DeepSeek V4 Pro 1.6T · Chip 对比

DeepSeek V4 Pro 1.6T — B300 vs GB200 NVL72

B300(NVIDIA Blackwell)与 GB200 NVL72(NVIDIA Blackwell)在 DeepSeek V4 Pro 1.6T 上的正面 AI 推理基准测试对比。涵盖各类 LLM 工作负载的延迟、吞吐量与成本。使用下方图表控件切换序列、精度和指标——交互方式与主推理图表相同。

在 18–203 tok/s/user 交互性区间的低端,即 DeepSeek V4 Pro 1.6T 上以 64 tok/s/user 运行时:B300 达到 3048 tok/s/Chip($0.21/百万 token),GB200 NVL72 达到 9108($0.06/百万)。GB200 NVL72 每 token 成本低 278%;GB200 NVL72 每 Chip 吞吐量高出 199%。

以 111 tok/s/user 为目标在 DeepSeek V4 Pro 1.6T 上运行时,B300 产出 1513 tok/s/Chip(每百万 token $0.41),GB200 NVL72 产出 6055($0.08)。GB200 NVL72 每 token 成本低 389%;GB200 NVL72 每 Chip 吞吐量高出 300%。

在 DeepSeek V4 Pro 1.6T 上以 157 tok/s/user 交互性运行时,B300 吞吐量为 870 tok/s/Chip,每百万 token 成本 $0.72;GB200 NVL72 吞吐量为 523 tok/s/Chip,成本 $0.99。B300 每 token 成本低 37%;B300 每 Chip 吞吐量高出 67%。 (数据反映此 URL 的默认 8k/1k · fp4 选择——如果您在控件中更改序列、精度或模型,下方表格和图表会自动更新。)

查看每美元性能对比 →

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)
B300:3047.8GB200 NVL72:9108.2
B300:1513.0GB200 NVL72:6055.1
B300:870.4GB200 NVL72:522.7
Cost ($/M tok)
B300:$0.214GB200 NVL72:$0.057
B300:$0.415GB200 NVL72:$0.085
B300:$0.721GB200 NVL72:$0.990
tok/s/MW
B300:1604089GB200 NVL72:4870672
B300:796291GB200 NVL72:3238016
B300:458110GB200 NVL72:279500
Concurrency
B300:~60GB200 NVL72:~9943
B300:~8GB200 NVL72:~1815
B300:~3GB200 NVL72:~41

推理性能

不同模型、硬件配置和服务参数下的推理性能指标。

厂商:
部署模式:
投机解码: