Gigabyte recently announced something pretty incredible: the TRX50 AI TOP (trillions of operations per second), a new motherboard that is specifically designed to train and fine tune AI locally.
The company says it designed the TRX50 to help utilise the maximum capacity of your CPU, GPU, memory, and SSD. With the right hardware, the company says it can support up to 236B-parameter large language models (LLMs) of AI training and fine-tuning, all on-device, which can be helpful for privacy and security.
The TRX50 is designed to work with AMD’s Threadripper and Threadripper Pro CPUs, so some of the most high-end parts on the market, unsurprisingly. The E-ATX board has four PCIe x16 slots, four NVMe M.2 slots, eight DIMM slots, two 10 Gbps Ethernet interfaces, WiFi 7 via a Qualcomm QCNCM865 module, and more. There’s also surge protection to make sure things don’t get too crazy.
Get on OpenAI’s level
Training up LLMs is notoriously tricky and resource-intensive, which is one of the reasons that OpenAI and its startup rivals are raising hundreds of millions of dollars right now. To this end, Gigabyte provides a helpful UI alongside new workflows and real-time progress monitoring.
The TRX50 supports many of the top open-source models out there, with a special focus on models aimed at beginners. The AI TOP Tutor can also offer advice on how to set up, tweak, and generally play with a model out of the box.
On top of these, Gigabyte also offers other DIY-friendly features, like EZ-Debug Zone to help debug LEDs and control buttons, the WIFI EZ-Plug design to consolidate Wi-Fi antenna plugs, and more.
Democratizing AI training
Like Bitcoin mining of the past few years, it’s hard not to see the AI frenzy as following a similar pattern of huge buzz and excitement. Of course, crypto is predominantly about the value of coins rising, so there are key differences, but the hardware needs are pretty similar.
Many of the Nvidia GPUs and various CPUs from different manufactures that found their way into bitcoin mining operations are now being repurposed towards AI, which has coincided with a general reduction in interest and investment in crypto as a sector.
While Gigabyte is a big name, the TRX50 is clearly a very niche product. On a boarder scale, Apple has been building its M-series laptops to handle a lot of AI-related workloads, including training, and the M4 chip comes with a variety of AI features.
