AI hardware and compute news
Hardware & Compute follows the physical side of AI: GPUs and custom chips from Nvidia, AMD, Intel, Google, Amazon and the start-ups, the data centres being built to run them, and the power, water and supply chains they depend on. It is for engineers, infrastructure buyers, investors and energy watchers who need to understand the compute race behind every model launch. Stories cover chip announcements, export rules, capacity deals, data-centre sites and prices, with the numbers as reported and links to the companies' own figures where they publish them.
This week in Hardware & Compute
seven days to 6 October65 Hardware & Compute stories this week, from 52 publishers. The biggest: SpaceX launches Google AI chips into orbit for Project Suncatcher test. SpaceX launched Google's TPUs into orbit on October 1, 2026 via Falcon 9 from Vandenberg Space Force Base. The test advances exploration of orbital data centers, which could offer scalable AI computing using solar power and avoid terrestrial energy and political constraints, though challenges like radiation, cooling, and debris remain. Also this week: NVIDIA launches 64 GB DGX Spark desktop AI system starting at $4,999; NVIDIA Blackwell GPUs power OpenAI's GPT-6 Astra Ultrafast with up to 8x faster token generation and NVIDIA and CoreWeave launch Vera Rubin NVL72 and Vera CPU for agentic AI workloads.
Key topics
most mentioned this weekMost important
this week, by importance-
SpaceX launches Google AI chips into orbit for Project Suncatcher test
SpaceX launched a Falcon 9 rocket carrying Google's Tensor Processing Units (TPUs) as part of Alphabet's Project Suncatcher from Vandenberg Space Force Base in California on October 1, 2026. The mission marks the first…
12 sources primary sourcenote.comfinance.yahoo.comcoinpedia.orgblog.google +7 more -
NVIDIA launches 64 GB DGX Spark desktop AI system starting at $4,999
NVIDIA announced the 64 GB configuration of its DGX Spark desktop AI system, featuring the GB10 Grace Blackwell Superchip, DGX OS, and full NVIDIA AI software stack. The system will be available starting October 23,…
12 sources primary source HN 6techgenyz.comservethehome.compcmag.comnote.com +5 more -
NVIDIA Blackwell GPUs power OpenAI's GPT-6 Astra Ultrafast with up to 8x faster token generation
NVIDIA announced that its Blackwell GPUs are powering OpenAI's GPT-6 Astra Ultrafast, enabling token generation up to eight times faster than the Standard mode. The performance boost targets latency-sensitive workloads…
4 sources primary sourcegithub.comnote.comibtimes.co.ukgeeky-gadgets.com -
NVIDIA and CoreWeave launch Vera Rubin NVL72 and Vera CPU for agentic AI workloads
NVIDIA and CoreWeave announced new infrastructure for agentic AI at CoreWeave Fully Connected. The companies introduced the NVIDIA Vera Rubin NVL72 systems with Spectrum-X networking and the NVIDIA Vera CPU, designed…
3 sources primary sourcenetworkworld.comNVIDIA Blogfinance.yahoo.com -
TSMC manufactures NVIDIA and Apple’s AI chips via foundry model
TSMC, a Taiwanese semiconductor manufacturer, specializes in producing chips designed by other companies like NVIDIA and Apple under a pure-play foundry model. Unlike chipmakers that design and sell their own products,…
3 sourcesfool.comnote.com
All Hardware & Compute stories
newest first-
Applied Materials and Intel expand AI chipmaking collaboration at epic Center
Applied Materials and Intel announced on October 6, 2026, that they are deepening their partnership to accelerate AI chipmaking innovations. The collaboration ties Applied’s EPIC Center in Silicon Valley with Intel’s…
1 sourceUnite.AI -
Xeal to deploy 100,000 Nvidia GPUs in EV charging pods across US
Xeal plans to install 100,000 Nvidia GPUs in Latient Pods at its 1,600+ US EV charging sites, repurposing underused electrical capacity. Each pod holds up to 48 Nvidia Hopper or Blackwell Ultra GPUs, offering sub-20ms…
1 sourceTom's Hardware -
Meta’s Muse boosts AMD stock as AI agents shift workload to CPUs
AMD and Intel are seeing a surge in demand as AI agents like Meta’s Muse and OpenAI’s Dots shift computational workloads from GPUs to CPUs. Muse, launched in early September, topped the Apple App Store in under two…
