Semiconductor & Chip Industry — 2026年5月25日 週次レポート
重要な発見
重要な発見（13件）
- 1.The U.S. Department of Commerce announced 9 letters of intent totaling $2.013 billion in federal incentives under the CHIPS and Science Act, a significant expansion from the previously reported $277 million letter of intent [1].
- 2.The SIA projects the U.S. will triple domestic semiconductor manufacturing capacity from 2022 to 2032 (203% growth) and grow its share of advanced logic manufacturing below 10nm to 28% of global capacity by 2032, up from 0% in 2022 [2].
- 3.Imec announced a world-first quantum dot qubit device using High NA EUV lithography on May 18, 2026, describing the most advanced lithography as key for scaling quantum technology, corroborated by IEEE Spectrum's May 19, 2026 report on imec's 15-year Moore's Law roadmap [4] [5].
- 4.NVIDIA launched the Ising open model family on April 14, 2026 — described as the world's first open AI models for quantum computing — delivering up to 2.5x faster and 3x more accurate quantum error correction decoding than the open source standard pyMatching, with adopters including Fermi National Accelerator Laboratory, Harvard, and Lawrence Berkeley National Laboratory [6] (company announcement — may reflect promotional framing).
- 5.NVIDIA's analysis of the Blackwell GB300 NVL72 platform shows tokens per second per GPU rising from 90 on Hopper HGX H200 to 6,000 on Blackwell, resulting in nearly 35x lower cost per million tokens [13] (company announcement — may reflect promotional framing).
- 6.Imec demonstrated the first 3D implementation of a CCD for AI memory applications on May 12, 2026, and IC-Link joined the TSMC 3DFabric Alliance to accelerate advanced packaging and 3D IC innovation for AI, HPC, mobile, and automotive applications [4].
- 7.JEDEC previewed an LPDDR6 roadmap expanding LPDDR into data centers and processing-in-memory, and advanced the DDR5 MRDIMM ecosystem with new memory interface logic, with May 2026 forums focused on next-generation memory for AI, server, cloud, and mobile computing [7].
- 8.Intel launched its Core Ultra Series 3 processors on May 20, 2026, integrating CPU, GPU, and NPU for edge AI robotics compute across hospitality, manufacturing, healthcare, and education use cases [9] (company announcement — may reflect promotional framing).
- 9.ASML announced a strategic partnership with Tata Electronics on May 16, 2026 to advance the semiconductor manufacturing ecosystem in India, marking a concrete advance beyond policy announcements toward equipment and manufacturing partnerships [10].
- 10.NVIDIA and Emerald AI are collaborating with AES, Constellation, Invenergy, NextEra Energy, Nscale Energy & Power, and Vistra to pioneer AI factories as flexible grid participants, built on the NVIDIA Vera Rubin DSX AI Factory reference design [11] (company announcement — may reflect promotional framing).
- 11.Semiconductor Engineering reported that chiplet costs can rise with chiplet-based designs, that low-temperature solders are suddenly critical for chiplets and photonics due to warpage, heat, and brittleness challenges, and that foundry capacity constraints at leading-edge nodes are creating multi-die design opportunities for smaller chip developers [8].
- 12.The WSTS Spring 2026 Semiconductor Forecast is scheduled for publication on June 2, 2026, with the 72nd Forecast Conference having taken place May 18-21 in Toronto [3].
- 13.ABB Robotics reported achieving 99% simulation accuracy using NVIDIA Omniverse libraries in RobotStudio HyperReality, with outcomes including up to 50% reduction in product introduction cycles and up to 80% reduction in commissioning time [15] (company announcement — may reflect promotional framing).
エグゼクティブサマリー（10件）
- •U.S. semiconductor industrial policy is intensifying: the CHIPS Program Office announced $2.013 billion in new letters of intent across 9 recipients, the SIA applauded CHIPS R&D quantum computing awards on May 21, 2026, and top business groups urged Congress to extend and expand the semiconductor tax credit on May 12, 2026 — representing a coordinated policy execution push [1] [2].
- •Quantum computing is emerging as a focal point for semiconductor R&D investment this period, with imec's world-first High NA EUV quantum dot qubit fabrication, NVIDIA's Ising AI models for quantum error correction, and explicit U.S. government CHIPS R&D awards targeting quantum leadership all converging simultaneously [4] [6] [2].
