- Mediatek challenges Broadcom in the evolutionary hardware strategy of Google AI
- The new generation TPU aims for better treatment of AI and a reduction in costs
- Google diversifies the supply of AI fleas, reducing Broadcom dependence
Mediatek and Google collaborate to develop new generation tensors’ treatment units (TPUS), a decision which means that Google will likely reduce its dependence on Broadcom in its AI hardware commercial strategy.
The objective of this collaboration is to design an affordable and energy efficient TPU, produced by TSMC and will probably be launched as part of the 7th generation TPU series from Google.
Although Google will lead the global design, Mediatek will manage the entry / output modules to facilitate communication between the processor and the peripherals, marking a passage from the previous Broadcom role in the basic development of TPU, although Google does not cut links with Broadcom completely.
Develop more efficient AI processors
Google TPUs, designed to require AI and essential calculations for the training of the best LLM, will see the improvement in efficiency and performance, meeting the growing demand for powerful AI.
Google has developed its own AI server chips for internal use and cloud services in order to reduce dependence on external suppliers like NVIDIA, whose domination of the IA equipment contrasts with Google thrust for a more self-suisse ecosystem via TPU investments.
Meanwhile, Mediatek has expanded its presence in the development of AI fleas, including partnerships in AI superordinaters.
In Arly 2025, the company has announced a collaboration on the IA supercomputer project, with its latest partnership with Google aligning with its wider strategy to strengthen its position in the AI industry.
If this collaboration is successful, it can influence the development of the best LLM models by making the treatment of AI more effective.
Via UDN
