Google Quantum AI researchers led by Volodymyr Sivak reported this month that reinforcement learning can enable quantum systems to run longer by improving error correction, building on a December 2024…
Marvell Technology announced its Teralynx T100 switch ASIC, a monolithic 100.4 Tb/sec chip built on TSMC's 3nm process with 512 200 Gb/sec SerDes, offering 25% lower power consumption than chiplet-bas…
AI chips now account for roughly one-third of Taiwan Semiconductor Manufacturing Co.'s (TSMC) total revenue, CEO CC Wei said during an earnings call, as the foundry announced an additional $100 billio…
Nvidia's Kevin Deierling and Everpure's Shawn Rosemarin argue that the primary obstacle to enterprise AI production is not compute power but unrefined data, coining the term "AI-ready data" as a durab…
Ethernet switch sales surged 39.8% to $15.36 billion in Q1 2026, outpacing the overall server market's 30.4% growth, driven by AI infrastructure demand. Datacenter Ethernet switch revenue rose 61% to …
Three high-performance computing experts—Jack Dongarra, Torsten Hoefler, and Satoshi Matsuoka—are questioning the necessity of GPUs in AI and scientific computing in a forthcoming paper, prompted by t…
AMD, Broadcom, Meta, Microsoft, Nvidia, and OpenAI formed the Optical Compute Interconnect Multi-Source Agreement to standardize co-packaged optics for AI scale-up networks, settling on a slow-and-wid…
AMD acquired MEXT, a startup that extends DRAM main memory to flash storage, to address soaring DRAM prices driven by GenAI demand. DRAM now accounts for up to 90% of server costs, with prices rising …
Hewlett Packard Enterprise (HPE) announced upgraded high-performance computing (HPC) hardware and software focused on security, data sovereignty, and multi-tenancy at the ISC High Performance 2026 con…
Hewlett Packard Enterprise (HPE) built an AI-first support platform on premises using its GreenLake Intelligence and Private Cloud AI with Nvidia, cutting token costs by over 30 times and saving nearl…
Everpure CTO Rob Lee placed Nvidia at the center of the vendor's AI strategy during the Accelerate Conference in Las Vegas, citing a shift from traditional HPC to accelerated computing. The company un…
Hewlett Packard Enterprise CEO Antonio Neri said at Discover 2026 that the $14 billion Juniper Networks acquisition was perfectly timed to strengthen its datacenter networking portfolio for AI workloa…
AI chip startup Tensordyne unveiled its Napier inference engine, which converts matrix multiplication to logarithmic addition to dramatically boost performance and efficiency. The company claims the a…
QuEra announced its Libra fault-tolerant quantum system, which will be hosted on Amazon Braket in 2028. Libra is a neutral-atom megaquop-class system with over 256 logical qubits and a logical error r…
QuEra announced its new Libra fault-tolerant quantum system, which will be hosted on Amazon Web Services' Braket service starting in 2028. The neutral-atom system is QuEra's first fault-tolerant quant…
Gartner forecasts datacenter power consumption will rise significantly by 2027, driven by AI servers, but the overall share of global electricity remains modest compared to other sectors. Concerns abo…
Oracle Corp. reported a $54.75 billion capital expenditure surge in fiscal 2026 as it builds five massive datacenters totaling 5.5 gigawatts of capacity to support AI processing, with a $638 billion r…
Amazon Web Services has begun shipping its Graviton5 Arm server CPU in new M9g and M9gd instances, featuring 192 Neoverse V3 cores across four chiplets linked by die-to-die interconnects. The chip del…
Genomics teams running GPU-accelerated short-read sequencing pipelines in the cloud are experiencing failure rates of 15 to 40 percent, paying for compute on failed runs that must be restarted from sc…
Broadcom reported $22.19 billion in revenue for its fiscal second quarter, a 47.9 percent increase year over year, driven by surging demand for its custom AI chip design and networking components. The…