Hello, We found that some standard nVidia tests fail with dual nVidia RTX 4090 system. It is CUDA 11.8, driver 520.61.05 running on Linux dl 5.15.0-52-generic #58-Ubuntu SMP Thu Oct 13 08:03:55 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux TEST-1: executables_v2/bin/x86_64/linux/release/simpleP2P [e…

It looks like nVidia bug and unlikely that Bizon or their training could help here. This is the ticket number: “NVIDIA PSIRT” PSIRT@nvidia.com ; Bug report: standard nVidia P2P tests: 3902559

Hi vasilii.shelkov Thank you for the additional information. We can reproduce this issue on our systems. This is under investigation.

Hello @abchauhan – How do I receive an update for this issue? I tried sending email to PSIRT@nvidia.com about issue 3902559, but haven’t hear back yet. Is there a portal I can sign in to see the trace of updates? Thanks

@abchauhan This is a very serious issue and has already been reproduced by a number of people: Problems With RTX4090 MultiGPU and AMD vs Intel vs RTX6000Ada or RTX3090 | Puget Systems It seems new A6000 Ada cards are also affected on AMD CPUs. I have ordered 8 such cards, so I hope this will be fix…

We can not confirm that the RTX 6000 Ada GPUs have this problem on AMD EPYC or WRX80 based CPUs. P2P copy has not to be disabled when using RTX 6000 Ada GPUs. More important: the transfered data is correct. Multi RTX 6000 Ada setups seem to work without problems. Some findings for the multi RTX 40…

I have just run into this issue as well using a 2X 4090 setup with i9-10980XE CPU. I too have been using the simpleP2P test which is broken when p2p is enabled. Doing a bit of hacking, it does appear that cudaMemcpyPeer works as expected. In my case the cudaMemcpyPeer gave 10.54GB/s vs 12.5 Gb/s wh…

@abchauhan Any idea on when we should expect a fix? Can also confirm the same issue with p2p on a multi RTX 4090 setup with AMD WRX80.

Hi all, Apologies for the delay. Feedback from Engineering is that Peer to Peer is not supported on 4090. The applications/driver should not report this configuration as peer to peer capable. The reporting is being fixed and future drivers will report the following instead :- I. # ./simpleP2P [./…

What about other cards? 4080,70,60? new gen A6000 and all others that are coming up?

Standard nVidia CUDA tests fail with dual RTX 4090 Linux box

Graphics / Linux Linux

suprnerd February 21, 2023, 12:53am 13

Is this a hardware or software/firmware issue?

Topic		Replies	Views
TAO5 - Detectnet_v2 - MultiGPU TAO API Stuck - EXTRA GPU TAO Toolkit	13	1473	November 7, 2023
Parallel training with 4 cards 4090 cannot be performed on AMD 5975WX， stuck at the beginning CUDA Programming and Performance	13	6404	February 6, 2023
我用的CPU是AMD 5975WX，显卡是4块4090。cuda版本为cuda12，pytoch版本为2.0 Linux chinese	6	2183	December 29, 2022
One GPU NOT capable of Peer-to-Peer (P2P) CUDA Programming and Performance	22	5834	November 27, 2018
Low P2P GPU bandwidth performance between GeForce GPUs CUDA Programming and Performance	20	2498	October 9, 2024
P2P not working for P600s? CUDA Programming and Performance	7	2003	April 5, 2018
P2P Communication Fails 1080ti->1080ti. IOMMU & ACS disabled Linux	1	1855	February 15, 2020
multi-GPU Peer to Peer access CUDA SDK example not working, why? CUDA Programming and Performance	13	5426	February 26, 2015
Peer access not supported between devices CUDA Programming and Performance	11	7697	November 9, 2017
Partial fail of peer access in 8 Volta GPU instance (p3.16xlarge) on AWS -> huge slowdown CUDA Programming and Performance	32	4214	March 10, 2018

Standard nVidia CUDA tests fail with dual RTX 4090 Linux box

Related topics