Back to home
RTX 5090
2 articles tagged with this topic
NVIDIAGemma
NVIDIA NVFP4 Puts 26B Model on Consumer GPU With Under 1% Accuracy Loss
NVIDIA's NVFP4 Gemma-4-26B shrinks to 18.8GB for consumer GPUs with <0.7% accuracy loss. 4-bit is now optimal, but also an ecosystem lock-in.
3h ago2 min read
QwenUnsloth
Qwen3.6-27B Quantized Fits Single Consumer GPU: Local Deployment Sweet Spot
Unsloth Q5-quantized Qwen3.6-27B runs stably on a single RTX 5090 across 19 rounds. Mid-size model local deployment is hitting the cost-capability swe
3h ago2 min read