Launching PCs for Generative AI Research
We have opened an online store for generative AI PCs.
https://shop.qualiteg.com/When starting research on AI or LLMs (large language models), one of the first challenges researchers and developers face is choosing the right hardware. For research in this field in particular, simply picking a high-performance computer is not enough; you need a specialized configuration tailored to the goals and requirements of your research. Yet deciding on the right PC configuration for a given research purpose is a significant challenge for many researchers and developers.
Building your own machine is an option, but GPUs and their peripheral components are expensive, and a wrong choice can end up costing a great deal. (Will the GPU even fit in the case? What about cooling? If you plan to install two or more GPUs, you also need to consider the CPU, bus width, and more. For those new to AI machines, the bar is high.)
Our lineup starts with affordably priced PCs whose specs let you comfortably work with LLMs in the 13-billion-parameter class. The models are equipped with either one or two A4000 GPUs.
The biggest difference between us and PC retailers or HPC (high-performance computing) vendors is the comprehensive, integrated hardware/software technical support we provide. We have deep knowledge and experience not only in GPU hardware but also in the software stack. For LLMs in particular, we bundle the Express Edition (free bundled version) of our flagship product, ChatStream(TM). This means you can start evaluating streaming chat with a variety of open-source LLMs right out of the box, without tedious installation work or complicated configuration.
We can also build custom configurations, such as Xeon-architecture generative AI PCs with four GPUs that can handle larger models like Meta's Llama. Please feel free to contact us.
