- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
We are using an off the shelf, industrial system built by a well known manufacturer. The application is an AI based Machine Vision solution. The system uses an i7-1370PRE processor. We are running a large set of performance tests and are carefully monitoring CPU temperature. All cores are operating within a range of 90C to 65C. Well within spec.
But at around 90C core temp, we see normal operation of the CPU but excessive throttling of the GPU. Our machine learning model uses the GPU very effectively but the high rate of throttling at just 85C was shocking and results in a huge impact to our performance numbers.
Our only recourse was to down clock the GPU from 1.4GHz to 1GHz while keeping the CPU clock at 1.9GHZ. Note that we are not running in turbo mode.
Is there any solution or workaround to the GPU throttling at 90C core temp?
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello machine_vision,
Thank you for posting in Intel Communities.
Base on your inquiry, we have specific forum for these issues or inquiries. I will be transferring this thread for faster response.
Please wait for their reply within 1 to 2 business days.
Best regards,
Michael L.
Intel Customer Support Technician
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
...S
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Scott, Good question. Our only means of verification that the GPU is being throttled is to monitor the temperature & performance of that core. Then to down-clock the GPU till we see consistent performance. When we do this, we do see (in correlation to the down-clocking) that specific core temp reducing. Happy to post all the detail graphs. If you can provide us with a more precise process for confirming GPU throttling, we can check.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
You could use an app like ThrottleStop to verify whether the cause is Temperature, Current or externally (VRM, etc.) related.
...S
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you. As a note, we run Linux, Ubuntu 22.04. Is ThrottleStop available for Linux?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I honestly don't know. Not a Linux expert.
Sorry,
...S
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Appears to be windows only:
https://www.techpowerup.com/download/techpowerup-throttlestop/
Doc (not an Intel employee or contractor)
[Maybe Windows 12 will be better]
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Looks like there is a Linux version:
https://github.com/agoose77/throttlestop
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello @machine_vision,
Thank you for contacting Intel Embedded Community.
You may try with the latest driver for Linux:
https://dgpu-docs.intel.com/installation-guides/index.html
Best regards,
![](/skins/images/7FC17B7B85029576C25F1E43CE255B51/responsive_peak/images/icon_anonymous_message.png)
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page