GeekRadar / The comparison desk
COMPARE.
Tested systems
Cisco

NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e · 16 total

Tested system

Processor
AMD EPYC
Graphics
NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e
Accelerator count
16
Nodes
2

Recorded results

Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.

MLPerf Inference / gpt-oss-120bOffline · v6.1 · Tokens/s

NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e · 16 total

94,249.30* Tokens/s

Accuracy target / result
exact_match: 83.57105741751104
Weight precision
mxfp4
Division
closed
System type
datacenter
Availability
available
Cores per processor
128
Processors per node
2
Nodes
2
Total accelerators
16
Software stack
vLLM
Operating system
Ubuntu 22.04
MLPerf Inference / gpt-oss-120bServer · v6.1 · Tokens/s

NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e · 16 total

73,357.60* Tokens/s

Accuracy target / result
exact_match: 83.43465835281891
Weight precision
mxfp4
Division
closed
System type
datacenter
Availability
available
Cores per processor
128
Processors per node
2
Nodes
2
Total accelerators
16
Software stack
vLLM
Operating system
Ubuntu 22.04
MLPerf Inference / llama2-70b-99.9Offline · v6.1 · Tokens/s

NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e · 16 total

99,683.80* Tokens/s

Accuracy target / result
ROUGE1: 44.4169 ROUGE2: 22.068 ROUGEL: 28.7957 TOKENS_PER_SAMPLE: 286.6
Weight precision
fp8,fp4
Division
closed
System type
datacenter
Availability
available
Cores per processor
128
Processors per node
2
Nodes
2
Total accelerators
16
Software stack
vLLM
Operating system
Ubuntu 22.04
MLPerf Inference / llama2-70b-99.9Server · v6.1 · Tokens/s

NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e · 16 total

95,104.60* Tokens/s

Accuracy target / result
ROUGE1: 44.6044 ROUGE2: 22.1926 ROUGEL: 28.9466 TOKENS_PER_SAMPLE: 284.7
Weight precision
fp8,fp4
Division
closed
System type
datacenter
Availability
available
Cores per processor
128
Processors per node
2
Nodes
2
Total accelerators
16
Software stack
vLLM
Operating system
Ubuntu 22.04
MLPerf Inference / llama3.1-8bOffline · v6.1 · Tokens/s

NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e · 16 total

163,292* Tokens/s

Accuracy target / result
ROUGE1: 38.5675 ROUGE2: 16.0757 ROUGEL: 24.5063 ROUGELSUM: 35.5154 GEN_LEN: 8141814
Weight precision
fp8,fp4
Division
closed
System type
datacenter
Availability
available
Cores per processor
128
Processors per node
2
Nodes
2
Total accelerators
16
Software stack
vLLM
Operating system
Ubuntu 22.04
MLPerf Inference / llama3.1-8bServer · v6.1 · Tokens/s

NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e · 16 total

161,110* Tokens/s

Accuracy target / result
ROUGE1: 38.6437 ROUGE2: 16.1045 ROUGEL: 24.5278 ROUGELSUM: 35.5601 GEN_LEN: 8147699
Weight precision
fp8,fp4
Division
closed
System type
datacenter
Availability
available
Cores per processor
128
Processors per node
2
Nodes
2
Total accelerators
16
Software stack
vLLM
Operating system
Ubuntu 22.04

Processing & platform

Processor
AMD EPYC
Graphics
NVIDIA H200-SXM-141GB + AMD Instinct MI350X 288GB HBM3e

Connectivity & software

Operating system
Ubuntu 22.04

System configuration

Reported host core count
128
Processors per node
2
Accelerator count
16
Nodes
2
Software stack
vLLM
Submission division
closed

* Provisional data