- •Advanced packaging and 3D IC integration remain a primary competitive axis: imec's 3D CCD for AI memory, IC-Link's entry into the TSMC 3DFabric Alliance, JEDEC's LPDDR6 and DDR5 MRDIMM roadmap advances, and Semiconductor Engineering's ongoing chiplet economics coverage collectively signal that heterogeneous integration is now central to AI workload performance scaling [4] [7] [8].
- •NVIDIA's AI infrastructure ecosystem continues to expand rapidly, with first Vera CPUs delivered to Anthropic, OpenAI, and a third leading AI lab on May 18, 2026, Google Cloud announcing A5X bare-metal instances powered by Vera Rubin NVL72, and a joint NVIDIA-Google Cloud developer community surpassing 100,000 developers [12] [18] (company announcements — may reflect promotional framing).
- •Energy efficiency and grid integration are becoming first-order semiconductor infrastructure design priorities: NVIDIA's collaboration with six major energy companies on flexible AI factory grid assets and the metric of tokens per second per watt as the defining measure of modern compute infrastructure reflect a structural shift in how AI semiconductor deployments are evaluated [11] (company announcement — may reflect promotional framing).
- •Physical AI and robotics simulation are creating new categories of semiconductor demand: Intel's Core Ultra Series 3 targeting edge AI robotics, NVIDIA's Isaac and Omniverse frameworks being adopted by ABB, FANUC, KUKA, and Yaskawa with a combined 2 million robot install base, and Altera's FPGA AI Suite for physical AI systems signal a broad new application frontier [9] [14] (company announcements — may reflect promotional framing).
- •Geographic diversification of semiconductor manufacturing is advancing concretely: the ASML-Tata Electronics partnership for India manufacturing, continued U.S. CHIPS Act funding, and Semiconductor Engineering's reporting on Taiwan and Europe packaging buildout reflect systematic effort to broaden the global manufacturing footprint [10] [1] [8].
- •High-NA EUV is transitioning from research milestone to production-relevant technology, with imec using it for quantum dot qubit fabrication, IEEE Spectrum projecting only seven years to the next transistor evolution, and Semiconductor Engineering identifying new mask technology challenges from high-NA EUV adoption converging with curvilinear adoption and data volume growth [4] [5] [8].
- •Chiplet integration is moving from a design trend to a systemic engineering and economic challenge: Semiconductor Engineering's May 2026 coverage documented rising chiplet costs, growing interconnect standard confusion, new workflow requirements for multi-die assemblies, and low-temperature solder reliability problems — indicating the industry is encountering second-order complexity from its packaging strategy [8].
- •The WSTS Spring 2026 Semiconductor Forecast publication scheduled for June 2, 2026 represents the next major data point for quantifying the ongoing industry expansion cycle initiated in 2025, following record semiconductor materials revenue of $73.2 billion and 25% quarter-over-quarter global sales growth reported in the prior period [3].
市場動向
U.S. CHIPS Act Funding and Domestic Capacity Expansion
The U.S. Department of Commerce announced the signing of 9 letters of intent to provide $2.013 billion in federal incentives under the CHIPS and Science Act, reflecting continued momentum in domestic semiconductor manufacturing investment [1]. This aligns with the Semiconductor Industry Association's projection that the U.S. will triple its domestic semiconductor manufacturing capacity from 2022 to 2032, representing a 203% growth rate — the largest projected percent increase in the world over t…
Quantum Computing Breakthroughs Emerge from Semiconductor R&D
Significant quantum computing milestones are emerging from semiconductor research institutions. Imec announced a world-first quantum dot qubit device using High NA EUV lithography on May 18, 2026, describing the most advanced lithography as key for scaling up quantum technology [4]. The EU-supported semiconductor-based quantum pilot line 'SPINS' was also launched to strengthen Europe's leadership in quantum computing [4]. On the AI-for-quantum front, NVIDIA launched the Ising open model family o…
Advanced Packaging and 3D IC Innovation Accelerate for AI Workloads
Advanced packaging and 3D integration continue to intensify as the semiconductor industry addresses AI workload demands. Imec's IC-Link joined the TSMC 3DFabric Alliance, aiming to accelerate advanced packaging and 3D IC innovation for AI, HPC, mobile, and automotive applications [4]. Imec also demonstrated the first 3D implementation of a CCD for AI memory applications on May 12, 2026, paving the way for addressing the memory wall for AI-specific workloads [4]. JEDEC previewed an LPDDR6 roadmap…
