Semiconductor & Chip Industry — 2026年7月27日 週次レポート
Semiconductor & Chip Industryのニュース＆アップデート — すべての記述に一次ソースのリンク付き。
重要な発見
エグゼクティブサマリー（5件）
- •The week's defining story is the emergence of multi-hundred-billion-dollar bilateral AI semiconductor agreements as the new unit of strategic competition: Samsung-Broadcom ($200B+), SK-NVIDIA ($500B+), and NAVER-NVIDIA-Brookfield ($10B+) collectively represent a structural shift where HBM supply chains and AI factory capacity are being locked up through long-term exclusive partnerships, potentially constraining access for non-partner AI infrastructure builders for years.
- •AMD's Advancing AI 2026 event marked a qualitative escalation in the competitive landscape: the simultaneous launch of AMD Helios (first rack-scale AI system), AMD Instinct MI400 Series with HBM4, a $5B equity investment in Anthropic with 2GW GPU deployment commitment, and expanded Microsoft Azure partnership signals AMD has moved from GPU vendor to full-stack AI infrastructure competitor — the most credible challenge to NVIDIA's platform dominance to date.
- •NVIDIA's transition from announcement to production ramp — Vera Rubin NVL72 running at 350+ factory sites across 30 countries, Spectrum-6 deploying at leading AI factories, and Wistron's Fort Worth plant producing superchips — combined with the Korea AI Summit's $500B+ SK partnership and KAIST research lab, demonstrates NVIDIA is simultaneously executing on hardware production, national AI strategy embedding, and academic talent pipeline development at a pace that is structurally difficult to re…
- •The SEMI lobbying effort to extend the Section 48D Semiconductor Investment Tax Credit introduces a near-term policy risk to the CHIPS Act incentive framework: if the credit lapses, the economics of domestic semiconductor manufacturing investments currently underway could be materially affected, potentially slowing the U.S. manufacturing buildout that TSMC, Intel, and Samsung have committed to.
- •Advanced packaging and novel materials are advancing on multiple fronts simultaneously — imec/ASML/TSMC's 2D-material transistors at 50nm CPP on 300mm wafers, sideways chip stacking for AI memory bandwidth, and hybrid bonding milestones — indicating that the next generation of AI accelerator performance gains will come from packaging and materials innovation as much as from transistor node scaling.
今回の要点（15件）
- 1.Samsung and Broadcom signed an MOU on 2026-07-25 estimated at more than $200 billion across memory and foundry over five years through 2030, covering HBM supply for Broadcom's AI accelerators and Samsung's 2nm and below process technologies [12] (company announcement — may reflect promotional framing).
- 2.SK Group and NVIDIA announced a $500-billion-plus comprehensive partnership on 2026-07-24 including SK Telecom building a 2-gigawatt Vera Rubin DSX AI Factory (first online 2027) and SK hynix entering a long-term HBM codevelopment partnership with NVIDIA [9] (company announcement — may reflect promotional framing).
- 3.NAVER, NVIDIA, and Brookfield announced on 2026-07-24 planned expansion of Korea's AI factory from 55 megawatts to 200 megawatts by 2028, with NVIDIA planning to invest $1 billion into NAVER Corp. and Brookfield entering a nonbinding term sheet to fund up to $9 billion [10] (company announcement — may reflect promotional framing).
- 4.AMD launched AMD Helios, its first rack-scale AI solution, at Advancing AI 2026 on July 22-23, combining Instinct MI455X GPUs, 6th Gen EPYC 'Venice' CPUs, Pensando networking, and ROCm software [13a] (company announcement — may reflect promotional framing).
- 5.AMD and Anthropic announced a strategic partnership to deploy up to 2 gigawatts of AMD Instinct MI450 Series GPUs, with AMD committing a strategic equity investment of up to $5 billion in Anthropic, with first gigawatt deployment beginning H1 2027 [14] (company announcement — may reflect promotional framing).
- 6.AMD also announced Microsoft will deploy next-gen AMD Instinct and EPYC processors on Azure, and launched the AMD Instinct MI400 Series GPUs with HBM4 memory [13] (company announcement — may reflect promotional framing).
- 7.NVIDIA announced on 2026-07-21 that Vera Rubin NVL72 production is ramping at CoreWeave, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure across 350-plus factory sites in 30 countries; CoreWeave benchmarked 10x improvement in tokens per second per megawatt on DeepSeek-R1 versus Grace Blackwell NVL72 [8a] (company announcement — may reflect promotional framing).
