Welcome Here  Shenzhen Mingjiada Electronics Co., Ltd.

Recently, it was reported that U.S. chip giant AMD announced the acquisition of Toronto-based AI chip company Untether AI and confirmed the news to CRN. However, this was not a typical acquisition. To be precise, AMD only took on Untether AI's entire engineering team and did not acquire anything else. Therefore, Untether AI stated in its announcement that after the transaction was completed, its speedAI AI products and imAIgine SDK would no longer be supplied or supported.

Untether AI stated that while this transaction marks the end of its journey, it takes pride in the pioneering research that laid the foundation for its work in advancing cutting-edge AI chip technology. Untether AI expressed gratitude for the dedication of its engineering team, as well as the support of its customers, partners, and investors, and looks forward to this world-class team contributing to AMD.

AMD stated in a press release that it has entered into a strategic agreement with Untether AI to acquire a talented team of AI hardware and software engineers. Going forward, the team will focus on enhancing AI compiler and kernel development capabilities, as well as strengthening AMD's digital and SoC design, design verification, and product integration capabilities. AMD is delighted to welcome the team and their unique expertise into the company.

Untether AI specialises in designing chips specifically for AI inference. While GPUs like NVIDIA Blackwell Ultra and AMD Instinct MI350 excel in training AI models, Untether AI performs better in inference and is more energy-efficient. Typically, Untether AI places AI chips close to storage chips to reduce latency and power consumption.

As AI chips continue to evolve, power consumption is increasing, prompting the search for more efficient alternatives. Recently, AMD has been actively acquiring companies in the AI sector, focusing not only on the raw computational power of AI GPUs but also on how to expand capabilities and improve efficiency.