1 sourceCNBC Technology -
Nokia CEO says data center construction could double without supply constraints
Nokia’s CEO, Justin Hotard, told CNBC that supply chain bottlenecks—particularly shortages of memory chips and energy—are slowing data center construction by half. He argued that if these constraints were lifted,…
1 sourceCNBC Technology -
Ghost launches $3,499 AI agent computer with $11M seed
Ghost, a startup founded by 19-year-old Zain Javaid, has emerged from stealth to announce its first product, Core, a dedicated personal computer for running AI agents. The company simultaneously revealed an $11 million…
1 sourcetechcrunch.com -
Open-source demo runs Qwen models in browser via WASM and WebGPU
A GitHub repository demonstrates how to load and run large language models directly in a web browser using Llama.cpp compiled to WebAssembly (WASM) and WebGPU. The proof-of-concept uses two Qwen models: the smaller…
1 source primary sourcegithub.com -
Open-source voice assistant Speakrail runs locally on an RTX 4090
Speakrail is a full-duplex voice assistant that runs entirely on a user’s machine, requiring only an RTX 4090 GPU. Released via GitHub, it supports real-time conversation, interruptions, and backchanneling like…
1 source primary sourcegithub.com -
AMD benchmarks Gorgon Halo AI chips against Intel, ahead of RTX Spark launch
AMD shared benchmarks for its new Gorgon Halo chips, positioning them as a competitor to upcoming RTX Spark devices launching later this week. The company compared its Ryzen AI Max+ Pro 495 to Intel’s Core Ultra X9…
1 sourceTom's Hardware -
ResearchAndMarkets adds GPUaaS forecast: 27.2% CAGR to $118.8B by 2036
ResearchAndMarkets.com listed a new report titled "GPU as a Service (GPUaaS) Market Size, Share & Trends Analysis - Global Opportunity Analysis and Industry Forecast (2026-2036)." The report estimates the global GPUaaS…
1 sourcefinance.yahoo.com -
Nvidia partners with Microsoft and Valar to speed up US nuclear builds
Nvidia is positioning itself as a central player in accelerating the US nuclear energy sector to meet rising AI power demands. The company launched the PROMETHEUS project with Idaho National Laboratory in February…
1 sourcecryptobriefing.com -
Huawei's Atlas 960 Super Pod cuts power use by over 550 kW per machine
Artificial intelligence workloads are driving rapid growth in data‑center power demand. Analysts estimate that by 2030 AI facilities could consume more than 945 terawatt‑hours each year, with roughly 80% of that energy…
1 sourcegeeky-gadgets.com -
Nvidia raises Shield TV Pro price by $100 to $299.99 due to AI-driven component costs
Nvidia has increased the price of the Shield TV Pro from $199.99 to $299.99, effective October 2, 2026, citing rising component costs driven by AI industry demand. The device, first launched in 2019 with a Tegra X1+…
5 sourcesnotebookcheck.netWired AITom's Hardwarevideocardz.com +1 more -
Malaysia’s Data center colocation market grows 18.6% annually through 2031
Malaysia’s data center colocation market is projected to expand at a compound annual growth rate of 18.6% from 2025 to 2031, driven by AI workloads, hyperscale investments, and government incentives. By 2031, the…
1 sourcefinance.yahoo.com -
AM Intelligence orders 20,000 NVIDIA Rubin GPUs for India and Malaysia
AM Intelligence has placed two binding orders for approximately 20,000 NVIDIA Rubin GPUs, expanding its total committed capacity to around 29,000 units and close to 100 MW across India and Malaysia. The new capacity…
1 sourceciol.com -
User runs Qwen Flash Next Q4 on Mac Mini M5 with SSD streaming
A Reddit user detailed how they ran Qwen Flash Next Q4, a 125 billion parameter open-weight model, on a Mac Mini M5 with 64GB RAM and SSD streaming. The setup leverages hybrid linear attention and an n-gram lookup…
1 sourcefreshworktree.com -
Musk, Huang warn China may close lithography gap by 2030
Elon Musk and Jensen Huang warned that China could significantly narrow its advanced lithography gap by 2030, creating long-term uncertainty for the AI chip supply chain. This assessment coincides with ASML’s plans to…
