TSMC Partners Up With NVIDIA & Broadcom to Develop Cutting-Edge Silicon Photonics

11.09.2023 - 13:27 / wccftech.com

Korean media reports that TSMC has teamed up with Broadcom & NVIDIA to develop cutting-edge silicon photonics, in order to provide massive transmission speeds for AI computing.

For those unaware, silicon photonics is a next-gen adaption of the traditional copper transmission cables, which combines laser and silicon technology to provide a solution that guarantees high data transfer speeds. Silicon photonics have gained immense traction in the industry, particularly with the advent of the AI boom, since it plays a crucial role in data centers and HPC machines.

Taiwan Economic Daily reports that TSMC, Broadcom, and NVIDIA are forming a partnership with the aim of innovating the "silicon photonic" industry. TSMC has been rumored to allocate 200 professionals for R&D, primarily focusing on the tech in silicon photonic integration into high-speed computing chips, which could potentially lead to a major breakthrough for the AI industry.

TSMC is apparently highly optimistic towards next-gen silicon photonics development since the company's VP Yu Zhenhua has stated that the development could solve several problems currently existing in the industry:

If we can provide a good silicon photonics integrated system, we can solve the two key issues of energy efficiency and AI computing power. This will be a new one.

Paradigm shift. We may be at the beginning of a new era.

The semiconductor industry sees silicon photonics as the future, mainly due to the benefits it brings onboard. The conventional "electricity" data transmissions come with compromised transfer speeds, & the industry is demanding 'higher computational speeds" given that genAI development has reached unprecedented levels. To cater to this, silicon photons convert electricity to light, which guarantees faster speeds with a cost-efficient method. Intel also recently detailed its latest Silicon Photonics silicon at Hot Chips 2023 which is based on RISC and features an 8 core / 528 thread config.

NVIDIA and partners are working on expanding the technology from the 45-nanometer process up to the 7-nanometer, which will ultimately bring in significant performance uplift onboard. The process is expected to come into implementation by 2024, with mass production expected by 2025. Hence we could see next-gen AI GPUs to boast significant transmission speeds in the upcoming years.

Similar to the "dot.com" bubble, the rapid boom of the AI industry made it evident that innovation will occur, and the integration of silicon photonics is a prime example. With the passage of time, one should expect huge developments, especially in the department of HPC, since that is the way to lead in the future of genAI.

News Source: Taiwan Economic Daily