In the chip world, one branch does the actual 'computing' — from the NVIDIA GPUs that trained ChatGPT, to the CPUs in servers and phones, to the chips Google and Amazon design themselves. This is where a $5 trillion company was born, where NVIDIA holds an almost total grip on the market — and where its biggest customers are trying to 'escape' by building their own chips.
Contains
Theme index· base 100 · USD total return
No index history for this theme yet.
Why is Logic, Compute & Connectivity Processors moving?
Q2 2026
▲2▼2
AI chip demand surges, but memory shortage and Nvidia pricing pressure bite
▲
AI infrastructure spending boom Hyperscaler AI spending could hit $755B in 2026, with Nvidia posting $81.6B quarterly revenue and AI infrastructure potentially reaching $3–4T annually by 2029. This confirms massive demand for logic and connectivity chips.
It shows the primary demand driver for the sector.
▲
Intel-Apple foundry deal and custom AI chip momentum Intel's Apple foundry deal, Qualcomm's and AMD's data center push, and Broadcom's and Marvell's custom AI chips (Broadcom up 143%) validate the theme, showing broadening opportunities beyond Nvidia.
It highlights new competitive dynamics and revenue streams.
▼
Memory shortage squeezes consumer demand A severe memory shortage—Apple's 'hundred-year flood'—is raising device costs, forcing smartphone makers to cut 2026 shipments up to 30%, and squeezing consumer demand for logic chips.
It is a major supply-side risk affecting the entire sector.
▼
Nvidia pricing pressure and export-control enforcement Custom chips and new entrants pressure Nvidia's dominance and pricing (B200 rental rates fell from $6.11 to $4.22), while Taiwan's raid on Super Micro over alleged Nvidia chip smuggling to China escalates export-control enforcement.
It shows competitive and regulatory headwinds for the sector leader.
Latest
▲3▼1
AI chip demand broadens; China access opens; memory and digestion risks build
▲
China may let Alibaba and ByteDance buy Nvidia's new RTX Pro 5500 chips Chinese regulators signaled Alibaba and ByteDance may be allowed to buy Nvidia's new RTX Pro 5500 processors, asking about order sizes and uses. This could reopen a huge market that U.S. export rules had largely closed, adding demand for logic and compute chips.
A new potential reopening of China demand directly affects the theme's biggest end-market.
▲
Custom AI chip demand broadens to Google TPUs and Marvell Piper Sandler sees Google's external TPU sales jumping from about $8 billion in 2026 to $104 billion by 2028, and Marvell began shipping custom AI chips to Google. This shows the custom processor market widening beyond Nvidia, lifting the whole logic and connectivity group.
New third-party forecasts and actual Marvell shipments show custom compute demand broadening.
▲
TSMC 2nm output and possible new Texas plants expand leading-edge supply TSMC may reach 120,000 wafers a month of 2nm output by end-2026 as Apple, Nvidia, AMD, Qualcomm and MediaTek raise orders 10-20%. It is also weighing tens of billions in new Texas plants. More leading-edge capacity supports future processor supply.
New capacity plans and faster 2nm ramp address the supply constraint that has limited the theme.
▼
Nvidia's Huang warns of a future 'digestion' period and memory costs stay high Nvidia's CEO said the AI buildout will eventually create more computing capacity than needed, though not for two to three years. Meanwhile Micron earnings loom with memory prices still elevated. This is a real counterweight: today's demand is strong, but a future slowdown and high memory costs could pressure chipmakers.
It is the main new cautionary signal and balances the positive demand news.
Q3 2026
▼3▲1
AI chip demand broadens but memory and China risks bite
▲
Broadening AI chip demand and mega-deals Broadcom's Apple pact, Nvidia's $500B+ partnerships, and hyperscaler spending near $830B show demand spreading beyond GPUs to CPUs, TPUs, and custom chips. Nvidia's quarterly revenue hit $96.2B, up 106%.
This is the core positive force: massive, broadening demand for logic and connectivity chips.
▼
Memory shortage squeezes margins and earnings Memory prices jumped 5x–7x, pressuring chipmaker margins (Nvidia guided to ~74%) and potentially cutting S&P 500 earnings by ~10%. This shortage raises costs and limits consumer device demand.
It is a major new risk that directly threatens profitability across the sector.
▼
China competition and export restrictions China's free Kimi K3 model and DUV lithography progress triggered selloffs and threaten ASML's dominance. US-China restrictions cost Nvidia $400M and zero China data center revenue.
