Intel and Google deepen AI CPU push: why the next AI trade may spread beyond GPUs

Intel and Google are expanding their partnership around AI-focused CPUs and custom infrastructure processors. The key takeaway is that AI deployment is spreading beyond training into inference, enterprise workloads and broader data-centre computing. That could help parts of the server CPU, cloud infrastructure and networking stack, while putting some pressure on names that have been treated as the only AI winners.

Winners

AI server CPU and inference names

If AI demand is moving further into inference, enterprise deployment and general-purpose compute, CPUs become more important alongside accelerators. Intel is the direct headline winner because Google is continuing to deploy Xeon chips and use Intel’s newer processors. AMD can benefit too because the market may start giving more credit to the broader server CPU space rather than focusing only on GPUs.

Names: $INTC (Intel), $AMD (Advanced Micro Devices)

Cloud and data-centre infrastructure beneficiaries

Balanced AI systems need more than accelerators. They also need servers, infrastructure design, power efficiency and enterprise deployment support. Alphabet benefits because this partnership may improve the economics and performance of Google’s AI infrastructure. Dell and HPE can benefit if customers increasingly invest in full AI server stacks built for inference, hybrid workloads and data-centre refresh cycles.

Names: $GOOGL (Alphabet), $DELL (Dell Technologies), $HPE (Hewlett Packard Enterprise)

Networking and connectivity names tied to AI buildout

As AI infrastructure expands beyond training clusters, networking remains critical. More inference traffic, enterprise AI deployment and larger-scale distributed compute can drive demand for switching, routing and data-centre connectivity. That is supportive for companies exposed to AI-related network upgrades.

Names: $ANET (Arista Networks), $CSCO (Cisco Systems)

Losers

GPU pure-play expectations trade

This is not necessarily bearish for AI overall, but it could shift investor attention away from a GPU-only narrative. If the market starts pricing in a more balanced AI architecture where CPUs and infrastructure processors matter more, some of the premium attached to the most crowded GPU-linked trades could come under pressure, especially after strong runs.

Names: $NVDA (NVIDIA), $SMCI (Super Micro Computer)

Rival x86 CPU share-gain expectations

Intel’s deal with Google strengthens the idea that Intel can still defend and possibly rebuild share in AI-related server compute. That can challenge the view that Intel keeps losing ground everywhere in data-centre infrastructure. AMD remains a broader AI infrastructure beneficiary, but in relative terms this specific headline improves Intel’s competitive positioning.

Names: $AMD (Advanced Micro Devices), $NVDA (NVIDIA)

Custom AI chip and accelerator competitors

Intel and Google are expanding co-development around custom infrastructure processing, which suggests Google is investing more deeply in alternative ways to optimise workloads. If large cloud customers increasingly build more of their own infrastructure layer, some expectations around third-party custom silicon suppliers may face tougher comparisons.

Names: $MRVL (Marvell Technology), $AVGO (Broadcom)

#StockMarket #Trading #Investing #DayTrading #SwingTrading #AI #ArtificialIntelligence #Semiconductors #Intel #Google #CloudComputing #DataCenter #CPUs #Inference #TechStocks #USStocks