NuAxi Stream - Real-Time Semiconductor Intelligence

Thu, Oct 15

19:30Impact score 60
ASML struggling to keep up with demand for chipmaking equipment as AI spending continues to boom
ASML published its Q3 2026 earnings on October 14, 2026, saying the unprecedented spending and investment brought about by the AI boom is not slowing down, with customers already lining up to place orders for its advanced chipmaking tools for 2029 and 2030 output; CEO Christophe Fouquet said the company would struggle to meet demand despite plans announced in July to raise capacity by 30% in 2027 and another 30% in 2028. The company also confirmed it will raise prices starting in 2027, with the full effect showing in 2028. The Q3 2026 report detailed a net profit of €4.96 billion on net sales of €11.7 billion; the company projects Q4 sales of €14.2 billion to €15.2 billion ($15.9 billion to $17 billion) and boosted its full-year sales projections from at least €43 billion to €44-45 billion. ASML's shares rose 2% on Wednesday, up almost 70% so far this year, with its €586 billion market valuation making it Europe's largest firm by market cap. Additionally, Terafab will reportedly receive its first EUV chipmaking tool from ASML in mid-2027; CFO Roger Dassen said it is "still kicking the tires on the exact size of that (Terafab) opportunity."
advanced chipmaking toolsEUV chipmaking toolASMLTom's HardwareCapacity Change
19:15Impact score 60
TSMC confirms it is considering Terafab chipmaking expansion with Elon Musk
Summary: TSMC CEO C.C. Wei confirmed on Thursday, October 15, 2026 that the company is considering expanding its American production capacity beyond Arizona, with Texas as one option; teams have been sent to Texas for evaluation, and TSMC has discussed with a customer attempting to build a fab called Terafab (Elon Musk). If TSMC joins the project, it would insist on operating the fab. Discussions are ongoing and no agreement has been reached yet.
TSMCTom's HardwareCapacity Change
18:30Impact score 0
Russian chipmakers demo homegrown RISC-V and Arm chips with retro games
Russian chipmakers demonstrated homegrown RISC-V and Arm chips with retro games at the Microelectronics 26 (ME26) forum in Sochi on October 11, 2026, as reported in an article published October 15, 2026. The demo featured a Baikal Electronics BE-U1000 microcontroller-based NES clone, an Elvis Hypercom-U SoC, and MCST's Elbrus-based PCs. The source does not state the cause. The event status is observed demonstration.
Tom's Hardware
16:57Impact score 0
RT NEWS: The advanced packaging race is heating up, but #TSMC isn’t fazed by the growing buzz around #Intel’s EMIB. CEO C.C. Wei says TSMC alone can...
According to a repost from TrendForce, TSMC CEO C.C. Wei said that despite the growing buzz around Intel's EMIB advanced packaging technology, TSMC is not fazed, and stated that TSMC alone can mass-produce 5x reticle-size packaging at high yields—and it is already pushing toward 9x. This is a statement by TSMC about its own packaging capability and technology position; the source does not state the cause beyond the reported remarks, and no specific price or volume changes are given.
5x reticle-size advanced packaging9x reticle-size advanced packagingIntel EMIBTSMC, IntelTrendForce Official (Market Data)
12:07Impact score 0
Intel says AI 'gold rush' is over, pushes prompt routing across open hardware
Intel stated at the OCP Global Summit in San Jose that the AI gold rush is over and that AI is moving away from giant models in centralized data centers, promoting prompt routing across open hardware; the source does not state the cause.
DigiTimes Asia
11:45Impact score 58
Two chipmaking bets, one U.S. ambition! Washington is backing #Intel’s advanced-node push while doubling down on #GlobalFoundries’ specialty chips—...
According to TrendForce's post dated Oct 15, 2026 at 3:45 AM (shown as 6h old), Washington is backing Intel's advanced-node push while doubling down on GlobalFoundries' specialty chips—now bolstered by Japan's planned $21B investment in up to two new fabs in New York, with Renesas and Apple already named among the initial customers. The event is in a planned status and has not been implemented.
specialty chipsadvanced-nodeIntel, GlobalFoundries and 4 totalTrendForce Official (Market Data)Capacity Change
10:00Impact score 20
RT NEWS: #Intel CEO Lip-Bu Tan says Chinese frontier models like #DeepSeek, Kimi, and GLM are “good enough” with lower token costs and very fast ado...
