.jpg%3Fwidth%3D720%26quality%3D80%26disable%3Dupscale&s=UEuv_BApK-Q8WZ76w7PHLHhHSrHK0gadRZ5SZJusLHM)
The new Boise-based research hub will target memory, compute and packaging as AI systems push existing technologies.
A major memory chip manufacturer is investing $10 billion in a new research organization focused on developing advanced memory technologies and architectures for AI systems. The investment matters because memory has become a fundamental bottleneck in AI infrastructure, with the challenge shifting from simply making faster chips to rethinking how memory and computing components are physically integrated and positioned relative to each other. Current AI workloads require systems to hold and access more data closer to processors than existing designs allow, and energy consumption from moving data between memory and compute has become as critical as the computation itself. The research hub will operate on a 10-year horizon, signaling that the company expects future AI systems will require architectural changes that conventional manufacturing improvements alone cannot deliver.

As AI clusters scale, network silicon becomes the bottleneck. Here’s a look at Cisco’s Silicon One G300 and P200, alongside Broadcom’s Tomahawk 6 and Nvidia’s Spectrum-6.

TerraPower's nuclear power plant possesses a strategic advantage over competitors, especially when chasing after data center deals.

The AI buildout shows no signs of slowing. And with hundreds of billions of dollars a year going into data centers and GPUs, compute has become the single biggest cost for anyone building AI products. But for all that spending, there still isn’t a straightforward way to put a price on compute — or for firms to hedge their exposure when the price changes. Silicon Data […]
Want to go deeper than the news? Explore live, cohort-based AI courses taught by practitioners.
Browse AI courses on Maven