AI Factory Infrastructure Evolves with Next-Generation Compute Platforms
The transformation of AI data centers into high-efficiency, grid-integrated infrastructure is advancing with new hardware generations and partnerships. NVIDIA and Google Cloud announced A5X bare-metal instances powered by NVIDIA Vera Rubin NVL72 rack-scale systems at Google Cloud Next, delivering up to 10x lower inference cost per token and 10x higher token throughput per megawatt than the prior generation [12] (company announcement — may reflect promotional framing). NVIDIA's analysis of the Bl…
Physical AI and Simulation-First Robotics Reshape Industrial Semiconductor Demand
Physical AI — the convergence of AI with robotics and industrial automation — is driving new semiconductor requirements for edge compute, simulation, and AI inference. At Hannover Messe 2026, NVIDIA and partners demonstrated AI-driven manufacturing including humanoid robots, vision AI agents, and factory digital twins built on NVIDIA Omniverse libraries and Isaac simulation frameworks [14] (company announcement — may reflect promotional framing). ABB Robotics reported achieving 99% simulation ac…
競合動向
NVIDIA Vera Rubin Platform Deepens AI Infrastructure Ecosystem
Continuing from the previous period, NVIDIA's Vera Rubin platform remains central to AI infrastructure expansion. New this period, NVIDIA and Google Cloud announced at Google I/O 2026 that their joint developer community has surpassed 100,000 developers, with further collaboration on agentic and physical AI. The NVIDIA Vera Rubin DSX AI Factory reference design continues to underpin partnerships with major energy companies including AES, Constellation, Invenergy, NextEra Energy, Nscale Energy & …
U.S. CHIPS Act Funding and India Semiconductor Expansion Advance
Continuing from the previous period, the U.S. Department of Commerce announced the signing of 9 letters of intent to provide $2.013 billion in federal incentives under the CHIPS and Science Act, a new specific figure not reported in the prior period. The Semiconductor Industry Association applauded CHIPS R&D awards to advance U.S. leadership in quantum computing on May 21, 2026, and top business groups urged Congress to extend and expand the semiconductor tax credit on May 12, 2026. The SIA's su…
Quantum Computing Receives Major Semiconductor Industry Investment
A significant new development this period is the U.S. government's $2 billion investment in quantum computing, reported by AnySilicon. Separately, NVIDIA launched the Ising open model family on April 14, 2026, described as the world's first open AI models for quantum processor calibration and error correction, delivering up to 2.5x faster and 3x more accurate decoding than the current open source standard pyMatching. Adopters include Fermi National Accelerator Laboratory, Harvard, Lawrence Berke…
Advanced Packaging, 3D IC, and Next-Gen Memory Standards Accelerate
Continuing from the previous period with new activity, imec's IC-Link joined the TSMC 3DFabric Alliance, confirmed again this period, targeting advanced packaging and 3D IC innovation. Imec also demonstrated the first 3D implementation of a charge-coupled device for AI memory applications on May 12, 2026, described as addressing the memory wall for AI-specific workloads. JEDEC previewed an LPDDR6 roadmap expanding into data centers and processing-in-memory, and advanced the DDR5 MRDIMM ecosystem…
Physical AI and Robotics Simulation Drive New Chip Platform Demand
A growing trend this period is the convergence of physical AI, robotics simulation, and semiconductor platforms. NVIDIA introduced frontier models including Cosmos 3, Isaac GR00T N1.7, and Alpamayo 1.5 at GTC 2026, alongside the Physical AI Data Factory Blueprint and Omniverse DSX Blueprint for AI factory digital twin simulation. ABB Robotics, FANUC, KUKA, and Yaskawa, with a combined install base of over 2 million robots, are integrating NVIDIA Omniverse libraries and Isaac simulation framework…
制度・規制動向
U.S. CHIPS Act Funding Advances with $2 Billion in New Incentives
The U.S. Department of Commerce announced the signing of 9 letters of intent to provide $2.013 billion in federal incentives under the CHIPS and Science Act, representing a significant new funding milestone beyond the previously reported individual awards [1]. This development corroborates the continuing trend of CHIPS Act deployment, with the Semiconductor Industry Association simultaneously applauding CHIPS R&D awards to advance U.S. leadership in quantum computing in a press release dated May…
High-NA EUV and Quantum Dot Qubits Mark Lithography Frontier Advances