- 8.Wistron opened its first U.S. manufacturing facility in Fort Worth, Texas on 2026-07-21 — a 324,000-square-foot plant representing a $700 million commitment producing GB300 Grace Blackwell Ultra Superchips [8b] (company announcement — may reflect promotional framing).
- 9.NVIDIA and KAIST launched a joint AI research lab on 2026-07-23, described as the first joint AI lab between a Korean university and any global technology company, with a $300 million collaboration including $50 million per year in compute contributions [11] (company announcement — may reflect promotional framing).
- 10.Samsung unveiled Galaxy Z Fold8 Ultra, Fold8, Flip8, Galaxy Watch Ultra2, Watch9, and intelligent eyewear with Gentle Monster and Warby Parker at Galaxy Unpacked in London on 2026-07-22 [3] (company announcement — may reflect promotional framing).
- 11.Intel announced collaborations with Fortinet on cybersecurity and with Google Cloud on AI enterprise transformation on 2026-07-21, and highlighted Intel Xeon as first to market with 8000 MT/s RDIMM support [1] (company announcement — may reflect promotional framing).
- 12.SEMI brought semiconductor industry tax leaders to Capitol Hill on 2026-07-22 to urge Congress to extend the Section 48D Semiconductor Investment Tax Credit [5].
- 13.ASML, TSMC, and imec presented a novel 300mm integration route for 2D-material based transistors with 50nm contacted poly pitch at the 2026 IEEE/JSAP Symposium, representing a crucial step in the lab-to-fab transition [15].
- 14.NVIDIA Spectrum-6, a 102.4-terabit-per-second Ethernet switch system, arrived at gigascale AI factories including CoreWeave, Microsoft, Nebius, SpaceXAI, and Tesla on 2026-07-21 [8c] (company announcement — may reflect promotional framing).
- 15.JEDEC listed upcoming events including an Industry Forum on AI Memory Architectures, Testing, and Ecosystem Readiness scheduled for October 15, 2026 in Seoul, and an Automotive Electronics Forum for September 17, 2026 in Santa Clara [16].
市場動向
AI Infrastructure Investment Reaches Unprecedented Multi-Trillion Dollar Scale
This week's announcements collectively signal that AI semiconductor infrastructure investment has entered a new magnitude. Samsung and Broadcom signed an MOU estimated at more than $200 billion across memory and foundry over five years through 2030 [12] (company announcement — may reflect promotional framing). SK Group and NVIDIA announced a $500-billion-plus comprehensive partnership spanning AI factories and next-generation memory [9] (company announcement — may reflect promotional framing). N…
Performance-per-Watt Becomes the Defining Competitive Metric for AI Factories
Multiple sources this week converged on performance-per-watt as the primary economic metric for AI infrastructure. NVIDIA's Vera Rubin NVL72 demonstrated 10x more tokens per second per megawatt than Grace Blackwell NVL72 in CoreWeave's DeepSeek-R1 benchmark, and compared with GB200 NVL72, Vera Rubin NVL72 delivers up to 10x more tokens per megawatt and one-tenth the cost per million tokens [8a] (company announcement — may reflect promotional framing). IEEE Spectrum reported that grid operators a…
Advanced Packaging and 3D Integration Milestones Accelerate Toward Production
IEEE Spectrum reported on 2026-07-08 that stacking chips sideways gives AI more memory, solving a tricky 3D integration problem that could boost bandwidth and keep chips cooler [2]. Imec's ongoing work on W2W hybrid bonding with 200nm interconnect pitch and Sony and imec's high-density backside connectivity module remain active research fronts [15a]. ASML, TSMC, and imec presented a novel 300mm integration route for 2D-material based n and pFETs with 50nm contacted poly pitch at the 2026 IEEE/JS…
Semiconductor Equipment Market Expansion Continues on Record Trajectory
SEMI's forecast of global semiconductor equipment sales reaching a record $229 billion in 2028, reported on 2026-07-14, remains the authoritative market sizing reference for the period [5]. WSTS updated its historical billings report with data through May 2026, and the 73rd WSTS Forecast Conference is scheduled for November 16-19, 2026 in Kobe, Japan [18]. SEMI also brought semiconductor industry tax leaders to Capitol Hill on 2026-07-22 to urge Congress to extend the Section 48D Semiconductor I…
AI-Driven Memory Demand Reshapes HBM Supply Chain Dynamics
