Wedbush Has a Message for AMD Investors After Google Move
I'm LongbridgeAI, I can summarize articles.Wedbush analyst Matt Bryson suggests AMD could play a deeper role in Google's AI infrastructure, potentially designing its next-generation Tensor Processing Units (TPUs). This collaboration would expand AMD's opportunity beyond CPUs and GPUs into custom AI silicon. While unconfirmed, such a partnership would highlight AMD's growing expertise in ASIC design and packaging, diversifying revenue sources beyond Instinct GPUs as hyperscalers demand integrated compute platforms.
Advanced Micro Devices could be positioning itself for a much deeper role inside Google's AI infrastructure after reports suggested Alphabet (GOOGL) is working with AMD on a future generation of its tensor processing units. Wedbush says such a collaboration could represent a meaningful shift in the semiconductor landscape by expanding AMD's opportunity beyond CPUs and GPUs into custom AI silicon design.
AMD is best known for its EPYC server processors and Instinct AI accelerators, but it has been broadening its data-center portfolio into networking, advanced packaging and rack-scale systems. That matters because hyperscalers increasingly need tightly integrated compute platforms rather than standalone chips.
SemiAnalysis reported that Google may use AMD to help design its 10th-generation TPU, drawing on AMD's CPU expertise as well as its packaging and networking capabilities.
Wedbush analyst Matt Bryson said, We would find it noteworthy if Google collaborated with AMD vs. Broadcom, Marvell, or Intel, adding that the speculation highlights the growing importance of ASIC design and reusable IP blocks.
Google unveiled its eighth-generation TPUs in April, with separate chips targeting AI training and inference. Alphabet has also been expanding the TPU business beyond internal use, saying it plans to deliver TPU hardware directly to select customers' data centers beginning this year.
AMD, meanwhile, is already seeing explosive AI infrastructure growth. Its data-center revenue surged 107% year over year to $6.7 billion in the second quarter, driven by EPYC CPUs and Instinct GPUs.
Investor Takeaway
For AMD investors, the importance of a Google TPU partnership would extend well beyond incremental chip revenue.
A confirmed design role would show AMD can monetize CPU IP, advanced packaging, networking and ASIC expertise alongside its existing accelerator business. That could open another route into hyperscaler AI spending without requiring every dollar to come from Instinct GPUs.
Investors should watch for confirmation from Google or AMD, future TPU design disclosures and any evidence that AMD is winning similar custom-silicon work elsewhere.
The main risk is that the collaboration remains limited or speculative. But if AMD becomes a recurring design partner for hyperscalers, its addressable AI market could broaden materially beyond the Nvidia-style GPU competition investors typically focus on.