Hardware Choice: R9700 32GB vs W7800 48GB for AI Workloads
0xkbose · reddit · 2026-09-02
A user plans to upgrade a workstation for local inference and PyTorch training, debating between dual AMD R9700 32GB or dual W7800 48GB cards to pair with an existing RTX 3090 24GB. Workloads include GGUF inference (multimodal/coding models) and medical imaging (2D ultrasound/3D CT/MRI) plus clinical text training. The user seeks feedback on practical ROCm/PyTorch/MONAI training differences and known issues between R9700 (RDNA4) and W7800, planning to use the AMD cards as a main ROCm pair.
More from Infra
- Data Infrastructure is the Real Trend Amidst AI Hype — YvesMulkers · 2026-09-02
- Local Book Translation Pipeline: Gemma Translates, Qwen Edits on Dual P40s — neowisard · 2026-09-02
- Merge Gateway Adds Claude Fable 5.1 with 75% Cheaper Cache Reads — shensi · 2026-09-02
- Gemini Adds Agent-Based Video Analysis, Cuts Token Usage by 88% — The Decoder · 2026-09-02
- Editing Earlier Turns Invalidates Subsequent Thinking Blocks — eyishazyer · 2026-09-02
- High traffic from AI apps challenges traditional $5 VPS deployments — NielsRogge · 2026-09-02