1 sourcedigitimes.com -
Nvidia may accelerate co-packaged optics rollout, boosting Lumentum, GF Securities says
Nvidia is expanding co-packaged optics faster than anticipated, according to an estimate by GF Securities. The brokerage projects Nvidia's co-packaged optics switch shipments will hit 15,000 units in the fourth quarter…
1 sourcefinance.yahoo.com -
Oura delays IPO as AI wearables face privacy backlash and design failures
Smart ring maker Oura pulled its planned initial public offering on Tuesday, citing uncertainty in the IPO market, though analysts called the last-minute move unusual. The delay comes as the broader AI wearables…
1 sourceCNBC Technology -
Toshiba plans to double HDD production for AI data centers by FY2027
Toshiba has announced plans to double the production capacity of hard‑disk drives (HDDs) used in artificial intelligence data centers by the end of fiscal year 2027. The company aims to raise its share of storage…
1 sourceasia.nikkei.com -
SpaceX may lease AI compute to Microsoft for $150B deal
SpaceX’s AI compute leasing unit discussed a potential $150 billion deal with Microsoft this summer, according to reports. The talks stem from Microsoft’s need for 3 gigawatts of capacity to address a shortfall, though…
2 sourcesfinance.yahoo.comcryptobriefing.com -
Strata runtime doubles Qwen3.8-Flash-Next speed on RTX 4070
A Reddit user reports that the Strata inference runtime significantly outperforms llama.cpp when running the Qwen3.8-Flash-Next model on consumer hardware. The user tested a 125B-parameter Mixture of Experts (MoE)…
1 sourcecarteakey.dev -
Meta releases open-source Muse Gadgets SDK and free Muse Home Link for U.S. subscribers
Meta announced the open-source Muse Gadgets project, providing ESP32 firmware and Linux SDKs to enable developers to build hardware that connects its Muse AI agent to displays, sensors, and smart home devices. As part…
6 sourcesThe DecoderTechCrunch AIUnite.AIThe Verge AI +2 more -
OpenAI’s Jalapeño ASIC beats Nvidia on efficiency in AI chip design
OpenAI’s vice president of hardware, Richard Ho, discussed the company’s Jalapeño inference ASIC—developed with Broadcom—in an interview with Tom’s Hardware. The chip, unveiled at Hot Chips 2026, leverages AI within…
1 sourceTom's Hardware -
gufo-Qwen3.6-35B-A3B-Q6dense - 3095tok/s prefill; 190 tok/s decode on Strix Halo
The Gufo project posted benchmark results for its vertical local inference engine built for AMD Strix Halo hardware (Ryzen AI MAX+ 395 with Radeon 8060S, up to 128 GiB unified memory). The engine uses speculative…
2 sources primary sourcegithub.com -
Volantis raises $88M Series A to build photonic AI inference system
Volantis, a San Francisco semiconductor firm, announced a $88 million Series A on October 1, 2026, co-led by Lachy Groom and Abstract Ventures. The round includes investors like John Doerr, VXI Capital, and angel…
3 sourcesnote.comreuters.comUnite.AI -
Nvidia-backed Nebius buys Inferize to cut AI inference GPU costs
Inferize’s technology targets cold-start delays, which leave GPUs unused while models load or capacity scales. The deal integrates Inferize’s tools into Nebius Token Factory, a managed production-inference platform,…
2 sourcescalcalistech.comfinance.yahoo.com -
Nvidia’s stock hits May record as quarterly revenue jumps 41 percent
Nvidia’s stock reached a new record of $237.87 on October 2, surpassing its May 14 peak of $236.54, despite the company’s share price remaining nearly unchanged since then. The chipmaker’s market capitalization now…
1 sourcefool.com -
Microsoft and LG embed AI into washing machines via Voice Live hub
Microsoft and LG Electronics are collaborating to bring AI into smart homes through appliances, not phones or speakers. At the Microsoft Industry Summit in Seoul, they unveiled Voice Live, a speech-to-speech AI agent…
1 sourcefinance.yahoo.com -
AMD updates Ryzen AI Developer Platform OS with ROCm 10.0 and Linux 7.2
AMD released an updated version of its Ryzen AI Developer Platform OS on September 2026. The new version replaces the Linux 6.18 LTS kernel with Linux 7.2, citing improvements in AMDGPU and AMDKFD drivers as well as…