It highlights new competitive and geopolitical threats that could undermine Western chip leadership.
▼
Signs of cooling demand and margin pressure Nvidia's CEO warned of eventual demand 'digestion,' and Broadcom's margins are already declining. These signals suggest the AI boom may cool, posing a risk to future growth.
It provides a forward-looking counterweight: even strong demand may not sustain current growth rates.
Semtech Fair Value Raised to US$210.33 as Analysts Back Data Center and LoRa
Semtech's modeled fair value has been lifted from US$204.83 to US$210.33, reflecting a modestly higher assessment of the stock's worth in the updated framework. The revision follows a broad pattern of raised price targets from Oppenheimer, Roth Capital, Morgan Stanley, UBS, Seaport Research, and BMO Capital, tied largely to stronger expected contributions from data center and LoRa. Oppenheimer noted that data center exposure has shifted from about 5% of sales a couple of years ago to a figure it expects to exceed 40% by the end of the current year, while Roth Capital and UBS said data center and LoRa together are approaching 60% of the mix and supporting expanding gross margins over time. Morgan Stanley kept an Equal Weight rating even after lifting its target, and Roth Capital flagged potential confusion around guidance following the divestiture of cellular modules, which removes at least US$40m of quarterly sales and associated earnings. Alongside the fair value change, the revenue growth assumption moved from 27.53% to 34.83%, the net profit margin assumption from 24.19% to 30.20%, the future P/E from 58.82x to 32.02x, and the discount rate from 11.27% to 11.40%.
Arm's Data-Center Chip Push Tests Its Licensing Moat
Arm Holdings is moving beyond designing processors for others by selling its own data-center CPU, a shift that could capture more revenue per system but changes its relationship with the customers that license its technology. In its fiscal 2027 first-quarter results released July 29, Arm reported $1.29 billion of revenue, with royalties of $715 million, up 22%, and licensing and other revenue of $574 million, up 23%, while data-center royalty revenue more than doubled. Arm said its AGI CPU had been delivered to initial customers and that it had secured capacity for a $1 billion revenue opportunity across fiscal 2027 and fiscal 2028, with management describing demand exceeding $2 billion over those two years, a figure it characterized as demand commentary rather than recognized revenue or a guaranteed order book. At a September 30 market-data snapshot, Arm traded at roughly 59 times enterprise value to trailing revenue and about 205 times trailing free cash flow, and the company had 15,521,694 shares sold short at the September 15 settlement with 3.9 days to cover. The bull case rests on Arm earning royalties from customers' processors while using its own product to accelerate adoption of its architecture, while the bear case is a gradual erosion of neutrality as customers grow reluctant to share roadmaps with a supplier competing for the same server budget.
Artificial Intelligence › AI Compute & Accelerator Silicon Competition
ARM · Capital · Positive Arm reported fiscal Q1 revenue of $1.29B with royalties up 22% and licensing up 23%, plus a $1B capacity-secured revenue opportunity.
ARM · Competition · Neutral Arm's move to sell its own data-center CPU could capture more revenue per system but risks eroding its licensing neutrality with customers who compete for the same server budget.
Qualcomm Unveils Snapdragon Platforms for On-Device Agentic AI
Qualcomm Incorporated unveiled new Snapdragon platforms designed to run AI agents directly on consumer devices at its Snapdragon Summit, as the company positions agentic AI as its next growth phase beyond smartphones. CFO Akash Palkhiwala said the two new chips target the high-end smartphone market with best-in-class gaming and productivity performance and set up the platform for agentic AI. The company is expanding across AI-powered devices, automotive systems, data centers, and physical AI applications, and earlier this month announced a multi-generation collaboration with Amazon to develop customised silicon for large-scale AI data centers focused on AI inference. Qualcomm's forward GAAP P/E of 15.59x sits 6.76% below its 5-year average of 16.72x and below the sector's forward GAAP P/E of 29.51. Hedge fund ownership slipped from 78 funds at the end of Q4 2025 to 71 at the end of Q1 2026, and short interest reached 3.53% of the float as of August 31, 2026.
Artificial Intelligence › Agentic AI & Autonomous Workflows ▲Technology
QCOM · Technology · Positive Qualcomm unveiled new Snapdragon platforms designed to run AI agents directly on consumer devices, positioning agentic AI as its next growth phase.
