Llama 3.3 70B · Chip comparison
Llama 3.3 70B — B200 vs MI300X
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and MI300X (AMD CDNA 3) on Llama 3.3 70B. 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.
Near the low end of the 14–101 tok/s/user interactivity band, at 36 tok/s/user on Llama 3.3 70B: B200 runs 6,670 tok/s/chip at $0.072 per million tokens, MI300X runs 1,823 at $0.14. B200 is 50% cheaper per token; B200 delivers 3.7× the tok/s/chip.
Setting 58 tok/s/user as the target on Llama 3.3 70B, B200 produces 5,052 tok/s/chip ($0.095 per million tokens) and MI300X produces 1,144 ($0.23). B200 is 59% cheaper per token; B200 delivers 4.4× the tok/s/chip.
At 80 tok/s/user interactivity on Llama 3.3 70B, B200 delivers 3,781 tok/s/chip at $0.13 per million tokens; MI300X delivers 679 tok/s/chip at $0.39. B200 is 67% cheaper per token; B200 delivers 5.6× the tok/s/chip at this point. (Numbers reflect the default 8k/1k · fp8 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B2006,670MI300X1,823 | B2005,052MI300X1,144 | B2003,781MI300X679 |
| Cost ($/M tok) | B200$0.072MI300X$0.14 | B200$0.095MI300X$0.23 | B200$0.13MI300X$0.39 |
| tok/s/MW | B2003,900,512MI300X1,311,463 | B2002,954,131MI300X823,315 | B2002,211,020MI300X488,526 |
| Concurrency | B200~47MI300X~27 | B200~32MI300X~16 | B200~25MI300X~8 |