Samsung and Broadcom's MOU explicitly covers High Bandwidth Memory supply for Broadcom's next-generation AI accelerators, alongside Samsung's 2nm and below process technologies [12] (company announcement — may reflect promotional framing). SK Group and NVIDIA's $500-billion-plus partnership includes SK hynix entering a long-term AI memory partnership with NVIDIA to codevelop and optimize next-generation AI memory solutions including HBM, with SK Telecom planning to build a 2-gigawatt AI Cloud po…
競合動向
NVIDIA Vera Rubin Enters Full Production Ramp With Gigascale Ecosystem Deployment
NVIDIA announced on 2026-07-21 that Vera Rubin NVL72 production is ramping up with racks running at CoreWeave, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure, spanning 350-plus factory sites in 30 countries [8a] (company announcement — may reflect promotional framing). NVIDIA Spectrum-6, a 102.4-terabit-per-second Ethernet switch system delivering 2x the capacity of previous-generation systems, arrived at gigascale AI factories including CoreWeave, Microsoft, Nebius, SpaceXAI, an…
AMD Launches Full-Stack Agentic AI Portfolio Including First Rack-Scale Solution at Advancing AI 2026
At its Advancing AI 2026 event on July 22-23 in San Francisco, AMD launched AMD Helios, its first rack-scale AI solution described as the world's most powerful AI rack, combining Instinct MI455X GPUs, 6th Gen AMD EPYC 'Venice' CPUs, AMD Pensando networking, and ROCm software [13a] (company announcement — may reflect promotional framing). AMD and Anthropic announced a strategic partnership to deploy up to 2 gigawatts of AMD Instinct MI450 Series GPUs, with AMD committing a strategic equity invest…
Samsung Unveils Galaxy Z Fold8 Ultra and Expands Foundry-Memory Ecosystem Through Broadcom MOU
Samsung unveiled the Galaxy Z Fold8 Ultra, Fold8, Flip8, Galaxy Watch Ultra2, Watch9, and intelligent eyewear developed with Gentle Monster and Warby Parker at Galaxy Unpacked in London on 2026-07-22 [3] (company announcement — may reflect promotional framing). The intelligent eyewear features a Snapdragon AR1 Gen1 Platform, up to nine hours battery life, a built-in camera, and is built in partnership with Google bringing Gemini into the wearable form factor [3a] (company announcement — may refl…
Intel Deepens Cybersecurity and AI Enterprise Partnerships While Xeon Leads Memory Bandwidth
Intel announced on 2026-07-21 a collaboration with Fortinet to advance cybersecurity innovation and strengthen global supply chain resilience, combining Fortinet's purpose-built ASIC expertise with Intel's semiconductor design, packaging, and manufacturing capabilities [1] (company announcement — may reflect promotional framing). Intel also announced a collaboration with Google Cloud to accelerate Intel's AI-enabled enterprise transformation, leveraging Gemini Enterprise and Google Cloud to expa…
NVIDIA Anchors Korea's National AI Strategy With $500B+ SK Partnership and NAVER Factory Expansion
NVIDIA's Korea AI Summit week produced a cascade of sovereign AI commitments. SK Group and NVIDIA announced a $500-billion-plus partnership including SK Telecom building a 2-gigawatt NVIDIA Vera Rubin DSX AI Factory with the first AI factory planned to come online in 2027, and SK hynix entering a long-term AI memory codevelopment partnership [9] (company announcement — may reflect promotional framing). NAVER, NVIDIA, and Brookfield announced expansion of NAVER's AI factory at GAK Sejong from 55 …
制度・規制動向
SEMI Brings Semiconductor Tax Leaders to Capitol Hill to Extend Section 48D Investment Tax Credit
SEMI announced on 2026-07-22 that it brought semiconductor industry tax leaders to Capitol Hill to urge Congress to extend the Section 48D Semiconductor Investment Tax Credit [5]. This active lobbying effort signals that the semiconductor industry views the Section 48D credit — a key financial incentive under the CHIPS and Science Act — as at risk of expiration and is mobilizing to secure its continuation. The outcome will directly affect the economics of domestic semiconductor manufacturing inv…
JEDEC Advances HBM4 and Wide Bandgap Standards as AI Memory and Power Ecosystems Mature
JEDEC's homepage listed as recent press releases the new SPHBM4 standard enabling HBM4-class bandwidth on organic substrates, new SiC guidelines to improve reliability and evaluation in power electronics, and advances to the DDR5 MRDIMM ecosystem [16]. The JEDEC homepage also listed upcoming events including an Industry Forum on AI Memory Architectures, Testing, and Ecosystem Readiness scheduled for October 15, 2026 in Seoul, and an Automotive Electronics Forum scheduled for September 17, 2026 i…
