DeepSeek R1 · Chip comparison
DeepSeek R1 — B300 vs MI355X
Head-to-head AI inference benchmark comparison of B300 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) on DeepSeek R1. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
At 60 tok/s/user interactivity on DeepSeek R1, B300 delivers 5,239 tok/s/chip at $0.12 per million tokens; MI355X delivers 1,942 tok/s/chip at $0.21. B300 is 44% cheaper per token; B300 delivers 170% more tok/s/chip at this point.
B300 posts 1,537 tok/s/chip for $0.41 per million tokens at 108 tok/s/user on DeepSeek R1; MI355X posts 1,028 tok/s/chip for $0.41. Cost per token is essentially tied; B300 delivers 50% more tok/s/chip.
Throughput at 157 tok/s/user on DeepSeek R1: B300 hits 899 tok/s/chip, MI355X hits 406. Costs per million tokens land at $0.70 and $1.03 respectively. B300 is 32% cheaper per token; B300 delivers 121% more tok/s/chip. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B3005,239MI355X1,942 | B3001,537MI355X1,028 | B300899MI355X406 |
| Cost ($/M tok) | B300$0.12MI355X$0.21 | B300$0.41MI355X$0.41 | B300$0.70MI355X$1.03 |
| tok/s/MW | B3002,757,405MI355X929,125 | B300809,048MI355X491,854 | B300473,375MI355X194,384 |
| Concurrency | B300~195MI355X~36 | B300~41MI355X~9 | B300~5MI355X~4 |