1 sourcephoronix.com -
Microsoft hints at AI PC upgrades in October event
Microsoft is preparing to unveil new AI features for its PCs at an upcoming event in October. The company has been pushing AI integration into its hardware over the past two years, and this event could mark the next…
1 sourcedigitaltrends.com -
Oracle's Wisconsin AI datacenter may miss 2027 target due to power delays
Oracle's Wisconsin AI datacenter, known as Project Lighthouse, faces a potential delay in its planned 2027 customer delivery. Research firm Aterio warns that the project is at risk because the necessary transmission…
1 source HN 42theregister.com -
Musk says SpaceX may run Nvidia VR72 at 250kW, report says
SpaceX CEO Elon Musk stated on X that the company is "cautiously optimistic" about running its version of Nvidia's VR72 AI system at an average power level of 250 kilowatts. He described this potential milestone as a…
2 sourcesinvezz.comfinance.yahoo.com -
Digital Realty secures $1.4B in new data center leases
Digital Realty, a major data center landlord, reported strong demand in the second quarter. The company signed leases worth $1.4 billion annually in base rent, a record for its portfolio. It also saw a 25.4% cash…
1 sourcefinance.yahoo.com -
Sony adds QSSR AI upscaling to PS5 via AMD collaboration
Sony announced Quick Spectral Super Resolution (QSSR), an AI upscaling feature for the standard PS5, developed with AMD under Project Amethyst. The technology aims to improve graphical sharpness by dynamically…
3 sourcesTom's HardwareUnite.AItheverge.com -
Google shifts AI workloads to TPUs, cutting Nvidia dependency
Alphabet’s Google Cloud revenue grew 82% in the summer quarter, driven by AI infrastructure powered by its own TPUs. Meanwhile, Nvidia’s data center revenue surged 105.8%, fueled by demand from hyperscalers—including…
1 source247wallst.com -
OpenAI deploys Jalapeño ASICs with AMD EPYC Turin CPUs, not Nvidia Vera
OpenAI is deploying its new Jalapeño ASIC chips internally alongside AMD EPYC Turin host CPUs, each equipped with 1.5TB of memory. Richard Ho, OpenAI's VP and Head of Hardware, told Tom's Hardware Premium the choice…
1 sourceTom's Hardware -
Apex Compute builds open-source Mesa Vulkan driver for its hardware
Apex Compute, a California-based startup founded in 2024, is developing an open-source Mesa-based Vulkan driver for its AI accelerators. The company aims to support real-time edge AI inferencing and has already…
1 sourcephoronix.com -
Sea deploys Nvidia Vera Rubin chips to scale AI agents
Sea, the Singapore-based parent company of Shopee, Garena, and Monee, has announced it will deploy Nvidia’s next-generation Vera Rubin platform. This makes Sea the first company in Southeast Asia to adopt the new…
1 sourcee.vnexpress.net -
AMD’s AI chip deals push revenue up 107% but leave valuation at 229x
AMD’s Data Center revenue jumped 107% last quarter, driven by AI chip demand from OpenAI and Meta. OpenAI committed to 6 gigawatts of AMD’s MI450 GPUs, while Meta pledged up to 6 GW, alongside Anthropic’s planned 2 GW…
4 sourcesfinance.yahoo.comfool.comaol.com -
Microsoft says Meta will not buy new chips to scale Muse agent
Meta confirmed that it will not purchase new chips from Microsoft to support its recently launched personal AI agent, Muse. The agent, which debuted on September 8, 2026, quickly rose to the top of app‑store charts and…
1 sourcecryptobriefing.com -
Jacobs launches digital twin platform for AI data centers on NVIDIA Omniverse
Jacobs Solutions announced a new data center digital twin platform built on NVIDIA Omniverse libraries during the GTC 2026 keynote. The solution is designed to simulate and optimize gigawatt-scale AI factories before…
1 sourcecryptobriefing.com -
PaleBlueDot AI raises $200M Series C at $3.2B valuation for compute expansion
PaleBlueDot AI announced on October 1, 2026, it closed a $200 million Series C round at a $3.2 billion valuation, led by ComputeCore, with participation from existing investor B Capital and other global investors. The…
1 sourceUnite.AI -
Cerebras gains 5x inference throughput via disaggregation
