gpt-oss 120B · Performance per Dollar
gpt-oss 120B — GB200 NVL72 vs MI355X Performance per Dollar
Cost per million tokens of GB200 NVL72 (NVIDIA Blackwell) versus MI355X (AMD CDNA 4) on gpt-oss 120B. Large-hyperscaler-volume ownership TCO normalized by total tokens — performance per dollar across LLM workloads. Pick the more cost-efficient SKU at every target interactivity level. Use the chart controls below to switch sequences, precisions, and metrics — same interactions as the main inference chart.
GB200 NVL72: $0.013 per million tokens. MI355X: $0.022. Both at 127 tok/s/user on gpt-oss 120B, with GB200 NVL72 39% cheaper.
Around the middle of the 68–305 tok/s/user interactivity band — at 187 tok/s/user — GB200 NVL72 runs $0.018 per million tokens on gpt-oss 120B while MI355X runs $0.036. GB200 NVL72 is the cheaper choice by 50%.
On gpt-oss 120B at 246 tok/s/user, the per-million math comes out to $0.027 for GB200 NVL72 and $0.12 for MI355X; GB200 NVL72 delivers 4.2× the total tokens per dollar. (Numbers reflect the default 8k/1k · fp4 selection for this URL — table and chart below update if you change sequence, precision, or model in the controls.)
Chip pricing (owning hyperscaler): GB200 NVL72 $1.86/chip/hr · MI355X $1.50/chip/hr. Source: SemiAnalysis Market July 2026 Pricing Surveys & AI Cloud TCO Model.
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) |
|---|---|---|---|
| Dollars per Million Tokens | GB200 NVL72$0.013MI355X$0.022 | GB200 NVL72$0.018MI355X$0.036 | GB200 NVL72$0.027MI355X$0.12 |
| Concurrency | GB200 NVL72~489MI355X~18 | GB200 NVL72~109MI355X~7 | GB200 NVL72~53MI355X~16 |
