B300 Inference Throughput Hits Multiple Times DeepSeek's
teortaxesTex · x · 2026-07-08
The B300 NVL72 paired with Dynamo achieves an 8K context per card at 80 tps, while DeepSeek using DSpark reaches about 1K context at the same 80 tps. If DeepSeek utilizes the H800 or 950 series, the performance-per-watt gap is at least 4x, though they remain roughly on par in terms of capital expenditure per dollar.
More from Infra
- Engram's random reads don't suit SSDs; CPU-memory over NVLink could serve all 72 GPUs — bookwormengr · 2026-09-11
- 80% of the DIY LLM inference hype posters have already quit — it's brutally hard systems work — abhijithneil · 2026-09-11
- Hugging Face's Ultra Scale Playbook: a free book on training LLMs on GPU clusters — mdancho84 · 2026-09-11
- Is inference latency becoming the biggest bottleneck for production AI agents? — Euphoric_Sea632 · 2026-09-11
- LLM Serving Metrics Thread: Why TPOT and Uptime Make or Break User Experience — abhijithneil · 2026-09-11
- PlanetScale launches sharded Postgres: 768 servers acting as one, 1PB scale — dhruv2038 · 2026-09-11