QCOM · Demand · Positive Qualcomm announced a multi-generation collaboration with Amazon to develop customised silicon for large-scale AI data centers focused on AI inference.
AMZN · Demand · Positive Qualcomm announced a multi-generation collaboration with Amazon to develop customised silicon for large-scale AI data centers focused on AI inference.
Bank of America Raises AMD Price Target to $720 on AI CPU Opportunity
Bank of America reiterated a Buy rating on Advanced Micro Devices and raised its price target to $720 from $620, citing the company's opportunity to capture a larger share of the rapidly expanding server CPU market. The firm expects the server market to grow from approximately $61 billion to $211 billion by 2030, with roughly $180 billion of that total tied to AI workloads, and sees agentic AI boosting demand for CPUs that coordinate workloads and manage data alongside GPUs. AMD's data center revenue rose 107% in the second quarter to $6.72 billion, driven by EPYC server CPUs and Instinct GPUs, while client business revenue climbed 23% year over year to $3.1 billion on Ryzen demand. The stock has gained more than 180% in 2026 and trades at a forward price-to-earnings multiple of 39X, roughly double Nvidia's 19X, though AMD's gross margin of 54% trails Nvidia's roughly 75%. Hedge fund participation increased to 164 funds holding AMD as of the second quarter, up from 134 in the first quarter, while short interest stood at approximately 39.98 million shares as of September 15, about 2.46% of the float.
Chuangyao Technology Earnings Call Responds to SparkLink 2.0 and M&A Questions; First-Half Revenue and Net Profit Both Decline
Chuangyao Technology held its 2026 semi-annual earnings briefing on September 29 via the Shanghai Stock Exchange's SSE Roadshow Center, responding to investor questions on SparkLink 2.0 tape-out and M&A progress. In the first half of 2026, the company achieved operating revenue of 167 million yuan, down 9.02% year-on-year; net profit attributable to the parent was 11.7539 million yuan, plunging 63.16% year-on-year; non-GAAP net profit swung from 26.7156 million yuan in the same period last year to a loss of 2.7683 million yuan, a decline of 110.36%, which the company said was mainly due to a year-on-year increase in government subsidies during the period. Gross margin fell to 28.00% in the first half, down 3.28 percentage points year-on-year, and net margin was only 7.05%, down 10.36 percentage points from the same period last year; net cash flow from operating activities was 45.7005 million yuan, surging 181.95% year-on-year, but the company admitted this was mainly due to an increase in deposit interest actually received. In the second quarter, the company achieved operating revenue of 96.1239 million yuan, up 36.23% quarter-on-quarter, but net profit attributable to the parent was only 1.368 million yuan, down 93.2% year-on-year and down 86.83% quarter-on-quarter. The company confirmed that SparkLink 2.0 has been initiated and is under continuous R&D investment, with the protocol standard adding a network layer and featuring ranging and positioning as well as integrated sensing and communication functions, but regarding when tape-out and shipment will occur and when mass production will begin, it only said to refer to company announcements; EtherCAT slave controller chips have achieved sales to customers in industrial, medical, semiconductor and other fields, and have entered the domestic mainstream embodied robot industry chain, but related revenue still accounts for a relatively low proportion. On M&A progress, the company responded that its investment and M&A layout is advancing steadily, and it will actively yet prudently promote external expansion under strict risk control. As of September 30, the company's stock price closed at 31.99 yuan per share, down more than 50% from the issue price of 66.6 yuan per share, and remains below the issue price.
Intel Data Center Revenue Jumps 59% as CEO Flags Demand Outpacing Supply
Intel's Data Center and AI revenue rose 59% year-over-year to $6.3 billion in the second quarter, outpacing the company's overall topline growth of 25%, as the chipmaker pushes an AI-driven turnaround built on CPUs, ASICs, advanced packaging and its foundry network. CEO Lip-Bu Tan said Intel can currently meet only roughly half of processor demand, a supply constraint the article frames as a potential market-share opportunity rather than a pure setback. Intel holds a 49% stake in chipmaker Altera, which has confidentially filed for a U.S. IPO, and management has issued a mid-20% revenue expansion outlook for 2026. The stock closed at $120.23 on September 30, a 225.83% year-to-date return, giving Intel a market capitalization of $635.55 billion, roughly 11 times trailing revenue of $57.03 billion, with a forward P/E of 63.29x. Hedge fund holdings tracked by Insider Monkey rose to 138 in the second quarter of 2026 from 112 in the first quarter, with BlackRock the largest institutional stakeholder at 426.48 million shares, or 8.46% of outstanding shares, as of June 30.