Cerebras Systems announced on October 1, 2026, that it achieved a 5x increase in inference throughput using disaggregation—a technique that splits prefill and decode phases across separate hardware pools. The company’s…
1 sourceUnite.AI -
Nvidia authorizes $150 billion in stock buybacks, but AI hardware focus remains
Nvidia’s board approved a $150 billion increase to its share repurchase program, raising the total to $235 billion—the largest single buyback authorization in U.S. corporate history. The move reflects the company’s…
14 sourcesfool.comCNBC Technologyfinance.yahoo.combarrons.com +4 more -
Tenstorrent and Smallest.ai launch on-prem voice AI stack with 4x cost savings
Tenstorrent and Smallest.ai announced a partnership on October 1, 2026, to deliver an on-premises voice AI stack combining Smallest.ai’s Lightning V2 real-time text-to-speech model with Tenstorrent’s Galaxy Blackhole…
1 sourceUnite.AI -
Meta launches Ray-Ban Meta Gen 3 in India with Ranveer Singh
Meta has announced Ranveer Singh as the first brand ambassador for Ray-Ban and Ray-Ban Meta in India, marking a significant expansion for the company's AI hardware division. The announcement coincides with the launch…
1 source primary sourceMeta Newsroom -
Synopsys unveils Autopilot platform for AI-driven chip design agents
Synopsys introduced its Autopilot Platform and AgentEngineer solutions, a suite of domain-specific AI agents for chip development. The portfolio covers verification, system validation, implementation, analog and…
4 sources HN 52Tom's Hardwarenews.synopsys.comThe Decoder -
Deepseek releases open-source TileLang for Huawei Ascend chips
Deepseek and Huawei announced a collaboration to develop open-source software for Huawei’s Ascend AI chips, including TileLang, a programming language designed as an alternative to Nvidia’s CUDA. The release includes…
4 sourcesTom's Hardwarethedailyupside.comThe Decoderreuters.com -
Legato launches AI hearing glasses for $999
Legato announced its flagship AI hearing glasses, called Legato Frames, are now available for purchase starting at $999. The glasses target adults with up to moderate hearing loss and use an AI system that separates…
1 sourceTechCrunch AI -
NVIDIA outlines how AI factories boost return on investment with productivity, durability and flexibility
NVIDIA’s blog post explains how its AI factories maximize return on investment through three key factors: earning capacity, useful life, and demand. Each megawatt-scale factory costs about $60 million, and operators…
1 source primary sourceNVIDIA Blog -
Satlyt raises $8M to run AI on satellites with Google’s Gemma model
Satlyt, founded by former Google and SpaceX engineer Rama Afullo, has secured an $8 million seed round to develop software that enables AI workloads on satellites. The company’s approach avoids building its own…
1 sourceTechCrunch AI -
Opinion: Broadcom's AI revenue growth keeps me adding shares
The author explains why they continue to buy Broadcom shares, pointing to the company's role in custom AI accelerators and its strong financials. Broadcom reported fiscal Q3 revenue of $29.59 billion, up 85.5% YoY,…
1 source247wallst.com -
Call Center Doctors tests DeepSeek V4.1 Flash on Nvidia H200s, finds costs higher than claimed
The Call Center Doctors, a call center consultancy, rented four Nvidia H200 GPUs on September 27 to test DeepSeek V4.1 Flash against Anthropic’s Claude Opus 5.5. The goal was to verify claims that DeepSeek is 80x…
1 sourceTom's Hardware -
Shure launches IntelliMix Bar Pro with Microsoft Teams certification
Shure has released the IntelliMix Bar Pro, an all‑in‑one collaboration bar and touch panel for medium and large meeting rooms that is certified for Microsoft Teams. The device combines Shure’s IntelliMix View AI video…
1 sourcetechnologyrecord.com -
ASML’s EUV monopoly powers Nvidia’s AI chip demand
ASML, a Dutch company based in Veldhoven, dominates the market for extreme ultraviolet (EUV) lithography machines, which are essential for manufacturing advanced AI chips like those made by Nvidia. The company holds an…
2 sourcesirishtimes.comft.com
