hardware

Cerebras unveils CS-4, claims 30x speed edge over GPUs

Cerebras Systems unveiled the CS-4, its next-generation rack-scale AI accelerator, at the company's Supernova event in San Francisco on Tuesday, claiming the system delivers up to 30 times faster inference than GPU-based solutions. The announcement marks the chipmaker's most aggressive…

Lenovo posts record $26.9B quarter as AI revenue soars

Lenovo Group reported its strongest quarter in company history on Thursday, with revenue surging 43% year-on-year to $26.94 billion for the three months ending June 30, far exceeding analyst expectations of $22.3 billion. The results were powered by accelerating demand…

Morgan Stanley warns ‘chipflation’ pushing firms to rush hardware buys

A growing consensus among major Wall Street banks emerged this week that the global memory chip shortage — branded "chipflation" by Morgan Stanley — will persist as a multi-year headwind, driving enterprises to accelerate hardware purchases and prompting governments to…

Microsoft plans to unveil Maia 300 AI chip this fall

Microsoft is preparing to publicly unveil its next-generation Maia 300 AI chip this fall, potentially as soon as September, as the company looks to ramp up its homegrown silicon effort and reduce its dependence on Nvidia hardware.

Anthropic confirms in-house chip team to build custom silicon for Claude

Anthropic confirmed on Wednesday that it is building an in-house silicon team to design custom chips for its Claude AI models, the first time the company has publicly acknowledged the effort. The announcement signals a new front in the race…

OpenAI says it’s building a ‘family of devices’ for AI

OpenAI president Greg Brockman said the company is building a "family of devices" for interacting with its AI models and that people should "expect them soon," while also pushing back on Apple's trade-secrets lawsuit and acknowledging the ChatGPT desktop app's…

German startup launches room-temperature quantum computers

German quantum computing startup SAXON Q has announced the commercial availability of its SXQ128 and SXQ512 systems — 128-qubit and 512-qubit diamond-based quantum computers that operate entirely at room temperature, marking a leap from the company's previous 10-qubit-class machines to…

Nvidia confirms Vera Rubin is shipping to OpenAI, CoreWeave, and other major customers

Nvidia confirmed on Monday that its next-generation Vera Rubin AI platform has reached full production, with systems now shipping to customers including OpenAI, CoreWeave, Google Cloud, Microsoft Azure, Meta, and Dell, according to a briefing at the company's headquarters first…

OpenAI launches Codex Micro, its first hardware product

OpenAI's first branded hardware product, the Codex Micro, arrives today — and it is not the consumer AI device the industry has been anticipating. The compact programmable macro pad, built in partnership with keyboard maker Work Louder, is designed for…

Nvidia CEO says Vera Rubin AI chip is in production

Nvidia CEO Jensen Huang pushed back against reports that the company's next-generation Vera Rubin AI accelerator platform faces production delays, declaring that "giant amounts of production" are incoming and the system remains on track.

OpenAI denies Apple’s trade secret theft claims as legal battle escalates

OpenAI has denied allegations in a sweeping trade secrets lawsuit filed by Apple, declaring it has "no interest in other companies' trade secrets" as the two tech giants head toward what could become one of the most consequential legal battles…

Apple sues OpenAI alleging systematic trade secret theft

Apple filed a sweeping trade secrets lawsuit against OpenAI on Friday, July 10, accusing the AI company and its senior hardware executives of orchestrating a systematic campaign to steal proprietary information — a legal strike that could disrupt OpenAI's plans…

Apple’s M7 Ultra chip designed to support 1.5TB of memory, Gurman reports

Apple is designing its upcoming M7 Ultra chip to support up to 1.5 terabytes of unified memory, according to Bloomberg's Mark Gurman, a figure that would match the maximum RAM configuration of the 2019 Mac Pro for the first time…

Nvidia denies report that its Kyber AI rack is delayed to 2028

Nvidia pushed back Monday against a report from semiconductor research firm SemiAnalysis claiming its next-generation Kyber NVL144 rack-scale AI system has been delayed by more than 12 months to 2028. "Our roadmap is intact," a spokesperson told Seeking Alpha, in…

Nvidia scraps original four-chip Rubin Ultra design

NVIDIA has canceled the original four-chip design of its Rubin Ultra GPU just three months after unveiling it at GTC 2026, according to semiconductor research firm SemiAnalysis, which posted the findings on June 30. The revised chip reverts to a…

Apple Vision Pro chief Paul Meade leaves to lead OpenAI’s hardware division

Paul Meade, the vice president who led hardware engineering for Apple's Vision Pro headset and its forthcoming smart glasses, is leaving the company to join OpenAI's hardware division, Bloomberg reported Thursday. Meade is expected to exit Apple by next week…

Apple to skip M6 Pro and Max chips, jump to M7 in 2027

Apple is breaking from its established chip cadence by forgoing higher-end versions of its upcoming M6 processor and instead fast-tracking an AI-centric M7 line, according to a Bloomberg report published Tuesday.

Nvidia now powers over 80% of world’s top supercomputers

The 67th edition of the TOP500 list, unveiled at the ISC High Performance 2026 conference in Hamburg, Germany, on Sunday, cements Nvidia's dominance in supercomputing. The company's technologies now power more than 400 of the world's 500 fastest supercomputers —…

Foxconn pegs Nvidia Vera Rubin data center at $47B per gigawatt

Foxconn Hon Hai Precision Industry Co., Ltd. Chairman Young Liu disclosed that a one-gigawatt AI datacenter built around Nvidia's upcoming Vera Rubin architecture would cost approximately $47 billion, with annual electricity bills of roughly $1.3 billion — figures that illustrate…

Samsung builds world’s first 3D stacked transistor at record 42nm pitch

Samsung Electronics announced on Tuesday that its Semiconductor Research Center has demonstrated the world's first 3D Stacked Field-Effect Transistor at a gate pitch of 42 nanometers, a breakthrough presented at the VLSI Symposium 2026 in Honolulu that earned the conference's…

Huawei unveils Hongtu Plan spanning 200 chip types

Huawei on Friday launched what it calls the Hongtu Plan, a sweeping semiconductor and device strategy that aims to span 200 types of chips, 1,200 device categories, and more than 20 industry segments, the company announced at its annual Developer…

Apple’s top AI features in iOS 27 require 12GB devices

Apple confirmed on Monday at its Worldwide Developers Conference that the most powerful on-device AI capabilities coming in iOS 27 will be restricted to devices with at least 12GB of unified memory, leaving the standard iPhone 17 and all iPhone…