ON Semiconductor Switches Synaptics Deal to $123-Per-Share All-Cash Offer
ON Semiconductor said it will acquire Synaptics in an all-cash transaction at $123 per share, replacing its original offer of 1.350 ON Semiconductor shares for each Synaptics share held. The revised deal values Synaptics at $5.7 billion, down from the original $7 billion, and represents a nearly 16 percent premium over Synaptics' closing price on Thursday. Synaptics shares surged 14.08 percent on Friday to close at $121.10, extending their winning streak to a fourth straight session and bringing the week's gain to 17.9 percent. Synaptics President and CEO Rahul Patel said the all-cash structure provides value certainty at a meaningful premium, and ON Semiconductor will finance the transaction with cash on hand and committed financing from Morgan Stanley. The deal is expected to close in the middle of next year, subject to approval by Synaptics shareholders, with Synaptics' board joining ON Semiconductor's board. The acquisition followed Synaptics' fourth-quarter results, in which net losses widened by 9,419 percent to $447.4 million from $4.7 million a year earlier, dragged by a $425.3 million non-cash charge tied to a valuation allowance against US deferred tax assets, while revenue rose 8.9 percent to $308 million.
Artificial Intelligence › Edge & On-device AI Silicon Competition
ON · Capital · Neutral ON Semiconductor switched its Synaptics offer to an all-cash $123-per-share deal valued at $5.7 billion, down from $7 billion, financed with cash and committed financing.
SYNA · Capital · Positive Synaptics is being acquired in an all-cash deal at $123 per share, a nearly 16% premium to its prior close, providing value certainty.
Broadcom Guides Q4 Revenue to $34.8 Billion After Q3 Beat
Broadcom reported third-quarter fiscal 2026 non-GAAP earnings of $3.32 per share, up 96.4% year over year and 3.11% above the Zacks Consensus Estimate, with revenues surging 85.5% year over year to $29.59 billion and beating the consensus mark by 0.41%. AI semiconductor revenues jumped 221% year over year to $16.7 billion, representing 56% of total revenues, while Semiconductor Solutions revenues surged 127% year over year to a record $20.84 billion, or 70% of total revenues, and Infrastructure Software revenues increased 29% year over year to $8.75 billion, contributing the remaining 30%. For the fourth quarter of fiscal 2026, Broadcom expects revenues of approximately $34.8 billion, representing 93% year-over-year growth, including about $26.1 billion of semiconductor revenues and roughly $8.7 billion of Infrastructure Software revenues. The company raised its fiscal 2026 AI semiconductor revenue outlook to $58 billion, up 186% year over year and above its prior $56 billion view, and said it has secured supply to support approximately $115 billion of AI semiconductor revenues in fiscal 2027, with fiscal 2028 AI semiconductor revenues seen reaching $230 billion. Broadcom also generated record free cash flow of $13.67 billion, equal to 46% of revenues, paid $3.1 billion in dividends, reduced long-term debt by $5.6 billion, and its board approved a quarterly dividend of 65 cents per share.
AVGO · Capital · Positive Broadcom beat Q3 estimates with EPS up 96.4% and guided Q4 revenue to $34.8B, plus record $13.67B free cash flow and debt reduction.
AVGO · Demand · Positive AI semiconductor revenues jumped 221% to $16.7B and the company raised its fiscal 2026 AI revenue outlook to $58B with supply secured for $115B in fiscal 2027.
TSMC Weighs Additional Chip Plants in Texas With Tens of Billions in Spending
TSMC is considering building a new chip manufacturing plant in Texas, with expectations of adding several tens of billions of dollars in investment under a multi-year plan to expand its chipmaking business in the United States. Sources say the Texas project covers the construction of multiple chip plants, and if the investment goes ahead, each plant would cost at least 20 billion dollars. Previously, TSMC pledged to invest 265 billion dollars in Arizona, where several plants are currently under construction. The Texas investment depends on whether the U.S. government extends a tax credit measure for advanced manufacturing that is set to expire at the end of this year. The plan is still at an early stage and no details have been officially disclosed. TSMC is a key chip supplier to Nvidia and Apple, with revenue from wafer manufacturing in America accounting for more than 75% of its revenue from that business this year.
