Inference Chip Index

← All slices

deepseek-r1 · Interactive

deepseek-r1 / Interactive / Tokens/s

v6.0|closed|deepseek-r1|Interactive|Tokens/s|official-system

Metric: Performance_Result (Tokens/s) · winning: higher · 3 ranking-eligible rows · vendors 1 · families 2 · never mixed with other slices

All systems (official)
RankPerformanceUnitSystemSubmitterAcceleratorSource
1250,634.0000Tokens/sNVIDIA GB300 NVL72 (72x GB300-288GB_aarch64, TensorRT)NVIDIANVIDIA GB300source
2240,318.0000Tokens/sNVIDIA GB200 NVL72 (72x GB200-186GB_aarch64, TensorRT)NVIDIANVIDIA GB200source
34,934.6500Tokens/sG894-SD3-AAX7GigaComputingNVIDIA B300-SXM-270GBsource