semiconductors
-
GlobalFoundries and Marvell are expanding U.S. silicon-germanium manufacturing capacity in Vermont as AI data centers demand faster, more efficient optical networking.
-
Germany’s Federal Agency for Breakthrough Innovation, SPRIND, and the Netherlands’ new National Agency for Disruptive Innovation, NADI, have launched a €40 million program aimed at accelerating European AI chip design. The agencies plan to fund small engineering teams using artificial intelligence to speed development of processors for AI training and inference. Reuters reporting on the…
-
Alibaba unveiled the Zhenwu V900 AI processor and set a goal for Alibaba Cloud data center capacity to exceed 20 GW by 2032.
-
Coherent’s new PhotonLink platform brings lasers, silicon photonics, fiber and detectors together for co-packaged, near-packaged and chip-to-chip optical links in next-generation AI data centers.
-
FutureBit’s HashFly experiment uses a simulated fruit-fly neural network to explore Bitcoin hashing and the theoretical energy limits of computing beyond silicon.
-
Apple’s M6 Mac mini arrives September 22 with a 2-nanometer chip, faster AI hardware and a push toward always-on local AI computing.
-
Alibaba’s new T-Head Zhenwu V900 AI accelerator targets higher performance and massive cluster scale as the company expands its data-center hardware stack.
-
Bitmain’s new 8.9 J/TH S23 XP Hyd. gives the Bitcoin Mining Singularity model a new real-world test of whether ASIC efficiency gains are beginning to slow.
-
Researchers are using optimal-control mathematics to sharply reduce the simulated energy required to switch magnetic states, pointing toward more efficient future memory devices.
-
Researchers are using optimal-control mathematics to explore how magnetic memory could switch states with far less energy and move closer to a fundamental thermodynamic limit.
-
AMD crossed a $1 trillion market capitalization for the first time as renewed AI-hardware demand lifted semiconductor stocks.
-
d-Matrix is building memory-centric AI inference hardware designed to reduce latency and work alongside GPUs as AI systems scale.
