8 × NVIDIA RTX PRO 6000 Blackwell

Tested system

Processor: Intel(R) Xeon(R) Platinum 8592+
Graphics: NVIDIA RTX PRO 6000 Blackwell
Accelerator count: 8
Nodes: 1

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 / llama2-70b-99.9Offline · v6.1 · Tokens/s

8 × NVIDIA RTX PRO 6000 Blackwell

26,626.50^* Tokens/s

Accuracy target / result: ROUGE1: 44.7182 ROUGE2: 22.3154 ROUGEL: 29.104 TOKENS_PER_SAMPLE: 273.6
Weight precision: fp4
Division: closed
System type: datacenter
Availability: available
Cores per processor: 64
Processors per node: 2
Nodes: 1
Total accelerators: 8
Software stack: TensorRT 10.16, CUDA 13.2, cuDNN 9.21
Operating system: Ubuntu 24.04

MLPerf Inference / llama2-70b-99.9Server · v6.1 · Tokens/s

8 × NVIDIA RTX PRO 6000 Blackwell

27,340.90^* Tokens/s

Accuracy target / result: ROUGE1: 44.7475 ROUGE2: 22.3697 ROUGEL: 29.1476 TOKENS_PER_SAMPLE: 274.5
Weight precision: fp4
Division: closed
System type: datacenter
Availability: available
Cores per processor: 64
Processors per node: 2
Nodes: 1
Total accelerators: 8
Software stack: TensorRT 10.16, CUDA 13.2, cuDNN 9.21
Operating system: Ubuntu 24.04

Processing & platform

Processor: Intel(R) Xeon(R) Platinum 8592+
Graphics: NVIDIA RTX PRO 6000 Blackwell

Connectivity & software

Operating system: Ubuntu 24.04

System configuration

Reported host core count: 64
Processors per node: 2
Accelerator count: 8
Nodes: 1
Software stack: TensorRT 10.16, CUDA 13.2, cuDNN 9.21
Submission division: closed

* Provisional data