Tag
A user built a home server from an old PC using modified mining GPUs, achieving around 30 tokens per second inference for the Qwen 3.8 27B AI model through driver patches and hardware upgrades.
A user with a 5090 GPU discusses how to spend $4,000 to enhance their AI model testing and coding capabilities, including considerations for media creation and additional local compute power.
GPU Solutions offers a service to upgrade NVIDIA GeForce RTX 2080 Ti VRAM from 11 GB to 22 GB by replacing GDDR6 modules, including BIOS configuration and stability testing.
A discussion about upgrading from dual RTX 3090s to alternatives like dual A6000s, RTX 5090, or 48GB RTX 4090, likely for AI/ML workloads.
The author shares that upgrading from single 24GB GPU to dual 24GB GPUs improved LLM performance not through larger models but through parallelism, using a main orchestrator agent with smaller subagents, increasing overall throughput for coding tasks.