According to a news repost by TrendForce, Intel CEO Lip-Bu Tan said that Chinese frontier models such as DeepSeek, Kimi, and GLM are 'good enough,' with lower token costs and very fast adoption. He also posed the question of whether raw performance is still the moat, or whether cost is the new battleground. This is an executive commentary/opinion; the source does not state the cause, and no specific prices or supply-demand data are involved.
Chinese frontier AI modelsDeepSeekKimiIntelTrendForce Official (Market Data)
08:28Impact score 60
[News] TSMC Touts Unmatched 5x Reticle-Size Packaging Prowess and High Yields, Ready to Step In if Rivals Fall Short
Reported on October 15, 2026: TSMC CEO C.C. Wei highlighted the company's ability to mass-produce 5x reticle-size packaging with strong yields, a feat he says rivals have yet to match, and signaled TSMC could step in if rival providers, including Intel's EMIB, struggle to meet customer demand. TSMC is doubling down on advanced packaging despite its own CoWoS capacity constraints, with 9x solutions under development and 14x on the horizon. According to an Anue report in August, TSMC Vice President of Advanced Packaging Technology and Service Jun He said the company's 5.5x reticle-size CoWoS technology has already entered high-volume production, with yields exceeding 98% for multiple AI customers and reaching as high as 99% for some; advancing one generation per year, TSMC expects to scale up to 14x reticle-size packaging by 2029, according to Anue. Wei also reportedly expects the next-generation A14 (1.4nm-class) process to become larger than N2 on stronger customer engagement, slated for risk production in 2027 and high-volume manufacturing in 2028. Intel is reportedly accelerating its 14A roadmap: CEO Lip-Bu Tan is reportedly targeting risk production as early as 1Q27, while Intel's official timeline remains more conservative, with internal 14A products planned for 2H27 and high-volume manufacturing in 2028. According to a July TechPowerUp report citing KeyBanc, Intel reportedly plans to manufacture 80–90% of its upcoming Nova Lake compute tiles internally, reversing an earlier dual-source strategy under which TSMC's N2 process was expected to handle 60–70% and Intel's 18A the rest. Wei stressed that the foundry treats all customers equally, whether IDMs or fabless designers, while factoring their internal manufacturing capacity into its own capacity planning.
5x reticle-size advanced packagingCoWoSA14TSMC, IntelTrendForceCapacity Change
07:57Impact score 60
[News] TSMC Sends Team to Assess Texas Fab Expansion, Confirms Terafab Talks but Insists on Operating Control
TSMC is considering expanding its U.S. manufacturing footprint into Texas, with growing demand from American customers prompting the company to evaluate additional capacity beyond Arizona. On the October 15, 2026 third-quarter 2026 earnings call, Chairman and CEO C.C. Wei said TSMC has already sent teams to Texas to assess conditions for a potential expansion; he noted TSMC has a large U.S. customer base with clients urging the company to expand in the country, and highlighted support from the U.S. government as a factor in considering further investment, with Texas now among the locations under evaluation. On Terafab, Wei confirmed the company is in discussions over the project but said the outcome remains uncertain, and emphasized that any fab involving TSMC would have to be operated by the company itself. This follows a different position from Elon Musk on October 7: according to Reuters, Musk said his companies would build and operate Terafab, suggesting TSMC could potentially sublease part of the facility but would not take an operational role, leaving the potential partnership structure uncertain. Separately, according to Reuters citing Bloomberg, Intel CEO Lip-Bu Tan said Intel would remain involved in the Terafab project. TSMC also expects gross margin dilution from overseas fabs to be closer to 2 percentage points in 2026, widening to 3-4 percentage points over the next few years as more overseas fabs ramp up; it has nearly 20 fabs worldwide, including 13 in Taiwan and approximately five to six overseas facilities under construction. To support sustained demand for its N3 node, volume production at its new Tainan Science Park fab remains on track to begin in the first half of 2027, and it has decided to add another fab in Tainan with volume production scheduled for the first half of 2028; its second Arizona fab has completed construction with equipment installation underway and volume production expected in the third quarter of 2027, slightly ahead of schedule; and TSMC plans to begin 3nm volume production at its second fab in Japan in 2028.
N3 node3nmTSMC, IntelTrendForceCapacity Change