Semiconductors › Logic, Compute & Connectivity Processors Capital
Artificial Intelligence › AI Compute & Accelerator Silicon ▲Capital
2330.TW · Capital · Positive TSMC is weighing tens of billions in additional Texas chip plant investment, expanding its US manufacturing footprint.
2330.TW · Regulation · Neutral The Texas investment depends on the US government extending an advanced-manufacturing tax credit set to expire this year.
Huawei unveils Mate 90, debuts Kirin 9050 Pro chip with LogicFolding 3D architecture
Huawei Technologies officially announced the launch of its Mate 90 Series flagship smartphones, leaning on its self-developed operating system and a redesigned chip architecture amid US controls on advanced semiconductor exports. Richard Yu, chairman of Huawei's consumer business group, said the company still faces constraints on the supply of advanced processors, as advanced semiconductor capacity within China remains very limited, and its Ascend series of AI chips must also draw on output from the same group of fabs. The top-end Mate 90 Pro Max is powered by the Kirin 9050 Pro chipset, a centralized chip produced using an innovation Huawei calls LogicFolding, which restructures the logic circuit interconnects inside the chip so they stack vertically in three dimensions instead of the traditional two-dimensional horizontal layout, even though it requires a larger amount of silicon wafer per chip. The phones also run on Huawei's fully self-developed HarmonyOS 7, which breaks away from reliance on Google's Android framework and now supports more than 450,000 applications and services. On pricing, the Mate 90 Pro Max with 16 GB of RAM and 512 GB of storage is priced at 9,999 yuan, or about 1,491 US dollars, an increase of 2,000 yuan compared with the same-spec version of the previous generation at launch. The standard Mate 90 and the Mate 90 Pro start at 5,999 yuan and 6,999 yuan respectively. Huawei said the top model's overall performance is up 31% compared with its previous flagship. Yu said rising memory chip prices have added as much as 200 US dollars to the average cost of each device, significantly squeezing Huawei's margins and forcing the company to raise prices as well, though it is still trying to do so gradually.
MediaTek Dimensity 9500 Powers Samsung Galaxy Tab S12 Series
MediaTek announced that its flagship Dimensity 9500 platform powers the new Samsung Galaxy Tab S12 Ultra and Samsung Galaxy Tab S12+, bringing flagship performance and on-device intelligence to Samsung's premium tablet lineup. The chip is built on TSMC's 3nm process and features an All Big Core CPU architecture, the MediaTek NPU 990 and Generative AI Engine 2.0, and a flagship 12-core GPU with hardware-accelerated raytracing and Unreal Engine 5.5+ support. Jerry Yu, Corporate Senior Vice President and Head of Smart Life Business Group at MediaTek, said the platform brings the company's latest advances in flagship performance, on-device AI and power efficiency to the Galaxy Tab S12 Series, while Drew Blackard, SVP of Product Management at Samsung, said the partnership pairs flagship performance with on-device Galaxy AI. The Dimensity 9500-powered Galaxy Tab S12 series will be available at all major retailers starting Monday, October 7.
Artificial Intelligence › Edge & On-device AI Silicon ▲Technology
Artificial Intelligence › AI Compute & Accelerator Silicon ▲Technology
2454.TW · Demand · Positive MediaTek's flagship Dimensity 9500 platform was selected to power Samsung's new Galaxy Tab S12 Ultra and S12+ tablets.
005930.KO · Demand · Positive Samsung's new Galaxy Tab S12 Ultra and S12+ tablets launch with MediaTek's flagship Dimensity 9500, supporting its premium tablet lineup.
2330.TW · Supply · Positive The Dimensity 9500 chip powering the Galaxy Tab S12 series is built on TSMC's 3nm process, implying foundry demand for TSMC.
Qualcomm Renews Apple Patent License Effective April 2027
Qualcomm announced in September 2026 that it had renewed its global patent license agreement with Apple, effective April 1, 2027, a move that significantly reduces near-term uncertainty around its QTL royalties. The renewal comes alongside a multi-year custom chip partnership with Amazon's AWS, which Qualcomm is counting on to underpin its data center goals of US$5.0 billion in fiscal 2027 and US$15.0 billion in fiscal 2029. Qualcomm also presented at the Humanoids Summit in Seoul and is expanding into data center and robotics solutions as it pushes to pair stable licensing income with new AI and cloud computing revenue streams. The company's narrative projects $60.0 billion in revenue and $11.5 billion in earnings by 2029, requiring 10.8% yearly revenue growth and about a $2.2 billion earnings increase from $9.3 billion today. The biggest current risk, according to the analysis, is that diversification spending rises faster than new AI and cloud revenues materialize, while rising in-house chips at major OEMs could still pressure Qualcomm.