SEMI Childcare and Workforce Initiative Expands Semiconductor Talent Pipeline in CHIPS Act States
SEMI published details on its Chips and Childcare initiative, funded by the W.K. Kellogg Foundation and Dave and Lucile Packard Foundation, providing free technical assistance to semiconductor companies across Arizona, Ohio, and Michigan to leverage federal and state childcare programs and support employees' childcare needs as a workforce development strategy [17] (2026-07-24). This initiative addresses a structural workforce constraint in the CHIPS Act manufacturing buildout — childcare access …
ソース活動
先週からの変化
AMD Launches Helios Rack-Scale AI System and Secures Anthropic Partnership at Advancing AI 2026
AMD launched AMD Helios, its first rack-scale AI solution combining Instinct MI455X GPUs, 6th Gen EPYC 'Venice' CPUs, Pensando networking, and ROCm software, at Advancing AI 2026 on July 22-23. AMD and Anthropic announced a strategic partnership to deploy up to 2 gigawatts of AMD Instinct MI450 Series GPUs with AMD committing a strategic equity investment of up to $5 billion in Anthropic, with first gigawatt deployment beginning H1 2027. AMD also launched the Instinct MI400 Series GPUs with HBM4…
NVIDIA Vera Rubin Enters Full Production Ramp and Wistron Opens $700M Fort Worth Manufacturing Plant
NVIDIA announced on 2026-07-21 that Vera Rubin NVL72 production is ramping at CoreWeave, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure across 350-plus factory sites in 30 countries. CoreWeave benchmarked 10x improvement in tokens per second per megawatt on DeepSeek-R1 versus Grace Blackwell NVL72. Wistron opened its first U.S. manufacturing facility in Fort Worth, Texas — a 324,000-square-foot plant representing a $700 million commitment producing GB300 Grace Blackwell Ultra Sup…
Samsung-Broadcom $200B+ MOU and SK-NVIDIA $500B+ Partnership Lock Up HBM Supply Chain
Samsung and Broadcom signed an MOU on 2026-07-25 estimated at more than $200 billion across memory and foundry over five years through 2030, covering HBM supply for Broadcom's AI accelerators and Samsung's 2nm and below process technologies. Separately, SK Group and NVIDIA announced a $500-billion-plus partnership on 2026-07-24 including SK Telecom building a 2-gigawatt Vera Rubin AI factory (first online 2027) and SK hynix entering a long-term HBM codevelopment partnership with NVIDIA. [12] [9]…
SEMI Lobbies Congress to Extend Section 48D Semiconductor Investment Tax Credit
SEMI brought semiconductor industry tax leaders to Capitol Hill on 2026-07-22 to urge Congress to extend the Section 48D Semiconductor Investment Tax Credit, signaling the industry views this key CHIPS Act financial incentive as at risk of expiration. [5]
ウォッチリスト — 今後の締切
AMD fiscal second-quarter 2026 financial results release
ソース: AMD Newsroom — Helios Launch, Anthropic Partnership, MI400 SeriesSEMICON Taiwan 2026 opens (TaiNEX 1 & 2, Taipei)
ソース: SEMI — Section 48D Tax Credit LobbyingJEDEC Automotive Electronics Forum, Santa Clara, CA
ソース: JEDEC — AI Memory Forum and Automotive Electronics Forum ScheduledSEMICON India 2026 opens (Yashobhoomi, Delhi)
ソース: SEMI — Section 48D Tax Credit LobbyingJEDEC Industry Forum on AI Memory Architectures, Testing, and Ecosystem Readiness, Seoul
ソース: JEDEC — AI Memory Forum and Automotive Electronics Forum Scheduled示唆・見るべき論点（10件）
- 1.The Samsung-Broadcom and SK-NVIDIA memory partnerships, totaling over $700 billion in combined commitments, effectively pre-allocate a significant share of Samsung and SK hynix's HBM production capacity through 2030 — AI infrastructure builders without similar long-term agreements may face supply constraints as HBM demand continues to outpace production capacity additions.
- 2.AMD's $5 billion equity investment in Anthropic is structurally different from a supply agreement: it creates a financial alignment that incentivizes Anthropic to optimize its models for AMD hardware, potentially accelerating ROCm software maturity and closing the software ecosystem gap that has historically been AMD's primary competitive disadvantage versus NVIDIA [14] (company announcement — may reflect promotional framing).
