Running ComfyUI on a $320 AMD V620 32GB Card: Benchmarks & Setup Guide
Brave_Load7620 · reddit · 2026-07-31
A developer spent only $360 ($320 for the card + $40 for cooling mods) on an AMD Radeon Pro V620 workstation card with 32GB of VRAM, successfully configuring it with ROCm, Triton, and Sage Attention to run ComfyUI on Windows.
Setup & Performance:
- Identified as a gfx1030 architecture GPU, it runs alongside an RX 9070 XT without conflicts, primarily handling background image and video generation tasks.
- The author shares detailed launch parameters and environment variables, noting that enabling Sage Attention provided roughly a 16% speedup in sampler step time compared to standard SDPA.
- Benchmarks: At 1080p, Z-Image Turbo takes 59s cold (47s warm); SDXL anime-style images take only 25s warm; Ace-step music generation takes about 3 minutes for a 3-minute song. While not the absolute fastest, the large VRAM offers excellent stability and value for local pipelines.
Related event: Running Local LLMs on a $360 AMD V620 GPU(2 posts)→
More from Infra
- DeepSeek-V4-Flash Ported to Run on AMD Strix Halo APU — Fit-Produce420 · 2026-07-31
- Satya Nadella Shares Hyperscaler ROIC Dashboard Showing 29.7% Average — firstadopter · 2026-07-31
- Running ComfyUI and Local LLMs on a $360 AMD V620 GPU: Success and Benchmarks — Brave_Load7620 · 2026-07-31
- Yuanli Semiconductor Raises Over 700M RMB in Series A for Edge AI Chips — 创业邦 · 2026-07-31
- Dwarkesh Predicts 10x Compute Cost Surge: Single H100 Could Yield $250K Yearly — rohanpaul_ai · 2026-07-31
- Inner Mongolia Desert Wind Farms Directly Powering AI Datacenters in China — 96Stats · 2026-07-31