Wed, Oct 14

23:00Impact score 55
New motherboards, old RAM, plenty of life left in yesterday's PCs
According to a report published by The Register on Wednesday, 14 October 2026 at 16:00 UTC, Gigabyte recently announced new motherboards that use DDR4 RAM, a standard that launched in 2014. For AM4 processors, Gigabyte offers the new B450M H ARGB with two DDR4 slots and the B550 X-Eagle with four slots, while Intel-platform buyers can get the B760M Aorus Elite. The article states this is partly due to the current price rises and component shortages induced by large-scale, cloud-based LLMs. In the People's Republic of China, several vendors sell new DDR3 motherboards for circa 15-year-old server CPUs, such as Huananzhi's X79-4D or X79 Supreme, or the Machinist X79 V2.82H, all using Intel's LGA-2011 socket, which launched in 2011. The DDR5 standard was agreed in 2020, with mainstream PC adoption following in 2021. The article also argues that Linux lets older hardware remain in service, making second-hand kit more usable than ever.
DDR4DDR3DDR5Gigabyte, AMD and 5 totalThe Register
17:23Impact score 55
Intel, Micron Urge Taiwanese Chip Suppliers to Invest in US
Intel and Micron Technology urged Taiwanese semiconductor materials, components and equipment suppliers to establish operations in the United States, speaking at the SEMICON West 2026 Taiwan Supply Chain Forum in San Francisco on Oct. 12. Ben Fulmer, vice president of construction and sourcing at Intel Foundry, encouraged Taiwanese companies to invest locally so they are ready when selected as a supplier. Shane Smith, senior director of front-end materials strategic sourcing and supply assurance at Micron, said cooperation with Taiwanese suppliers had enabled the company to relocate a significant portion of its DRAM business to the US, noted that transporting all the equipment and materials is a challenge, and emphasized that investment in the US was necessary to develop the ecosystem. Both companies are expanding US manufacturing: Intel is investing more than $100 billion across Arizona, New Mexico, Oregon and Ohio, while Micron plans to invest more than $250 billion by 2035. The forum was organized to connect Taiwanese suppliers with US manufacturers, academic institutions and government agencies; Intel and Micron delivered keynote addresses. Participating Taiwanese companies included Topco Scientific, Challentech International, Gallant Micro Machining, United Integrated Services and Yesiang Enterprise. The urging is an appeal, not a confirmed investment commitment.
DRAMIntel, Micron TechnologyThe Elec
17:21Impact score 25
Applied Materials Establishes World’s Largest Semiconductor R&D Center
Applied Materials opened the EPIC Center in California with up to $5 billion investment, partnering with TSMC, Intel, Samsung, Micron, and SK hynix to accelerate the commercialization of advanced semiconductor technologies, particularly for AI leadership.
Applied Materials, TSMC and 6 totalThe ElecPartnership/M&A
15:00Impact score 60
Kontron moves VPX board production to Huntsville
Reported on October 14, 2026: Kontron has started US production of its rugged VX307H computing board at subsidiary Nextek in Huntsville, Alabama, and at the same time the product moves from build-to-order to build-to-stock to shorten delivery times for US defense customers. The VX307H is the first product in a broader expansion of Kontron's VPX manufacturing in the US, with the VX3060 and VX30101 planned to follow. Nextek handles the entire production chain in Huntsville, from PCB assembly to final assembly and test; the facility is certified to AS9100 and ISO 9001, is ITAR-registered and works to IPC workmanship standards. According to the company, defense customers are asking for greater supply-chain resilience, faster availability and closer access to support, and bringing VX307H production to the U.S. directly addresses those needs. The VX307H is a 3U VPX platform based on Intel's Xeon D-2700/2800 processors, designed in accordance with the SOSA standard.
