3090 Upgrade Dilemma: Is 24GB VRAM Enough or Jump to 32GB?
gtech02 · reddit · 2026-08-05
Currently running an RTX 3090 (24GB VRAM) for AI tasks, the author is planning to split their previous dual-boot workstation into two dedicated machines after a hardware failure: a Linux box for AI and a Windows box for gaming.
Key Points
- Confirms the necessity of sticking with NVIDIA for AI workloads to avoid weeks of setup headaches.
- Debates the hardware upgrade path: whether to invest heavily in an upcoming Blackwell GPU (anticipating >24GB VRAM) for the AI box, or keep the 3090 for AI and just buy a mid-range card (e.g., RX 9070) purely for gaming.
More from Infra
- celld Released: Open-Source Self-Hosted Distributed Durable Objects — jedisct1 · 2026-08-05
- Unicity Launches Multi-Tenant Agent OS with 1000x Density — JoshuaJBouw · 2026-08-05
- Infineon Q3 Revenue Hits €4.17B as AI Datacenter DemandOffsets Auto Weakness — tengyanAI · 2026-08-05
- MiniMax Open-Sources Video Model: Community Runs It on 5GB VRAM in 48 Hours — DavidmComfort · 2026-08-05
- Benchmarking MiniMax H3 on RTX 5090: Efficiency vs. Pixel Workload — Forward-Parsley-148 · 2026-08-05
- Triton Team Built Gluon Language to Unlock Blackwell's Full Potential — ycombinator · 2026-08-05