#780m — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #780m, aggregated by home.social.
-
Is anyone having trouble with an AMD Ryzen Laptop and the latest Linux Kernel? My laptop went haywire and it appears to be a regression. I thought my SSD had failed at first but it seems it might be the embedded GPU, a 780M, causing the trouble. I found this, which appears to be the trouble: https://gitlab.freedesktop.org/drm/amd/-/issues/3388
-
Then your Docker compose container should have:
image-name:
build:
context: .
devices:
- /dev/dri
- /dev/kfd
group_add:
- video
shm_size: 4G
environment:
- PYTORCH_HIP_ALLOC_CONF=garbage_collection_threshold:0.9,max_split_size_mb:512
- PYTORCH_CUDA_ALLOC_CONF=garbage_collection_threshold:0.9,max_split_size_mb:512
- HSA_OVERRIDE_GFX_VERSION=11.0.0 -
在Ryzen 7 #8845HS w/ Radeon #780M 用 #ComfyUI 生圖（ #Linux ）
https://blog.pastwind.org/2025/02/ryzen-7-8845hs-w-radeon-780mcomfyuilinux.html
試了很久才發現成功的方程式…這是因為每次安裝 #ROCm 都需要下載安裝超過30GB的檔案！！！
tl;dr 直接說結論
OS: Ubuntu 22.04（因為ROCm 6.1只支援此以下的版本）
ROCm: <= 6.1.2，6.2跟6.3都沒辦法正常運行
PyTorch: <= 2.4.1，2.5.1版會顯示不支援硬體的警告，圖片有時候無法正確產生。UserWarning: Attempting to use hipBLASLt on an unsupported architecture! Overriding blas backend to hipblas
2.6以上則完全無法正常運行。使用PyTorch官網的版本而不是AMD提供的。
https://pytorch.org/get-started/previous-versions/
ComfyUI: 當前版本v0.3.14可正常運行。 -
So, final update on #IntelArc A380 in #Kubevirt/#Kubernetes.
To put it lightly, the whole thing was a bust. Likely at least partially due to my motherboard (whatever comes with a Dell R730) not supporting resizable bar. To add to that, while Linux will try its best to resize the PCIe BAR even if the BIOS feature isn't enabled, it seems the architecture of the Intel Arc cards blocks this (something to do with the card having a built-in PCIe switch that has its own resizable bar which blocks the actual GPU's bar from being resized... or something?
Everything I've seen online seems to be point to resizable bar being almost a hard requirement for the Arc cards to do any good. So it's perhaps due to not being able to do that that the performance of the card has been absolutely terrible.
#DavinciResolve Studio would barely even run in a Linux VM (which may be as much to do with the vendor than the GPU itself). On Windows the performance was average at best, and the GPU wasn't able to handle rendering to any of the intermediate codecs (like DNxHR or Cineform) without Davinci crashing due to running out of VRAM.
I have parts coming somewhat soon to build a nice tidy 1U server based around the #TopTon #7840HS board, which looks exciting. I'm tempted to pop the GPU in there, but at the same time I'm pretty sure the #780M iGPU actually has better performance than the A380, so I'm probably just going to utilise the iGPU for that.
So might be sending the A380 onto its next (and hopefully Resizable Bar capable) home shortly.
-
👉 Minisforum EM780: un potente mini PC in formato tascabile con Ryzen 7 7840U
Mini PC silenzioso e ultra-compatto con processore AMD Ryzen 7 7840U, piccolo e portatile, potente e con ottima connettività.https://gomoot.com/minisforum-em780-un-potente-mini-pc-in-formato-tascabile-con-ryzen-7-7840u
#780m #7840u #AMD #bluetooth #compatto #COMPUTER #em780 #HDMI21 #minipc #minisforum #displayport #nvme #PC #ryzen #usb4 #vesa #wifi #windows11 #linux
-
The #Ryzen #7940HS is quite impressive from both a performance and efficiency perspective. It provides high performance generic #compute with 8 cores and 16 threads.
The included integrated #GPU is the #AMD #Radeon #780M, part of the 3rd generation of #NAVI, using the #RDNA3 architecture. It supports #DirectX12, #Vulkan, hardware #RayTracing and variable rate shading.
This processor is fantastic for laptops, mobile devices, and #NUC like desktop usage