We program software for GPUs and clusters since 2010. #GPGPU #OpenCL #CUDA #HSA #ROCM #OpenMP #MPI #HPC #LLVM

Europe
Joined January 2010
Include

Only show posts containing:

Exclude

Hide posts containing:

Time range
-
Minimum likes
Replying to @Mattermost
We downgraded to 11.7 because 11.8 was seriously misbehaving - slow, eating more memory, broken functionality, etc. We have the paid version. It looked like untested code hit production. And based on our long experience with MM, this is probably exactly what happened (again).
56
Replying to @elonmusk
@elonmusk Hey, you need more OpenCL magic from the world's best GPGPU-experts to handle more workloads on AI5 and AI6. What mailbox is not full?
70
Replying to @openprovider
@openprovider Jullie DNS heeft een probleem. Eigen subdomeinen doen het weer, maar ik kan jullie webpagina niet bezoeken.
1
33
The more news you consume, the less well-informed you are.
54
62
11
567
112,948
Replying to @hpcnotes
Nope, and you even get invited to the cooler parties! ;)
1
53
Replying to @PostNL
@PostNL Vandaag verwachtten we 4 grote pakketten. Toevallig konden we zien vanuit kantoor wat de postbode deed. Bij het kantoorgebouw naast ons werd bezorgd, daarna kregen we een melding dat we niet thuis waren, en reed hij weg. Dit is niet een eenmalig incident
1
36
Replying to @gitlab
The project management is only at a usable state when getting the ultimate plus special PM-addon, costing a total of $115 (!) per seat. The add-on "GitLab Enterprise Agile Planning" should be made available for Premium. Somebody did not think this through.
58
Dawn, we probably have some serious social media problems. #MyXAnniversary
221
Trust no one
50
113
21
1,188
100,095
We're at @hipeac - flight canceled, hotel overbooked, posters gone. But we're there!
1
1
1
513
In HPC, independent networking brands are soooo 2023! Context: investor.juniper.net/investo…
200
Replying to @Donzanoid
By that measurement Microsoft "technologies" are more interesting than shaders. Wait...
227
Stream HPC retweeted
Amazon details its 4th-generation Graviton server processor. #Arm #ArmNeoverse #Armv9 fuse.wikichip.org/news/7633/…
1
3
30
3,833
Stream HPC retweeted
Next week, we're launching the EA version of our product towards #DecentralizedAI. This local super AI app integrates SDXL, Llama, SAM, and a bag of GANs. Even 8GB low end game GPU can experience datacenter-level GPU + TensorRT performance. Support Mac/Win/Linux Stay tuned.
1
1
11
1,227
😂
2
29
356
24,929