Running DeepSeek v4 PRO on Mac M5 Max via SSD Streaming and DGX Station Optimization
antirez · x · 2026-08-16
The author explores using DGX Station to pool RAM and VRAM for large model inference, noting that while bandwidth is a bottleneck, good orchestration via DwarfStar allows Q2/Q3 quants to run well. A demo shows DeepSeek v4 PRO Q2 running on a 128GB Mac M5 Max via SSD streaming, successfully translating an Italian short story to English, proving the model's usability for such tasks.
Related event: DeepSeek v4 PRO Tested on Mac M5 Max via SSD Streaming(2 posts)→
More from Infra
- Three CoreWeave co-founders have sold $3.3B in shares since IPO amid credit worries — SumitGup · 2026-10-16
- Cherry pi on a Raspberry Pi: the tiny-model rumor developers are watching — sull · 2026-10-16
- CoreWeave's debt costs falling: DDTL spreads tightened from S+9.62% to S+2.25% — BenBajarin · 2026-10-16
- SoftBank lines up US IPOs for SB Energy and AI robotics spinoff Roze, valuations top $50B — nordicinst · 2026-10-16
- TRL reproduces Modal's mixed-precision async RL: plain GRPO collapses, SC+MIS holds 0.72 — QGallouedec · 2026-10-16
- Hugging Face scales Indic pretraining with 2.8T tokens and an open data pipeline — Hugging Face Blog · 2026-10-16