using Maximum performance globally, though it’s not really needed, but if you use Desktop heavily then do so. could also follow F33RT’y using the MSI Burner to set Voltage/Core Maximum Limit with CTRL + F and it should fix may be other issues, but will run 100% all the time.
Somehow after Firmware update in LatencyMon, my devices decided to use other core for ISR 23,24. don’t know why but Core0 is not used for ISR. don’t see any real issues with that, but it interesting why it make so on the firmware, bug or intentionally?
![[IMG]](https://i.imgur.com/cmd8Vye.png)
Page 6 of 7
Share This Page