Imec announced on May 18, 2026 a world-first quantum dot qubit device using High NA EUV lithography, describing the most advanced lithography as key for scaling up quantum technology [4] (company announcement — may reflect promotional framing). This development is corroborated by IEEE Spectrum's May 19, 2026 report on the next 15 years of Moore's Law according to imec, noting the industry is only seven years from the next evolution in transistors [5]. Semiconductor Engineering reported that mask…
Chiplet Economics and Multi-Die Integration Complexity Intensify
Semiconductor Engineering published multiple analyses in the May 14-21, 2026 period examining the growing complexity and cost challenges of chiplet-based designs. A May 21, 2026 report asked whether chiplet costs will change and whether it will matter, noting that costs can rise with chiplets [8]. A May 14, 2026 analysis highlighted that multi-die assemblies are facing full system-level challenges requiring new engineering workflows, while a May 18, 2026 piece noted that confusion is growing wit…
ASML-Tata Partnership Signals India's Entry into Semiconductor Ecosystem
ASML and Tata Electronics announced a strategic partnership on May 16, 2026 to advance the semiconductor manufacturing ecosystem in India, representing a notable geographic expansion of the global semiconductor supply chain [10]. This development is consistent with the broader supply chain diversification trend identified in the previous reporting period but marks a concrete new milestone, as ASML's involvement signals that India's semiconductor ambitions are advancing beyond policy announcement…
AI Factory Energy Integration Emerges as Semiconductor Infrastructure Priority
A new trend has emerged in the semiconductor and AI infrastructure space: the integration of AI compute facilities as active grid assets rather than passive power consumers. NVIDIA and Emerald AI announced at CERAWeek a collaboration with AES, Constellation, Invenergy, NextEra Energy, Nscale Energy & Power, and Vistra to pioneer flexible AI factories that connect to the grid faster and operate as flexible energy assets, built on the NVIDIA Vera Rubin DSX AI Factory reference design [11] (company…
ソース活動
重要な変化の整理
CHIPS Act Funding: $2.013 Billion in New Letters of Intent
更新The Department of Commerce announced the signing of 9 letters of intent to provide $2.013 billion in federal incentives under the CHIPS and Science Act, as reported across multiple CHIPS.gov pages. This represents an updated figure from the previously reported $277 million letter of intent, indicating continued and expanded CHIPS Act investment activity. [1]
ASML and Tata Electronics Strategic Partnership for India
継続監視ASML announced a strategic partnership with Tata Electronics on May 16, 2026, aimed at advancing the semiconductor manufacturing ecosystem in India. According to Semiconductor Engineering's weekly review, this Tata-ASML deal was highlighted as a notable industry development. This partnership continues to represent a significant expansion of semiconductor manufacturing infrastructure into emerging markets. [10] [8]
Imec Achieves World-First Quantum Dot Qubit Device Using High NA EUV
更新Imec announced on May 18, 2026 a world-first quantum dot qubit device fabricated using High NA EUV lithography, describing the most advanced lithography as key for scaling up quantum technology. This follows Imec's earlier 3D CCD demonstration for AI memory on May 12, 2026, and IC-Link joining the TSMC 3DFabric Alliance. Separately, IEEE Spectrum reported on May 19, 2026 that Imec outlined the next 15 years of Moore's Law, noting we are only seven years from the next evolution in transistors. [4…
SIA Applauds CHIPS R&D Awards for Quantum Computing Leadership
更新The Semiconductor Industry Association issued a press release on May 21, 2026 applauding CHIPS R&D awards aimed at advancing U.S. leadership in quantum computing. The SIA also published a blog on May 19, 2026 on strengthening U.S. semiconductor leadership through smart export control policy, and top business groups issued a press release on May 12, 2026 urging Congress to extend and expand the successful semiconductor tax credit. [2]
Intel Core Ultra Series 3 Launched for Edge AI and Robotics
新規Intel announced on May 20, 2026 the Intel Core Ultra Series 3, positioning it as a new standard for edge AI robotics compute. According to Intel's newsroom, the processor integrates a CPU, GPU, and NPU to deliver edge AI power across use cases spanning hospitality, manufacturing, healthcare, and education. Intel also announced on May 14, 2026 that it was named official compute partner of McLaren Racing. [9]
示唆・見るべき論点（9件）
- 1.The simultaneous convergence of imec's High NA EUV quantum dot qubit fabrication, NVIDIA's Ising AI error correction models, and U.S. CHIPS R&D quantum computing awards in a single reporting period suggests quantum computing is transitioning from long-horizon research to a near-term industrial policy priority — semiconductor companies should assess quantum-readiness in their R&D roadmaps [4] [6] [2].
