Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
mast.hpc.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
Mastodon instance to support and encourage communication among the High Performance Computing community and related fields. (Personal use permitted.) More community features at https://hpc.social

Administered by:

Server stats:

118
active users

mast.hpc.social: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

Dr. Moritz Lehmann

Oh look, makes CPUs now! And I got my hands on one! 🖖😋
Today I benchmarked on Nvidia's , both and with . Finally I can answer the question: How does that exotic 2-chip APU show up in ?
--> It's 2 separate devices, a GPU with 94GB @ 4TB/s and a 72-core CPU with 480GB @ 384GB/s. The NVLink interconnect between the two is much faster than PCIe, achieving ~380GB/s host<->device bandwidth, only limited by poor misaligned VRAM BW on the GPU or RAM BW.

Nvidia GH200 in nvidia-smi and lscpu
How Nvidia GH200 shows up in OpenCL, and benchmarks of GPU and CPU chips
How Nvidia GH200 GPU and CPU stack up in FluidX3D benchmarks
Nvidia GH200 NARM Neoverse-V2 CPU loaded with FluidX3D/OpenCL shown in htop, with all 72 cores fully loaded
Aug 18, 2024, 04:28 PM·Public·Web
3boosts·16favorites
Public
Dr. Moritz Lehmann

@sascha how do I compile OpenCLCapsViewer for CLI on ARM Linux? The x86 CLI executable obviously doesn't work, yet I'd still like to upload the specs it to your database! 🖖🤔

Public
Dr. Moritz Lehmann

For reference: Here's a photo I took of a GH200 module at ISC24 in Hamburg. GPU on top, CPU on the bottom under a big heatspreader. Much fingerprint magnet!

Public
Sascha Willems

@ProjectPhysX Sorry, no idea. Never did anything for an ARM based Linux platform.

ExploreLive feeds
About