From S3 to GPU in One Copy | Hacker News - But Better

From S3 to GPU in One Copy spiraldb.com

We optimized Vortex so it can stream data from object storage into GPU memory at 90 Gbps. The first step to get this performance is to make sure each byte is only copied once between the kernel and the GPU.

2 points•surprisetalk•6 days ago•0 comments•

#gpu #copy #one copy #onur #qcon #saticis #london #transcript

0 comments

No comments yet.

Related stories

Show HN: A Simple GPU Job Schedulergithub.com
Hacker News · 2 points · 4 days ago
Show HN: Napkin – a scratchpad that holds texts and images for copygithub.com
Hacker News · 3 points · 8 days ago
A 4 GB Laptop GPU Beats a 12-Core CPU by 4.3x on Gemma 4dev.to
DEV Community · 15 points · 14 days ago
Why GPU Availability Is Still the Biggest Bottleneck in ML Infradev.to
DEV Community · 1 points · 9 days ago
Sparse resources helped my GPU-driven renderer memory usagezino2201.substack.com
Hacker News · 1 points · 10 days ago
Build in the VM, Think on the Mac GPU: Debian 13 on Apple container With a Local Gemma 4dev.to
DEV Community · 9 points · 17 days ago