Kimi K3 2.8T · Chip comparison
Kimi K3 2.8T — B200 vs MI355X
Head-to-head AI inference benchmark comparison of B200 (NVIDIA Blackwell) and MI355X (AMD CDNA 4) on Kimi K3 2.8T. 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.
AgentX replays real coding-agent sessions rather than fixed-length prompts, so context grows turn over turn and most of each request is served from cache instead of being recomputed. That turns the comparison into a systems question: KV transfer between nodes, prefix-aware routing, and cache capacity all move the curve alongside raw chip throughput. Fixed-sequence workloads stay the clean baseline for kernel and silicon performance, so the two scenarios answer different questions about the same hardware. Learn more about AgentX →
Throughput at 57 tok/s/user on Kimi K3 2.8T: B200 hits 5,372 tok/s/chip, MI355X hits 9,014. Costs per million tokens land at $0.089 and $0.046 respectively. MI355X is 48% cheaper per token; MI355X delivers 68% more tok/s/chip.
B200 / MI355X on Kimi K3 2.8T at 100 tok/s/user: 2,677 / 5,436 tok/s/chip, $0.18 / $0.077 per million tokens. MI355X is 57% cheaper per token; MI355X delivers 103% more tok/s/chip.
Toward the upper edge of the 15–186 tok/s/user interactivity band, at 144 tok/s/user on Kimi K3 2.8T: B200 runs 1,751 tok/s/chip at $0.27 per million tokens, MI355X runs 2,854 at $0.15. MI355X is 47% cheaper per token; MI355X delivers 63% more tok/s/chip. (Numbers reflect the default agentic-traces · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Interpolated from real benchmark data. Edit target interactivity values below to compare at different operating points.
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | B2005,372MI355X9,014 | B2002,677MI355X5,436 | B2001,751MI355X2,854 |
| Cost ($/M tok) | B200$0.089MI355X$0.046 | B200$0.18MI355X$0.077 | B200$0.27MI355X$0.15 |
| tok/s/MW | B2003,141,551MI355X4,313,126 | B2001,565,781MI355X2,600,857 | B2001,023,717MI355X1,365,674 |
| Concurrency | B200~18MI355X~16 | B200~7MI355X~24 | B200~5MI355X~4 |