AI

Huawei has reportedly developed a chip packaging process technology for its Ascend 910D processor that is comparable to TSMC's leading-edge CoWoS technology.

'What does a human slowly going insane look like to a corporation? It looks like an additional monthly user.'

Contemporary AI models are not silly, Apple just does not have proper hardware to test their limitations, says professor Seokjun Kwon of Sungkyunkwan University.

AMD's software stack for AI keeps improving.

AMD plans to launch its second-generation rack-scale AI platform in 2027 with EPYC 'Verano' CPUs, Instinct MI500X GPUs, and Vulcano 800 GbE NICs, accelerating its roadmap to an annual cadence.

A team has created an application called PartCrafter that creates 3D models, in single or multiple parts, using AI based on 2D images for input.

At London Tech Week, Nvidia CEO Jensen Huang claimed that AI has advanced a millionfold over the past decade, likely referencing explosive growth in GPU performance and system scale.

Large reasoning models fail at complex tasks.

Tesla's Stress tool detects and disables faulty cores in Dojo wafer-scale processors, which power Dojo clusters with millions of cores, without interrupting AI training.

Kioxia is developing a 10M IOPS XL-Flash SSD with peer-to-peer GPU connectivity to eliminate bottlenecks and deliver low-latency, small-block access optimized for AI training and inference workloads.