3U VPX rugged computing boardKontron, NextekEvertiqCapacity Change
06:53Impact score 62
[News] ASML Warns of 2027–28 Supply Shortfall Despite 30% Annual Capacity Hikes; Pricing Impact Starts in 2027
On its earnings call, ASML warned that even with a planned 30% capacity increase in 2027 followed by another targeted 30% expansion in 2028, its planned capacity would fall short of customer demand in both 2027 and 2028; Reuters, citing CEO Christophe Fouquet, notes that orders for 2028 are already pouring in, with some customers looking to secure equipment as far ahead as 2029 or even 2030. ASML expects to ship around 65 systems, up more than 50% from last year, and the outlook for DUV immersion tools has improved significantly: after initially anticipating a decline in shipments, the company now expects volumes to remain roughly flat year over year. ASML confirmed it will introduce the NXE:3800F EUV scanner in 2027, boosting wafer throughput from today's 230 wafers per hour to 260 wph, a breakthrough the company said is expected to meaningfully scale customer production capacities starting in 2028 and beyond; it is also upgrading key DUV immersion tools scheduled for delivery in 2027 to a record 330 wph. On pricing, ASML said pricing discussions with major customers have concluded, with the impact expected to be minimal in 2026, beginning in 2027 and fully visible from 2028 onward; according to The Elec, ASML is reportedly seeking a 10% price hike on replacement parts supplied to South Korea, with Samsung Electronics and SK hynix already notified. Financially, ASML posted record third-quarter net sales of €11.66 billion, up from €9.33 billion in the previous quarter; net profit surged to €4.96 billion from €2.92 billion and gross margin expanded to 57.5% from 54%. It forecasts Q4 revenue of €14.2 billion–€15.2 billion ($15.9 billion–$17.0 billion) with gross margin around 57%, and raised the lower end of full-year revenue guidance by €1 billion to €44 billion–€45 billion while maintaining the upper end, with gross margin of approximately 56%. The price impact has not yet taken effect and the capacity expansions remain planned.
EUV scannersDUV immersion toolsreplacement partsASMLTrendForceCapacity Change
03:43Impact score 60
[News] SEMI Sees Americas Fab Tool Spending Double to US$40.6B by 2029 as TSMC, Intel, Micron, Samsung Expand
SEMI expects Americas 300mm fab equipment spending to more than double from US$18.1 billion in 2026 to US$40.6 billion in 2029, driven by U.S. investment and expansions by TSMC, Intel, Micron, and Samsung.
300mm fab equipmentTSMC, Intel and 5 totalTrendForceCapacity Change
03:00Impact score 10
Interview with Eric Hong, VP of Engineering at Mixel, a Silvaco Company
Eric Hong joins Mixel as VP of Engineering, bringing 30 years of experience to lead analog and mixed-signal IP development. He highlights the challenges customers face in delivering complex SoCs and emphasizes Mixel's end-to-end support approach to reduce integration risk and accelerate time-to-market.
MIPI PHYSerDes IPMixed-signal IPMixel, SilvacoSemiWikiMarket Analysis

