Focus ·
Marvell Data Center Revenue Rises 46%
Marvell's fiscal Q2 revenue rose 37% to $2.739 billion, led by 46% data center growth, while its fiscal Q3 revenue guidance midpoint reached $3.15 billion.
中文版40 linked articles
Chips & AI Compute
Semiconductors, accelerators, memory, data centers, energy, and the economics of AI infrastructure.
Latest intelligence
Focus ·
NVIDIA Reportedly Agrees to Buy Hugging Face for $12.9B
The Information reports that NVIDIA has agreed to acquire Hugging Face for $12.9 billion, potentially bringing a key open-model distribution layer into NVIDIA’s ecosystem.
中文版Focus ·
NVIDIA posts $96B revenue and lifts outlook
NVIDIA reported $96.221 billion in second-quarter revenue and $89.023 billion in data-center revenue, both more than doubling year over year. Its $108 billion third-quarter revenue guide exceeded public analyst expectations.
中文版Focus ·
AWS expands NVIDIA GPU plan beyond 3M
AWS and NVIDIA confirmed that AWS will deploy 2 million additional GPUs in 2027 and 2028, taking its publicly announced NVIDIA GPU plan since 2026 to more than 3 million units.
中文版Focus ·
Synopsys raises full-year revenue outlook
Synopsys reported third-quarter revenue of $2.477 billion and non-GAAP EPS of $3.91, above published consensus estimates. It set full-year revenue guidance at $9.690 billion to $9.740 billion and disclosed expected full-year Ansys revenue of $2.98 billion.
中文版Focus ·
OpenAI’s Jalapeño Debuts Against Blackwell
OpenAI released initial Jalapeño inference-chip results, claiming higher work per watt and lower end-to-end latency than NVIDIA Blackwell across three public models.
中文版Focus ·
Apple reshapes desktop Macs with M6 and M5 Ultra
Apple has launched Mac mini models with M6 or M5 Pro chips and Mac Studio models with M5 Max or M5 Ultra. The refresh raises local-AI performance and memory-bandwidth ceilings while also lifting the Mac mini’s U.S. starting price by $300 versus the 2024 M4 model.
中文版Focus ·
Xiaomi Launches Its O3 Flagship Chip
Xiaomi has released its second high-end flagship SoC, O3. Reports say it uses a 3nm process, has 24 billion transistors and is planned to debut in the Xiaomi 18 Fold, advancing the company's in-house chip program beyond O1.
中文版Focus ·
TSMC July Revenue Rose 44.7% Year on Year
TSMC reported July consolidated revenue of NT$467,580 million, up 44.7% year on year, while revenue for the first seven months rose 37.0%.
中文版Focus ·
SpaceX completes $60 billion Cursor acquisition
Cursor has confirmed that SpaceX formally acquired it. The deal moves the coding-agent company from compute partner to SpaceX business, at a $60 billion implied equity value.
中文版Focus ·
NVIDIA Moves Groq 3 LPX Into Full Production
NVIDIA says Groq 3 LPX is now in full production as an interactive inference accelerator for the Vera Rubin platform, while SpaceXAI plans to deploy Vera CPUs for agentic AI.
中文版Focus ·
Alibaba Plans AI Expansion Funding
Alibaba is reportedly preparing an equity offering to expand its AI investment, intensifying the capital race.
中文版Focus ·
NVIDIA Brings Major Capital Firms Into Compute Finance
NVIDIA has partnered with Apollo, BlackRock, Blackstone and other major capital firms to create independent AI compute financing platforms targeting more than $500 billion in third-party capital.
中文版Explore ·
IBM Targets Native Arm Execution in Mainframe Chips
IBM says it is developing a dual-architecture processor for future IBM Z and LinuxONE systems, with each core designed to natively execute Arm and IBM Z or LinuxONE instructions.
中文版Explore ·
Texas Compute Buildout Meets Community Approval
Texas is bringing grid, water, residential, cost and community-impact concerns into data-center development. Governor Abbott’s standards and remarks make local support an increasingly explicit condition for compute expansion.
中文版Explore ·
AMD gains CPU share in a shrinking market
Mercury Research data put AMD's Q2 desktop CPU share at 34.9% and server share at 34.5%, even as desktop CPU shipments fell 20% year over year.
中文版Focus ·
NVIDIA AI server prices rise by over 15%
Bloomberg reports that NVIDIA has told some major customers that servers containing its AI chips will rise by more than 15% in many cases as memory costs surge.
中文版Explore ·
OpenAI growth shifts toward enterprise and Codex
Bloomberg reports that OpenAI is on track for more than $40 billion in annualized revenue, roughly double its late-2025 level. Enterprise ChatGPT adoption and Codex are increasingly central to its commercialization.
中文版Focus ·
NVIDIA reportedly backs Stargate power developer
NVIDIA has reportedly agreed to invest $2 billion in Lancium and commit up to another $1 billion as the power developer secures additional planned electricity.
