One user's local AI rig: DGX Spark bandwidth disappoints, RTX Pro 6000 looks like a steal
EAccelerate_42 · x · 2026-09-09
- The author recounts how the local AI hardware landscape changed over 1.5 years, listing his rig: RTX Pro 6000, RTX 5000, 2x DGX Spark, M4 Max, and M4 Pro.
- Key observations: DGX Spark's memory bandwidth is poor — its CUDA cores can't shine, and even an M4 Max has double the bandwidth; MoE + speculative decoding advances are making high-RAM non-RTX-Pro systems increasingly capable. He thinks the RTX Pro 6000 would be an insanely good deal at $6-8k.
More from Infra
- MagicAILabs claims new recipe matches DeepSeek V4 Pro pretraining with 50x less compute, ~$0.5M — AccBalanced · 2026-09-09
- LM Studio ships Bionic 1.1.2 with faster sessions, better screen reader support, Linux builds — mattturck · 2026-09-09
- Navier-Stokes proof burned 300B output tokens, $20-30M at consumer API prices — soumitrashukla9 · 2026-09-09
- Nvidia and AMD fight over credit guarantees to bankroll customers' data centers — rohanpaul_ai · 2026-09-09
- DeepSeek 4 Flash runs all day on Spark: zero crashes, 98% cache hit, 40 tok/s — jasonkneen · 2026-09-09
- New report: custom ASIC market shifts from design wins to program responsibility — BenBajarin · 2026-09-09