4x3090 rig vs DGX Spark: the local LLM hosting trade-offs worth knowing
Dapper_Chance_2484 · reddit · 2026-10-17
A Reddit user weighs upgrading a 2x3090 rig to 4x3090 (possibly with a Threadripper) versus buying a DGX Spark-class device for hosting larger local models. Key trade-offs listed: distributed VRAM vs unified memory, high combined bandwidth vs low memory bandwidth, 1200-1400W vs <200W power draw, and Ampere vs Blackwell architecture. A representative dilemma for GPU-poor local deployment users, though no benchmarks or conclusions yet.
More from Infra
- Modal explains ImageFS, the lazy-loaded filesystem behind sub-second cold starts — josh_wills · 2026-10-17
- Blackwell Ultra GPU rental prices climb 17 straight days to an all-time high — McDonaghMatthew · 2026-10-17
- Stop posting tok/s screenshots: 5 metrics a credible local AI benchmark needs — EAccelerate_42 · 2026-10-17
- 3 models, 4 engines, 4 Macs: Draw Things hits 1.8x mflux speed on Apple Silicon — liuliu · 2026-10-17
- SurrealDB's new engines hit 28.3x RocksDB point reads, Agent Memory goes live — mattturck · 2026-10-17
- Zuckerberg & Wang on Muse Wonder: Meta's $100B AI Bet Past 5GW — RylanSchaeffer · 2026-10-17