Artificial Intelligence › Edge & On-device AI Silicon Competition
QCOM · Demand · Positive Multi-year custom chip partnership with Amazon's AWS underpins Qualcomm's data center revenue goals of $5.0B in fiscal 2027 and $15.0B in fiscal 2029.
QCOM · Regulation · Positive Renewed global patent license agreement with Apple effective April 2027 significantly reduces near-term uncertainty around QTL royalties.
AAPL · Regulation · Neutral Apple renewed its global patent license agreement with Qualcomm effective April 2027, a licensing/legal arrangement that reduces uncertainty but is neutral-to-mixed for Apple.
AMZN · Demand · Positive AWS signed a multi-year custom chip partnership with Qualcomm, adding a supplier relationship for its data center chips.
Qualcomm Renews Global Patent License With Apple From April 2027
Qualcomm said it has renewed its global patent license agreement with Apple, effective April 1, 2027. The updated deal covers Apple's continued access to Qualcomm's cellular technologies under a multi-year global patent license framework, and management indicated the renewed agreement is structured to support ongoing collaboration between the two smartphone heavyweights. The renewal reduces uncertainty around QTL licensing revenue tied to one of Qualcomm's largest customers and supports the view that its patent portfolio remains hard to replicate for premium smartphones. Qualcomm's narrative still hinges on reaching US$40b of non-handset QCT revenue by fiscal 2029 while managing pressure from reduced Apple modem exposure and smartphone cyclicality, so the extension helps visibility on QTL cash flows but does not replace the need to execute on automotive, IoT and the AWS-backed data center program. Investors will watch how Qualcomm breaks out QTL segment performance after 2027, including effective royalty rates and license coverage for 5G and future standards, along with any commentary on duration, scope or renewal terms with Apple in quarterly filings or earnings calls.
QCOM · Regulation · Positive Qualcomm renewed its global patent license agreement with Apple, reducing uncertainty around QTL licensing revenue from a top customer.
AAPL · Regulation · Neutral Apple is the counterparty to the renewed Qualcomm patent license, ensuring continued access to cellular technologies, but the article frames it mainly as a Qualcomm positive.
TCW Fund Flags ON Semiconductor's $7 Billion Synaptics Deal as Shares Fall 20%
ON Semiconductor Corporation announced in late June an all-stock acquisition of Synaptics valued at $7 billion, a deal that sent its shares down 20% and prompted several sell-side analysts to downgrade the stock. TCW Relative Value Mid Cap Fund disclosed the development in its second-quarter 2026 investor letter, noting that ON Semiconductor shares had rallied more than 100% in May and June on a data center story before the Synaptics announcement. Despite the late-June pullback, the stock still gained 52.7% over the quarter and was among the fund's best performers, alongside Centene Corporation and Jabil. The fund said the all-stock structure has caused investors to question the impact of dilution and the motivation behind the deal, making the ON Semiconductor investment story more convoluted. ON Semiconductor closed at $75.95 on September 29, 2026, with a $29.49 billion market capitalization and a roughly 39.45% year-to-date gain, within a 52-week range of $44.56 to $134.92. The TCW Relative Value Mid Cap Fund returned 14.10% in the quarter versus the Index's 13.40%, with portfolio stock selection contributing positively at plus 174 basis points.
Semiconductors › Logic, Compute & Connectivity Processors Capital
ON · Capital · Negative $7B all-stock Synaptics acquisition raised dilution and deal-motivation concerns, sending shares down 20% and prompting analyst downgrades.
SYNA · Capital · Neutral Subject of ON Semiconductor's $7B all-stock acquisition, but the article frames the impact on the acquirer rather than Synaptics itself.
Marvell Technology Shares Jump 4.7% After Citi Reiterates Buy Rating
Marvell Technology shares jumped 4.7% in the afternoon session after Citi reiterated a Buy rating and a $275 price target on the networking chips designer ahead of its October 6 Investor Day. Citi analyst Atif Malik highlighted artificial intelligence data center infrastructure networking opportunities, pointing to raised fiscal 2027 and 2028 revenue targets of approximately $12 billion and $18 billion, as well as potential 2030 earnings of $15 to $20 per share. The note also cited an update on a custom revenue target of more than $10 billion for fiscal 2029 resulting from an agreement with Google, along with approximately $300 million in scale-up optics sales. After the initial pop, the shares cooled down to $261.81, up 3.9% from the previous close. Marvell Technology is up 193% since the beginning of the year, but at $261.81 per share it is still trading 17.3% below its 52-week high of $316.43 from June 2026.