- 2.The expansion of U.S. CHIPS Act funding from the previously reported $277 million letter of intent to $2.013 billion across 9 new letters of intent, combined with active Congressional tax credit advocacy, signals that the CHIPS Act deployment phase is accelerating into a broader domestic capacity buildout — making the June 2 WSTS forecast a critical data point for validating whether demand is keeping pace with this supply-side investment [1] [3].
- 3.NVIDIA's Vera Rubin platform is establishing a multi-layered ecosystem lock-in: first CPU deliveries to Anthropic, OpenAI, and a third major AI lab, combined with Google Cloud A5X deployment and energy company grid integration partnerships, mean that switching costs for AI infrastructure customers are rising rapidly — creating a significant competitive moat against alternative accelerator platforms [18] [12] (company announcements — may reflect promotional framing).
- 4.The identification of tokens per second per watt as the defining metric of modern compute infrastructure — with NVIDIA claiming over 1 million times improvement from Kepler to Vera Rubin — indicates that energy efficiency is becoming a primary product differentiation axis in AI semiconductors, with implications for data center siting, utility partnerships, and regulatory energy compliance frameworks [16] (company announcement — may reflect promotional framing).
- 5.Semiconductor Engineering's documentation of multiple chiplet integration pain points simultaneously — rising costs, interconnect standard proliferation, workflow complexity, and low-temperature solder reliability — suggests the industry is approaching an inflection point where chiplet standardization bodies and materials innovation will become rate-limiting factors for the broader heterogeneous integration strategy [8].
- 6.ASML's strategic partnership with Tata Electronics, combined with imec's EU Chips Act NanoIC pilot line and continued U.S. CHIPS Act deployment, reflects a three-way geographic diversification of semiconductor manufacturing investment — Europe, India, and the U.S. — that is fundamentally reshaping supply chain risk profiles and creating new opportunities for equipment and materials suppliers in previously underpenetrated markets [10] [4] [1].
- 7.The adoption of simulation-first development methodologies in robotics — with ABB reporting 99% accuracy and up to 80% reduction in commissioning time using NVIDIA Omniverse — indicates that semiconductor platforms enabling high-fidelity physics simulation will become critical infrastructure for industrial automation, creating a durable demand vector for edge AI compute beyond traditional consumer and cloud workloads [15] (company announcement — may reflect promotional framing).
- 8.IEEE Spectrum's reporting that imec's roadmap places the industry only seven years from the next transistor evolution, combined with Semiconductor Engineering's coverage of 2nm process ramps and High NA EUV mask technology challenges, implies that the window for current-generation advanced node economics is narrowing — companies with multi-year fab investment cycles must factor ångström-era transition costs into their capital planning now [5] [8].
- 9.JEDEC's expansion of LPDDR6 into data center and processing-in-memory applications, alongside HBM ecosystem developments, signals an architectural convergence between client and server memory standards driven by AI workload requirements — memory suppliers that can serve both markets with shared development investment will hold a structural cost advantage [7].
信頼度サマリー
今週追跡された 18 件のソース15 件の監視対象 URL から、期間中に新着・更新が検出された記事数。
各ソースは信頼度レベルに応じて重み付けされています。単独ソースの主張は AI 合成時に未検証としてフラグ付けされます。
ソース
The U.S. Department of Commerce announced the signing of 9 letters of intent to provide $2.013 billion in federal incentives under the CHIPS and Science Act, representing a significant expansion of domestic semiconductor manufacturing investment.
関連: Supply Chain & PolicySIA applauded CHIPS R&D awards for quantum computing leadership on May 21, 2026, projects the U.S. will triple domestic manufacturing capacity by 2032 with advanced logic below 10nm reaching 28% of global capacity, and top business groups urged Congress to extend and expand the semiconductor tax credit on May 12, 2026.
