Anton Shilov

Intel will begin layoffs at its Intel Foundry division in mid-July as part of a new restructuring effort to cut costs while focusing on keeping engineering talent.

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

Oracle Cloud Infrastructure will be the first major cloud provider to deploy AMD's Instinct MI350X GPUs and Pensando Pollara 400GbE Ultra Ethernet NICs as part of a massive zettascale AI cluster.

AMD's software stack for AI keeps improving.

Will at least double performance compared to existing MI355X-based solutions.

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.

AMD’s Zen 6-based EPYC ‘Venice’ CPU pushes core count to 256, boosts performance by 70%, and features PCIe 6.0, 1.6 TB/s memory bandwidth, and a new SP7 packaging, according to AMD.

Micron has become the first DRAM vendor to begin sampling 36GB HBM4 memory with a 2048-bit interface and 2TB/s bandwidth.

PCI-SIG has revealed the specs for PCIe 7.0, and its plans for 8.0.

A lot of questions, a lot of answers.

As power consumption of accelerators for supercomputers continues to grow in the coming years, AMD expects zetta scale-class datacenters to consume 500 megawatts, require nuclear reactor.

An expensive delay

China stays mum regarding the performance of its latest supercomputing installations.

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.

Micron has petitioned the Supreme Court to reverse earlier rulings that granted YMTC’s legal team access to 73 pages of confidential 3D NAND documents, citing national security concerns.

Zotac turned its April Fool's joke into reality by launching a new fragrance line designed for gamers.

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.

Sounds like ASE liked the CPUs.

Shipments of AMD's GPUs in Q1 declined almost twofold from Q4 despite the launch of Radeon RX 9070-series products that AMD deems successful.

If ever.

TeamGroup has expanded its high-end PCIe 5.0 SSD lineup to six different families based on four controller vendors to cover a wide range of performance and price points.