中文版Explore ·
Europe Scales AI Subsidies Toward Gigafactories
The EU has opened a call for up to seven AI Gigafactories, with up to €10 billion in public support and a target of at least €20 billion in private investment. European AI policy is moving toward sovereign-scale compute supply.
中文版Explore ·
AT&T Uses Open Models to Cap Closed-Model Bills
The Information reports that AT&T plans to use more open models, including NVIDIA Nemotron, to keep employee spending on Anthropic and OpenAI models flat in coming years.
中文版Explore ·
Loudoun’s Permitting Shift Reprices Compute Siting
Loudoun has more than 250 data centers and 117 projects in its pipeline, but the end of by-right development makes local approval a distinct constraint on new capacity.
中文版Focus ·
Marvell Gives Google a Warrant for Up to $12.2 Billion in Stock
Marvell expanded its chip-development partnership with Google and granted it a warrant to buy up to $12.2 billion of Marvell stock.
中文版Explore ·
NVIDIA Expands Its Strategic Exposure Through Major Stakes
Bloomberg says NVIDIA disclosed stakes worth nearly $21 billion in SpaceX and $30 billion in Intel, making capital allocation a new variable in assessing its ecosystem strategy.
中文版Focus ·
Applied Materials grows revenue and raises guidance
Applied Materials reported 25% year-over-year growth in fiscal third-quarter revenue and issued stronger fourth-quarter guidance.
中文版Focus ·
CoreWeave doubles revenue as funding pressure rises
CoreWeave reported $2.575 billion in second-quarter revenue, up 112% year over year. Its expansion still comes with rising cash use, heavy capex plans and reliance on external financing.
中文版Focus ·
Reasoning-state replay becomes a secret-governance surface
Research says encrypted reasoning blocks in proprietary LLM APIs may move across sessions, users and models, while public repositories contained credential traces. OpenAI and Google document replay of such state in multi-step calls.
中文版Focus ·
Anthropic teams with Macquarie and GIC on data center platform
Anthropic, Macquarie Asset Management and GIC reportedly formed an entity to develop, operate and lease AI data centers for Claude demand, initially in the U.S.
中文版Focus ·
NVIDIA Launches Nemotron 3.5 Lightning
NVIDIA expanded the Nemotron 3 family with Nemotron 3.5 Lightning and introduced NeMo Switchyard for multi-model routing in agent workloads.
中文版Focus ·
Meta Opens Muse Glimmer for Local Agents
Meta released the roughly 30B-parameter Muse Glimmer as open weights for local agent workflows, with a quantized model designed to fit systems with 24GB or 32GB of memory or VRAM.
中文版Focus ·
IBM and Together AI Sign $240M Inference Deal
IBM confirmed a $240 million multi-year agreement with Together AI to build an open-model inference cluster on IBM Cloud using NVIDIA HGX B300 systems.
中文版Explore ·
EY Adds an Agent Economics Role to Track AI Costs
EY is hiring an “agent economics” leader to govern AI-agent usage, cost leakage, and value realization. Enterprise AI management may be shifting from deployment volume to workflow-level unit economics.
中文版Explore ·
Memory-price reversal raises AI compute costs
Stanford DAM data and a media analysis put DDR5 at roughly $11.41-$13.28 per GB, near normalized 2007 levels. Memory may be returning as a material cost constraint in AI infrastructure.
中文版Focus ·
Firebird launches an AI Factory in Armenia
Firebird says it has launched an AI Factory in Armenia using NVIDIA accelerated computing and Dell Technologies infrastructure, moving the project from construction plans to a stated launch.
中文版Focus ·
US Department of Energy introduces Genesis-Science-1
The US Department of Energy has launched the Genesis Open Models Initiative and introduced Genesis-Science-1 as its first open-weight model for scientific research.
中文版Focus ·
AMD Reportedly Acquires Taalas to Pursue Hard-Wired AI Inference
The Decoder reports that AMD is acquiring Taalas, whose chips hard-code model weights to trade flexibility for very high inference speed.
中文版Focus ·
Lam revenue reaches $23.2bn
Lam Research’s FY2026 revenue rose about 26% to $23.233 billion and net income reached $7.265 billion, while China represented 26% of June-quarter revenue.
中文版Focus ·
SK hynix Pledges $38.1B for Two Memory Plants
SK hynix has announced a $38.1 billion investment in two new DRAM and NAND facilities in South Korea, expanding its long-term manufacturing bet on memory chips.
中文版Focus ·
KLA grows revenue and lifts its quarterly outlook
KLA's FY2026 revenue reached $13.58 billion and fourth-quarter revenue rose 15% year over year. Its roughly $4.0 billion next-quarter revenue midpoint exceeds the disclosed consensus estimate.
中文版Focus ·
SpaceX and Tesla begin Terafab buildout
SpaceX says Terafab will be built in Grimes County, Texas. The advanced semiconductor project combines logic, memory, and advanced packaging, moving from plan to construction.
中文版