You can't run ChatGPT on a single GPU, but you can run some far less complex text generation large language models on your own PC. We tested oogabooga's text generation webui on several cards to see how fast it is and what sort of results you can expect.
https://www.tomshardware.com/news/running-your-own-chatbot-on-a-single-gpu
Login to add comment
Other posts in this group
AMD's strategy for RDNA 4 will prioritize the desktop market first.
Should you have the appropriate hardware configuration, a YouTuber finds a logical method for getting a Lossless Scaling Frame Gen without performance loss in any game.
You can update upscaling and frame-gen DLLs across DLSS, FSR, and XeSS through DLSS Swapper.
Intel significantly cut pricing for its newest Xeon 6 models. The flagship 128-core Xeon 6980P model is now available for $12,460, but it is still more expensive than AMD
Facebook is banning posts which mention various Linux related topics, sites, or groups.
Gorgeous aesthetics
Configurable in three layouts, Be Quiet’s Light Base 900 DX stands apart from the competition. We paired it with Intel’s i7-14700K and Radeon’s RX 7900 GRE to test its th