Tue, Oct 13

22:04Impact score 10
CASPA CEO Summit Preview with President Jimmy Cheng
CASPA held its 35th anniversary summit, featuring over 35 CEOs from Silicon Valley including Intel's Lip-Bu Tan and Cerebras's Andrew Feldman. This event highlights industry networking but lacks specific supply chain data. Neutral impact on trading decisions.
Intel, Cerebras SystemsEE TimesMarket Analysis
21:30Impact score 60
Microsoft plans multi-gigawatt rollout of next-gen Azure Maia AI accelerator racks
Microsoft plans a multi-gigawatt rollout of its next-generation Azure Maia AI accelerator, with production deployments planned for 2027. The company claims a 96-accelerator configuration will deliver up to seven times the token throughput of its Maia 200 predecessor and provide 40 percent more tokens per megawatt than leading GPU-based systems. Each Maia 300 accelerator will feature eight HBM4 memory stacks delivering up to 18 TB/s of peak memory bandwidth.
Azure Maia AI acceleratorMicrosoftThe RegisterCapacity Change
18:00Impact score 0
Intel’s Open Software Blueprint: Why Production AI Needs More Than GPUs
Intel proposes an open software blueprint that positions open software as connective tissue linking CPUs, GPUs, data, and business applications, emphasizing upstream contributions to frameworks like PyTorch, SGLang, and vLLM, aiming for Day 0 readiness, describing Xeon as the operational foundation for inference, extending to disaggregated inference, and presenting microservices and enterprise agent toolkits as building blocks; the article notes it is a blueprint, not proof of execution.
CPUGPUdataIntelSemiWiki
11:56Impact score 0
At OCP 2026, #Intel’s GPU D270 makes a case for lower-cost LPDDR5X, while #Microsoft’s Maia 300 doubles down on HBM4. Two very different paths to sc...
At OCP 2026, per a TrendForce post, Microsoft's Maia 300 packs eight HBM4 stacks with 288GB capacity, while Intel's GPU D270 takes a different route with up to 480GB of LPDDR5X. The source describes these as two very different paths to scaling AI inference. The source does not state the cause or any price impact.
HBM4LPDDR5XMicrosoft, IntelTrendForce Official (Market Data)

Analog	$0.68	−5.9%
Connector	$0.75	+0.2%
CPU	$386.74	−21.3%
DDR4	$88.40	−0.2%
DDR52026-W13	$14.89	−11.2%
GPU	$478.20	+8.2%
MCU	$0.68	−61.7%Unusual
MLCC	$0.01	+54.0%Unusual
MOSFET	$0.12	−82.8%Unusual
PMIC2026-W13	$0.44	−12.5%
SSD/HDD	$288.43	n/a

NuAxi Stream - Real-Time Semiconductor Intelligence

Thu, Oct 15

ASML struggling to keep up with demand for chipmaking equipment as AI spending continues to boom

TSMC confirms it is considering Terafab chipmaking expansion with Elon Musk

Russian chipmakers demo homegrown RISC-V and Arm chips with retro games

RT NEWS: The advanced packaging race is heating up, but #TSMC isn’t fazed by the growing buzz around #Intel’s EMIB. CEO C.C. Wei says TSMC alone can...

Intel says AI 'gold rush' is over, pushes prompt routing across open hardware

Two chipmaking bets, one U.S. ambition! Washington is backing #Intel’s advanced-node push while doubling down on #GlobalFoundries’ specialty chips—...

RT NEWS: #Intel CEO Lip-Bu Tan says Chinese frontier models like #DeepSeek, Kimi, and GLM are “good enough” with lower token costs and very fast ado...

[News] TSMC Touts Unmatched 5x Reticle-Size Packaging Prowess and High Yields, Ready to Step In if Rivals Fall Short

[News] TSMC Sends Team to Assess Texas Fab Expansion, Confirms Terafab Talks but Insists on Operating Control

Wed, Oct 14

New motherboards, old RAM, plenty of life left in yesterday's PCs

Intel, Micron Urge Taiwanese Chip Suppliers to Invest in US

Applied Materials Establishes World’s Largest Semiconductor R&D Center

Kontron moves VPX board production to Huntsville

[News] ASML Warns of 2027–28 Supply Shortfall Despite 30% Annual Capacity Hikes; Pricing Impact Starts in 2027

[News] SEMI Sees Americas Fab Tool Spending Double to US$40.6B by 2029 as TSMC, Intel, Micron, Samsung Expand

Interview with Eric Hong, VP of Engineering at Mixel, a Silvaco Company

Tue, Oct 13

CASPA CEO Summit Preview with President Jimmy Cheng

Microsoft plans multi-gigawatt rollout of next-gen Azure Maia AI accelerator racks

Intel’s Open Software Blueprint: Why Production AI Needs More Than GPUs

At OCP 2026, #Intel’s GPU D270 makes a case for lower-cost LPDDR5X, while #Microsoft’s Maia 300 doubles down on HBM4. Two very different paths to sc...

Chip Teardown Lab2 opened

Component price indexNPI, vs. prior period

Trending components