Head-to-head AI inference benchmark comparison of GB300 NVL72 (NVIDIA Blackwell) and TPU7x (Google Ironwood) on Qwen 3.5 397B-A17B. 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
Throughput at 201 tok/s/user on Qwen 3.5 397B-A17B: GB300 NVL72 hits 35411 tok/s/chip, TPU7x hits 34337. Per-million costs land at $0.02 and $0.01 respectively. TPU7x is 117% cheaper per token; GB300 NVL72 delivers 3% more tok/s/chip.
GB300 NVL72 / TPU7x on Qwen 3.5 397B-A17B at 320 tok/s/user: 22490 / 24540 tok/s/chip, $0.03 / $0.01 per million tokens. TPU7x is 145% cheaper per token; TPU7x delivers 9% more tok/s/chip.
Toward the upper edge of the 82β558 tok/s/user interactivity band, at 439 tok/s/user on Qwen 3.5 397B-A17B: GB300 NVL72 runs 13590 tok/s/chip at $0.05/M tokens, TPU7x runs 15745 at $0.02/M. TPU7x is 160% cheaper per token; TPU7x delivers 16% more tok/s/chip. (Numbers reflect the default agentic-traces Β· fp8 selection for this URL β table and chart below update if you change sequence, precision, or model in the controls.)
| Metric | Interactivity (tok/s/user) | Interactivity (tok/s/user) | Interactivity (tok/s/user) |
|---|---|---|---|
| Throughput (tok/s/chip) | GB300 NVL72:35411.0TPU7x:34337.4 | GB300 NVL72:22489.7TPU7x:24539.7 | GB300 NVL72:13589.6TPU7x:15745.0 |
| Cost ($/M tok) | GB300 NVL72:$0.018TPU7x:$0.008 | GB300 NVL72:$0.029TPU7x:$0.012 | GB300 NVL72:$0.047TPU7x:$0.018 |
| tok/s/MW | GB300 NVL72:16703317TPU7x:28448572 | GB300 NVL72:10608360TPU7x:20331166 | GB300 NVL72:6410185TPU7x:13044759 |
| Concurrency | GB300 NVL72:~171TPU7x:~22 | GB300 NVL72:~101TPU7x:~15 | GB300 NVL72:~51TPU7x:~10 |
Agentic inference metrics from the AgentX scenario and fixed-sequence inference metrics across models, hardware configurations, and serving parameters.
TCO $/chip/hr:
Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate cost per million tokens per decode chip or per prefill chip, rather than per total chip count. This makes direct token cost comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate input throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct input throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate output throughput per decode chip or per prefill chip, rather than per total chip count. This makes direct output throughput comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate power per decode chip or per prefill chip, rather than per total chip count. This makes direct power comparison with aggregated configs not an apples-to-apples comparison.
Note: Disaggregated inference configurations (e.g., MoRI SGLang, Dynamo TRTLLM) calculate Joules per decode chip or per prefill chip, rather than per total chip count. This makes direct Joules per token comparison with aggregated configs not an apples-to-apples comparison.
Matching measurements exist, but their chip series are hidden.
Shift+Scroll to zoom β’ Drag to pan β’ Double-click to reset β’ Click a point to pin tooltip
Qwen3.5 397B 8K / 1K Total Tokens per $1 TCO vs. Interactivity