Artificial Intelligence › AI Compute & Accelerator Silicon ▲Demand
MRVL · Capital · Positive Citi reiterated a Buy rating and $275 price target on Marvell ahead of its Investor Day, citing raised revenue and EPS targets.
MRVL · Demand · Positive The note cites a custom revenue target of over $10 billion for fiscal 2029 from an agreement with Google plus ~$300 million in scale-up optics sales.
GOOG · Demand · Positive Marvell's custom revenue target of more than $10 billion for fiscal 2029 stems from an agreement with Google, implying Google is a customer driving that demand.
C · Capital · Neutral Citi is the analyst firm reiterating the Buy rating and $275 target on Marvell, a passing role in the story.
Dosilicon's integrated storage-compute-connect chip R&D project has a planned cycle of about 42 months
Dosilicon released an investor relations activity record announcement stating that the company's integrated storage-compute-connect chip R&D project has a planned cycle of about 42 months, and the project is currently progressing in an orderly manner. Through advanced packaging, the project integrates storage, connectivity, and computing units into a single chip to create a low-power embedded intelligent computing chip featuring local intelligent processing, data privacy and security, low latency, and low power consumption, with research and development undertaken by the company's own team. The announcement said that the chip needs to go through multiple stages including architecture design, tape-out, hardware and software debugging, and customer verification, so there is uncertainty in the R&D rollout. Please refer to the company's disclosed information for specific progress.
688110.CG · Technology · Positive Dosilicon's integrated storage-compute-connect chip R&D project is progressing in an orderly manner, advancing its own chip technology development.
Microchip Technology announced the expansion of its 10BASE-T1S Single Pair Ethernet portfolio with new transceivers and endpoint devices, sending shares up 1.9% to $79.42. The expansion introduces the LAN8679 and LAN8680 physical layer transceivers alongside the LAN8660X/1X/2X family of endpoints with integrated transceivers to support zonal architectures. The new products are engineered to meet AEC-Q100 automotive qualification requirements and ISO 26262 functional safety standards while simplifying wiring and delivering flexible edge connectivity. The stock is up 22.1% since the beginning of the year but remains 22.8% below its 52-week high of $102.92 from May 2026.
MCHP · Technology · Positive Microchip expanded its 10BASE-T1S Single Pair Ethernet portfolio with new automotive-qualified transceivers and endpoint devices.
Qualcomm Unveils Snapdragon 8 Elite Extreme Gen 6 Chips With On-Device AI Agents
Qualcomm showcased phone-based AI agents at its Snapdragon Summit, unveiling the Snapdragon 8 Elite Extreme Gen 6 and Gen 6 chips that combine processing, graphics and sensing so agents can respond to context and act across apps. The wireless and edge computing chipmaker also reported more than 20 partners working with Adreno Neural Fusion, its AI rendering technology for mobile games. Shares rose about 0.3% to $188.04 at around 11:23 a.m. ET Tuesday, after closing at $187.48 on Monday. At that price the stock sits 7.34% above the chart's $175.18 GF Value estimate, meaning investors are already paying for some of the promise. The next proof point is widespread use in shipping phones, followed by evidence that it lifts chip revenue.
Artificial Intelligence › AI Compute & Accelerator Silicon ▲Technology
QCOM · Technology · Positive Qualcomm unveiled Snapdragon 8 Elite Extreme Gen 6 chips with on-device AI agents and Adreno Neural Fusion, with 20+ partners adopting the AI rendering tech.
