Head-to-head AI inference benchmark comparison of MI355X (AMD CDNA 4) and TPU7x (Google Ironwood) on Kimi K3 2.8T. Latency, throughput, and cost across LLM workloads. Use the chart controls below to switch sequences, precisions, and metrics β 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 β
At 62 tok/s/user interactivity on Kimi K3 2.8T, MI355X delivers 8572 tok/s/chip at $0.05 per million tokens; TPU7x delivers 6210 tok/s/chip at $0.05. TPU7x is 6% cheaper per token; MI355X delivers 38% more tok/s/chip at this point.
MI355X posts 5517 tok/s/chip for $0.08 per million tokens at 99 tok/s/user on Kimi K3 2.8T; TPU7x posts 4701 tok/s/chip for $0.06. TPU7x is 24% cheaper per token; MI355X delivers 17% more tok/s/chip.
Throughput at 136 tok/s/user on Kimi K3 2.8T: MI355X hits 1718 tok/s/chip, TPU7x hits 3252. Per-million costs land at $0.24 and $0.09 respectively. TPU7x is 176% cheaper per token; TPU7x delivers 89% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp4 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | MI355X:8572.2TPU7x:6210.1 | MI355X:5517.1TPU7x:4701.4 | MI355X:1718.1TPU7x:3251.7 |
| Cost ($/M tok) | MI355X:$0.049TPU7x:$0.046 | MI355X:$0.076TPU7x:$0.061 | MI355X:$0.243TPU7x:$0.088 |
| tok/s/MW | MI355X:4101542TPU7x:5145092 | MI355X:2639760TPU7x:3895149 | MI355X:822078TPU7x:2694048 |
| Concurrency | MI355X:~22TPU7x:~43 | MI355X:~24TPU7x:~26 | MI355X:~11TPU7x:~16 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
No measurements to plot for this selection. Review the benchmark controls above or adjust quick filters.