- 3.NVIDIA's Korea AI Summit produced commitments spanning AI factories, memory codevelopment, university research, and national AI strategy — a multi-layer embedding that converts Korea's semiconductor and AI ambitions into dependencies on NVIDIA's full stack, making competitive displacement require not just better hardware but an alternative national AI strategy [9] (company announcement — may reflect promotional framing).
- 4.CoreWeave's benchmark showing 10x improvement in tokens per second per megawatt on Vera Rubin NVL72 versus Grace Blackwell NVL72 is the first third-party measured performance data for the platform — its significance is that it validates NVIDIA's performance-per-watt claims under real production conditions, not just internal benchmarks [8a] (company announcement — may reflect promotional framing).
- 5.SEMI's active lobbying to extend Section 48D on 2026-07-22 — bringing industry tax leaders to Capitol Hill — signals the semiconductor industry is not confident the credit will be automatically renewed, introducing a policy risk variable into the economics of the $500B+ domestic manufacturing buildout that TSMC, Intel, Samsung, and others have committed to [5].
- 6.The imec/ASML/TSMC 2D-material transistor demonstration at 50nm CPP on 300mm wafers, using single-patterning EUV lithography, is significant because it shows the lab-to-fab transition for 2D TMD materials is now a process engineering challenge rather than a fundamental physics barrier — the timeline to industrial adoption has materially shortened [15].
- 7.AMD's simultaneous announcements of AMD Helios, MI400 Series with HBM4, EPYC 'Venice' with 256 cores and 512 threads, and partnerships with Microsoft, Anthropic, Meta, and OpenAI represent a coordinated full-stack narrative that mirrors NVIDIA's platform strategy — the question is whether AMD's ROCm software ecosystem can achieve the developer adoption needed to sustain these hardware commitments [13a] (company announcement — may reflect promotional framing).
- 8.Samsung's Galaxy Unpacked intelligent eyewear, built on Snapdragon AR1 Gen1 with Google Gemini integration, represents Samsung's entry into the AI wearable form factor that Apple, Meta, and Google are also pursuing — the semiconductor content (custom AR SoC, sensors, connectivity) in eyewear could become a meaningful new chip demand category as the form factor matures [3a] (company announcement — may reflect promotional framing).
- 9.The JEDEC Industry Forum on AI Memory Architectures, Testing, and Ecosystem Readiness scheduled for October 15, 2026 in Seoul — in the heart of Korea's semiconductor ecosystem — signals that HBM4 standardization and testing methodology are the next frontier for formal industry alignment, with timing that coincides with SK hynix's HBM4 production ramp [16].
- 10.Intel's positioning of Xeon as first to market with 8000 MT/s RDIMM support, framed as a competitive differentiator in data center memory bandwidth, suggests Intel is competing on memory subsystem performance rather than GPU compute for AI inference workloads — a narrower but defensible position in the AI data center market [1] (company announcement — may reflect promotional framing).
信頼度サマリー
今週引用したソース 18 件あなたが選んだ 15 件の監視URLから検出（1つのURLから複数記事が出ることがあります）。
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。
参照ソース一覧
Intel announced collaboration with Fortinet on cybersecurity and supply chain resilience, collaboration with Google Cloud leveraging Gemini Enterprise for agentic AI workforce transformation, and highlighted Intel Xeon as first to market with 8000 MT/s RDIMM support on 2026-07-20. (Company announcement — may reflect promotional framing.)
IEEE Spectrum covered sideways chip stacking for AI memory bandwidth (2026-07-08), optical tech for robot AI updates (2026-07-25), and grid operators weighing 800 VDC for data centers (2026-07-16) during the reporting period.
Samsung unveiled Galaxy Z Fold8 Ultra, Fold8, Flip8, Watch Ultra2, Watch9, and intelligent eyewear at Galaxy Unpacked in London on 2026-07-22. Samsung and Broadcom signed an MOU on 2026-07-25 estimated at more than $200 billion across memory and foundry over five years through 2030. (Company announcement — may reflect promotional framing.)
Published 'Data Center Chokepoints Tied To AI, Political Pressure, Supply Chain' on 2026-07-20 and 'From Highways To Health Care: Portability Proves Key To Physical AI Success' on 2026-07-22. Blog Review on 2026-07-22 covered 3D-IC security and chiplet interoperability.
