Morgan Stanley predicts agentic AI will broaden chip spending beyond GPUs to CPUs and memory, impacting data center infrastructure.
Why it matters
This report highlights a significant shift in the hardware landscape driven by advancements in AI. As AI systems become more autonomous, the computational demands will diversify, moving beyond specialized AI accelerators like GPUs to more general-purpose processors (CPUs) and memory. This will have profound implications for semiconductor manufacturers, data center operators, and the overall tech industry, potentially leading to new market opportunities and strategic adjustments in hardware development and investment.
As AI gets smarter and can do more on its own, it will need more than just special graphics chips. It will also need more powerful regular computer chips (CPUs) and memory. This means companies that make these chips will also benefit from the AI boom.
Anthropic's new AI model, Mythos, designed for cybersecurity, can identify thousands of software vulnerabilities, raising concerns about its potential to accelerate cyberattacks.
Read on Economic Times Tech →Marvell is reportedly in talks with Google to develop two AI chips, including a memory processing unit and a new TPU for AI model execution.
Read on Economic Times Tech →Tech layoffs are projected to exceed 73,000 in 2026, with major companies like Meta and Oracle citing AI and automation as key drivers for workforce reductions.
Read on Economic Times Tech →