
COMPUTE & SILICON
Researchers Achieve 10 Billion Write Cycles in Wurtzite Ferroelectric Memory
A new superlattice structure in aluminum scandium nitride memory enables higher endurance for AI hardware by confining atomic defects.

A new superlattice structure in aluminum scandium nitride memory enables higher endurance for AI hardware by confining atomic defects.

MIT researchers have developed a new fabrication process that allows for the mass production of flexible, transparent silicon-photonics chips.

Qualcomm is developing a new chip architecture to bring data center-level AI performance to mobile devices by optimizing memory-compute proximity.

Market analysts project that the capital expenditure cycle for generative AI will persist through 2028, driven by persistent demand for high-performance hardware.
Applied machine learning, filed daily.
Model releases, silicon, clinical deployment, and the policy shaping them. No digest padding.