Latest in AI

club-3090 Adds Experimental FP8 Support for Qwen3.6-27B

The open-source project club-3090 has rolled out experimental FP8 quantization support for Qwen3.6-27B. This update is highly anticipated by dual RTX 3090 users, allowing them to run the model with significantly reduced VRAM requirements. According to reports, the official Qwen3.6-27B-FP8 model performs virtually identically to the original unquantized BF16 version.