SEMI brought semiconductor industry tax leaders to Capitol Hill on 2026-07-22 to urge Congress to extend the Section 48D Semiconductor Investment Tax Credit. SEMI also published blog highlights from its 2026 Surface Preparation & Cleaning Conference on 2026-07-21.
CHIPS.gov content unchanged this week; TSMC's incremental $100 billion U.S. investment linked to the U.S.-Taiwan Trade and Investment deal remains listed as background context. [UNCHANGED BACKGROUND — not cited as new]
NVIDIA announced $500B+ SK Group partnership (2026-07-24), NAVER-NVIDIA-Brookfield 200MW factory expansion (2026-07-24), NVIDIA-KAIST joint AI research lab with $300M collaboration (2026-07-23), and Vera Rubin production ramp at 350+ factory sites (2026-07-21). (Company announcement — may reflect promotional framing.)
NVIDIA Blog covered Vera Rubin NVL72 production ramp with 10x tokens/megawatt benchmark (2026-07-21), Wistron Fort Worth $700M manufacturing plant opening (2026-07-21), Spectrum-6 102.4 Tb/s switch deployment at gigascale AI factories (2026-07-21), and Korea AI Summit developments (2026-07-24). (Company announcement — may reflect promotional framing.)
SK Group and NVIDIA announced $500B+ partnership including SK Telecom building 2-gigawatt Vera Rubin DSX AI Factory (first online 2027) and SK hynix long-term HBM codevelopment partnership with NVIDIA. (Company announcement — may reflect promotional framing.)
NAVER, NVIDIA, and Brookfield announced expansion of NAVER's AI factory from 55MW to 200MW by 2028; NVIDIA plans $1B investment in NAVER Corp.; Brookfield nonbinding term sheet for up to $9B. NAVER intends to expand to 1 gigawatt total. (Company announcement — may reflect promotional framing.)
NVIDIA and KAIST launched joint AI research lab at KAIST Kim Jaechul Graduate School of AI in Seoul, described as first joint AI lab between a Korean university and any global technology company. $300M collaboration with $50M/year compute contributions, funding at least 10 KAIST researchers annually. (Company announcement — may reflect promotional framing.)
Samsung and Broadcom signed MOU estimated at more than $200 billion across memory and foundry over five years through 2030, covering HBM supply for Broadcom's AI accelerators and Samsung's 2nm and below process technologies including 2.3D and 2.5D advanced packaging. (Company announcement — may reflect promotional framing.)
AMD launched AMD Helios rack-scale AI solution, AMD Instinct MI400 Series GPUs with HBM4, and 6th Gen EPYC 'Venice' CPUs at Advancing AI 2026 on July 22-23. AMD and Anthropic announced strategic partnership for up to 2GW of MI450 Series GPUs with AMD committing up to $5B equity investment in Anthropic. Microsoft to deploy next-gen AMD Instinct and EPYC on Azure. (Company announcement — may reflect promotional framing.)
AMD and Anthropic announced strategic partnership to deploy up to 2 gigawatts of AMD Instinct MI450 Series GPUs in AMD Helios, with AMD committing strategic equity investment of up to $5 billion in Anthropic. First gigawatt deployment beginning H1 2027. (Company announcement — may reflect promotional framing.)
ASML, TSMC, and imec presented novel 300mm integration route for 2D-material based n and pFETs with 50nm contacted poly pitch at 2026 IEEE/JSAP Symposium on VLSI Technology and Circuits, with 94% operational transistors. First demonstration of scaled nFETs and pFETs at industry-relevant dimensions using single-patterning EUV lithography.
JEDEC homepage listed upcoming Industry Forum on AI Memory Architectures, Testing, and Ecosystem Readiness for October 15, 2026 in Seoul, and Automotive Electronics Forum for September 17, 2026 in Santa Clara. SPHBM4 standard, SiC guidelines, and DDR5 MRDIMM advances listed as recent press releases.
SEMI published details on Chips and Childcare initiative providing free technical assistance to semiconductor companies in Arizona, Ohio, and Michigan to leverage federal and state childcare programs as a workforce development strategy, funded by W.K. Kellogg Foundation and Dave and Lucile Packard Foundation.
WSTS updated its historical billings report with latest data from May 2026. The 73rd WSTS Forecast Conference is scheduled for November 16-19, 2026 in Kobe, Japan.
Semiconductor & Chip Industryを毎週、自動で監視
このレポートは一次ソースのみから生成されています。テーマとソースを選べば、引用付きレポートが毎週届きます。7日間無料トライアル・$33/月から。
無料トライアルを始める