HPE Cray XD670 (8x NVIDIA H200-SXM-141GB): edge AI benchmarks
Key results
| Workload | Best result | Model · precision | Power · perf/W | Grade · date | Source |
|---|---|---|---|---|---|
| image/video generation | 0.0241 FPS batch unknown | MLPerf wan-2.2-t2v-a14b bfloat16 · KISS-V: Krai Inference Serving Solution for Video (Fast Mode) | unknown | B 2026-04-01 (published) | MLPerf Inference v6.0, open division, data-centre suite |
| LLM serving (MLPerf datacenter) | 25940 tok/s batch unknown | MLPerf llama2-70b-99.9 fp16 · vLLM v0.15.0-dev | unknown | B 2026-04-01 (published) | MLPerf Inference v6.0, open division, data-centre suite |
Best published result per workload, across model sizes. For like-for-like groups, open the per-workload tables below.
Show data — per-workload tables
object detection
No published object detection results for HPE Cray XD670 (8x NVIDIA H200-SXM-141GB) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
instance segmentation
No published instance segmentation results for HPE Cray XD670 (8x NVIDIA H200-SXM-141GB) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
pose estimation
No published pose estimation results for HPE Cray XD670 (8x NVIDIA H200-SXM-141GB) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
image classification
No published image classification results for HPE Cray XD670 (8x NVIDIA H200-SXM-141GB) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
semantic segmentation
No published semantic segmentation results for HPE Cray XD670 (8x NVIDIA H200-SXM-141GB) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
speech-to-text
No published speech-to-text results for HPE Cray XD670 (8x NVIDIA H200-SXM-141GB) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
LLM generation
No published LLM generation results for HPE Cray XD670 (8x NVIDIA H200-SXM-141GB) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.
image/video generation
| Like-for-like group | Best result | Model · precision · stack | Grade | Power | Source · date | Rank in group |
|---|---|---|---|---|---|---|
| unknown · b=unknown | 0.0241 FPS | MLPerf wan-2.2-t2v-a14b bfloat16 · KISS-V: Krai Inference Serving Solution for Video (Fast Mode) | B | unknown power n/a | MLPerf Inference v6.0, open division, data-centre suite 2026-04-01 (published) | 16 of 16 |
LLM serving (MLPerf datacenter)
| Like-for-like group | Best result | Model · precision · stack | Grade | Power | Source · date | Rank in group |
|---|---|---|---|---|---|---|
| LLM >40B · b=unknown | 25940 tok/s | MLPerf llama2-70b-99.9 fp16 · vLLM v0.15.0-dev | B | unknown power n/a | MLPerf Inference v6.0, open division, data-centre suite 2026-04-01 (published) | 80 of 93 |
How to read these numbers
Machine-readable: /find.json (see llms.txt); full tables are in “Show data”. Confidence grades: A audited MLPerf closed · B documented third-party · C community · D vendor-published · E estimate. Unknown values are shown as unknown, never estimated. Generated 2026-10-16 from public sources.