- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I'm a developer for indie game.
Our products have unstable rendering issue on specific polygon in Intel UHD Graphics.
In GeForce graphics card or software rendering, same product render all polygons normally.
- [Has issue] Intel UHD Graphics 620 in i7-8565U(Driver 27.20.100.8681)
- [Has issue] Intel UHD Graphics in i7-10510U(Driver 31.0.101.1999)
- [No problem] GeForce MX150
- [No problem] GeForce MX250
- [No problem] GeForce RTX2070 SUPER
- [No problem] GeForce GTX1060
- [No problem] Windows Advanced Rasterization Platform(WARP)
I've created code and binary for reproducing. Is there anyone who can check it and describe reasons?
IMHO, This problem is the bug on driver.
Binaries
Attached binaries have unstable rendering in Intel UHD Graphics.
Rendering flicker.mp4: Rendering result by D3D12HelloTriangle.exe on Intel UHD Graphics.
D3D12HelloTriangle.exe: Binary for reproducing which is built from reproduce/rendering-flicker-with-error branch in following repository.
This binary always render 197 polygons, But rendering on 196th polygon is unstable.
Reproducing code
reproduce/rendering-flicker-with-error branch on my repository.
This repository forked from Microsoft official sample repository.
I've just changed vertices in following sample.
Samples/Desktop/D3D12HelloWorld/src/HelloTriangle
Difference of code from official sample is commit 172e5d2c
196th polygon is here.
x1: (61.500000 - 640) / 640
y1: -(61.500000 - 360) / 360
z1: 0.500000
x2: (-62.500000 - 640) / 640
y2: -(61.500000 - 360) / 360
z2: 0.500000
x3: (-62.500000 - 640) / 640
y3: -(-62.500000 - 360) / 360
z3: 0.500000
This polygon will be rendered some frame, and will NOT be rendered another frame.
Unknown reason for fixing
If replace x1 of 196th polygon with this, this issue is fixed by unknown reason.
From: x1: (61.500000 - 640) / 640
To: x1: (60.500000 - 640) / 640
reproduce/rendering-flicker-without-error branch on the same repository.
I'm already contact Intel customer support, but rejected it because this question is about development.
I hope anyone who can describe the reason of this problem or fix it.
Best regards.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi all
Our development team is already aware of this issue and we have shared your most recent findings internally.
Unfortunately, the products (Intel® Core™ i7-8565U= Whiskey Lake and Intel® Core™i7-10510U = Comet Lake) have processor graphics that have been moved to a legacy software support model. During our triage we have also confirmed the issue is not reproducible in latest Intel products.
For more details please have a look at the following article:
https://www.intel.com/content/www/us/en/support/articles/000091662/graphics.html
"Intel will provide software support for affected products on critical fixes and security vulnerabilities only. Software updates for these products will move to a quarterly release cadence or with software releases as needed." FYI
We appreciate the time invested in following up this case along with us and for the time being we will be closing this thread
Thank you for reaching out Intel Game Developer Customer Support
Karen
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
HI @Sheile
Thank you for reaching out the Intel Game Developer Support
We appreciate you sharing all the pertinent information. We will proceed to replicate the issue with the code you're sharing in your repository and if we have questions we will reach out in this post
Additional to all the info you shared, please attach the SSU logs for every system that is showing the failure
https://www.intel.com/content/www/us/en/support/articles/000057926/memory-and-storage.html
It is required for us in case we need to report the bug to the development team
We appreciate your patience in advance, happy holidays
Karen
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you for your replying.
I've attached SSU logs on this thread.
Please check it and replicate with the binary and code.
Best regards.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Sheile !
Thank you for submitting the SSU logs.
Since yesterday I am trying to reproduce the flickering in the 196th poly without success.
I tried in different systems (12th, 11th, 8th gen UHD integrated graphics) and with the driver version where you see the issue and with the latest using the binary files you share and also tried with the example code from your repository without any luck.
Can you reproduce the issue 100% of the times?
Also, can you reproduce it with the latest version of the drivers?
I will continue to try to reproduce the issue incrementing the number of polygons (may be is a memory issue?)
-- Arturo
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Ok! Finally we replicate the issue!
It looks only happen in Intel base laptops.
We will continue the debug with the drivers team, we will be in touch soon
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thank you for replicating the issue and give some solution
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
@Arturo_Intel
Thank you for replication and I’m sorry for delaying response.
Can you reproduce the issue 100% of the times
Yes, I always can reproduce the issue on following laptops with these 197 polygons.
- ASUS ZenBook S13 UX392FN Intel UHD Graphics 620 in i7-8565U(Driver 27.20.100.8681)
- Mouse DIVE4N 20014N-CML-KK Intel UHD Graphics in i7-10510U(Driver 31.0.101.1999)
But I felt that probability of the randomly rendering on this polygon is based on the number of polygons while investigating.
First, I found this issue on our indie game product with more than 2000 polygons.
The same polygon was hardly rendered. (approx. <10%)
I want to decide the cause of polygon, reduce polygons and found this weird behavior.
Reducing polygons affect to the probability of rendering on this specified polygon.
Also, can you reproduce it with the latest version of the drivers?
In my laptop(ASUS ZenBook S13 UX392FN), I've upgrade driver and re-scan SSU logs.
And also my friend's laptop has same issue.
- [Upgrade driver] ASUS ZenBook S13 UX392FN Intel UHD Graphics 620 in i7-8565U(Driver 31.0.101.2114)
- [New] VAIO S15 Intel HD Graphics 530(Driver 26.20.100.7465)
I appreciate your support and the effort of driver team.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi all
Our development team is already aware of this issue and we have shared your most recent findings internally.
Unfortunately, the products (Intel® Core™ i7-8565U= Whiskey Lake and Intel® Core™i7-10510U = Comet Lake) have processor graphics that have been moved to a legacy software support model. During our triage we have also confirmed the issue is not reproducible in latest Intel products.
For more details please have a look at the following article:
https://www.intel.com/content/www/us/en/support/articles/000091662/graphics.html
"Intel will provide software support for affected products on critical fixes and security vulnerabilities only. Software updates for these products will move to a quarterly release cadence or with software releases as needed." FYI
We appreciate the time invested in following up this case along with us and for the time being we will be closing this thread
Thank you for reaching out Intel Game Developer Customer Support
Karen

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page