関連: Supply Chain & PolicyThe WSTS Spring 2026 Semiconductor Forecast is scheduled for publication on June 2, 2026, with the 72nd Forecast Conference held May 18-21 in Toronto, Canada.
関連: Market DataImec announced a world-first quantum dot qubit device using High NA EUV on May 18, 2026; demonstrated the first 3D CCD for AI memory on May 12, 2026; IC-Link joined the TSMC 3DFabric Alliance; and launched the EU-supported SPINS quantum pilot line.
関連: Manufacturing & R&DIEEE Spectrum reported on May 19, 2026 that imec's roadmap projects the industry is only seven years from the next evolution in transistors, with the next 15 years of Moore's Law mapped out.
関連: Manufacturing & R&DNVIDIA launched the Ising open model family on April 14, 2026, described as the world's first open AI models for quantum computing, delivering up to 2.5x faster and 3x more accurate quantum error correction decoding than pyMatching. (Company announcement — may reflect promotional framing.)
関連: Quantum ComputingJEDEC previewed an LPDDR6 roadmap expanding into data centers and processing-in-memory, advanced the DDR5 MRDIMM ecosystem with new memory interface logic, and announced May 2026 forums focused on next-generation memory for AI, server, cloud, and mobile computing.
関連: Memory StandardsReported chiplet cost challenges, low-temperature solder reliability issues for chiplets and photonics, mask technology challenges from High NA EUV adoption, HBM testing shifts, 2nm process ramps, Taiwan and Europe packaging buildout, and foundry capacity constraints creating multi-die opportunities for smaller developers.
関連: Manufacturing & ChipletsIntel launched Core Ultra Series 3 on May 20, 2026, integrating CPU, GPU, and NPU for edge AI robotics compute across hospitality, manufacturing, healthcare, and education. (Company announcement — may reflect promotional framing.)
関連: Chip Design & ProductsASML announced a strategic partnership with Tata Electronics on May 16, 2026 to advance the semiconductor manufacturing ecosystem in India.
関連: Supply Chain & Geographic ExpansionNVIDIA and Emerald AI announced collaboration with AES, Constellation, Invenergy, NextEra Energy, Nscale Energy & Power, and Vistra to pioneer flexible AI factories as grid assets, built on the NVIDIA Vera Rubin DSX AI Factory reference design. (Company announcement — may reflect promotional framing.)
関連: AI Factory Energy IntegrationNVIDIA and Google Cloud announced A5X bare-metal instances powered by Vera Rubin NVL72, delivering up to 10x lower inference cost per token and 10x higher token throughput per megawatt than prior generation, with joint developer community surpassing 100,000 developers. (Company announcement — may reflect promotional framing.)
関連: AI InfrastructureNVIDIA analysis shows tokens per second per GPU rising from 90 on Hopper HGX H200 to 6,000 on Blackwell GB300 NVL72, resulting in nearly 35x lower cost per million tokens. (Company announcement — may reflect promotional framing.)
関連: AI Factory EconomicsAt Hannover Messe 2026, NVIDIA and partners demonstrated AI-driven manufacturing including humanoid robots, vision AI agents, and factory digital twins built on NVIDIA Omniverse and Isaac frameworks. (Company announcement — may reflect promotional framing.)
関連: Physical AI & RoboticsABB Robotics reported 99% simulation accuracy using NVIDIA Omniverse libraries in RobotStudio HyperReality, with up to 50% reduction in product introduction cycles and up to 80% reduction in commissioning time. (Company announcement — may reflect promotional framing.)
関連: Physical AI & RoboticsNVIDIA described tokens per second per watt as the defining metric of modern computing infrastructure, noting that from the Kepler GPU to Vera Rubin, tokens generated within the same power budget has increased by more than 1 million times. (Company announcement — may reflect promotional framing.)
関連: AI Factory Energy IntegrationAnySilicon reported a U.S. investment of $2 billion in the quantum computing race, GlobalFoundries investing in Playground Global, and Intel and Qualcomm expressing interest in AI chip startup Tenstorrent as top news items.
関連: Market IntelligenceNVIDIA's first Vera CPUs arrived at Anthropic, OpenAI, and a third leading AI lab on May 18, 2026, and NVIDIA CFO Colette Kress discussed Q1 Fiscal Year 2027 chip revenue on May 21, 2026. (Company announcement — may reflect promotional framing.)
関連: AI Compute Platforms