Broadcom Q2 Revenue Jumps 85.5% to $29.59 Billion as AI Chip Sales Surge
Broadcom reported quarterly revenues of $29.59 billion, up 85.5% year on year, in a mixed quarter that saw inventory levels improve significantly but next-quarter revenue guidance slightly miss analysts' expectations. The company's Q3 AI semiconductor revenue of $16.7 billion grew 221% year-over-year and 54% quarter-over-quarter, and CEO Hock Tan said Q4 AI semiconductor revenue is expected to accelerate to $21.7 billion, up 236% year-over-year. Broadcom delivered the weakest performance against analyst estimates and the weakest guidance update among the nine processors and graphics chips stocks tracked, and its shares are down 4.8% since reporting, trading at $349.44. Across the group, revenues beat consensus estimates by 6.3% and next-quarter revenue guidance came in 6.8% above expectations, with share prices up 5.7% on average since the latest results. Intel posted the strongest quarter, with revenues of $16.13 billion, up 25.4% year on year and 11.7% above expectations, while Qualcomm reported revenues of $9.95 billion, down 4% year on year but 3% above estimates.
AVGO · Capital · Positive Broadcom's Q2 revenue jumped 85.5% to $29.59B with AI semiconductor revenue up 221% year-over-year, though guidance slightly missed and shares fell 4.8%
INTC · Capital · Positive Intel posted the strongest quarter in the group with revenues of $16.13B, up 25.4% year on year and 11.7% above expectations
QCOM · Capital · Neutral Qualcomm reported revenues of $9.95B, down 4% year on year but 3% above estimates
Microchip Expands 10BASE-T1S Single Pair Ethernet Portfolio With New Transceivers and Endpoints
Microchip Technology announced the expansion of its 10BASE-T1S portfolio with the introduction of the LAN8679 and LAN8680 10BASE-T1S PMD transceivers, alongside the LAN8660X/1X/2X family of endpoints with integrated transceivers. The LAN8679 transceiver provides a standardized Open Alliance 3-Pin interface in a compact 8-pin VDFN package, while the LAN8680 adds System Basis Chip features including power management, system wake/sleep control and watchdog supervision. The LAN8660X/1X/2X endpoint family integrates a PMD transceiver into a single-chip Remote Control Protocol solution, allowing compact edge nodes for control systems, lighting and audio without external PHY components. The devices are designed to meet AEC-Q100 and ISO 26262 functional safety standards and build on Microchip's VelocityDRIVE portfolio, targeting automotive zonal architecture, industrial automation, robotics and aerospace and defense. Using 10BASE-T1S multidrop topology, the solutions let multiple nodes, typically eight or more, share a single bus line, reducing wiring weight and system costs. The LAN8679, LAN8680 and LAN8660X/1X/2X are available in limited sampling.
Semiconductors › Analog, Power & Discrete Technology
MCHP · Technology · Positive Microchip expanded its 10BASE-T1S portfolio with new LAN8679/LAN8680 transceivers and LAN8660X/1X/2X endpoints, a product/R&D development.
Micron Earnings, U.S. Payrolls and RBA Rate Decision Headline Q4's First Week
Micron Technology reports its latest quarterly earnings after the bell on Wednesday, the next test for an AI trade that has shown signs of fatigue in recent months. The $1.2-trillion memory chip maker, a key supplier to Nvidia's AI processors, has seen its shares rise over +280% year-to-date, but they fell more than -6% during the quarter while the Philadelphia Semiconductor Index shed an even steeper -14%. Investors will also get September U.S. nonfarm payrolls on Friday, expected to rise by +100K jobs with unemployment at 4.2%, along with the PCE price index on Wednesday and Eurozone inflation data on Friday. The Reserve Bank of Australia is expected to hike its policy rate by a quarter point on Tuesday, taking it to a 15-year high of 4.6%. Futures pricing suggests a roughly 50/50 chance the Federal Reserve will again hike by a quarter point in October.
Artificial Intelligence › AI Compute & Accelerator Silicon Demand
MU · Capital · Neutral Micron's quarterly earnings report is the headline event, framed as a test for the AI trade after shares fell over 6% in the quarter.
AU-10Y.GB · Monetary · Positive RBA is expected to hike its policy rate by a quarter point to a 15-year high of 4.6%, pushing Australian bond yields up.
AMD acquires Fei-Fei Li's World Labs for 8.2 billion dollars, advancing Physical AI
Advanced Micro Devices, or AMD, announced on Monday, September 28, that it has reached an agreement to acquire World Labs, the artificial intelligence startup founded by Dr. Fei-Fei Li, in an all-stock transaction valued at 8.2 billion dollars. The move establishes AMD's foothold in research and development of Physical AI systems capable of understanding and simulating the real world. The acquisition will give AMD access to research in Spatial Intelligence, an emerging field focused on enabling AI systems to build three-dimensional models of environments and simulate physical interactions. World Labs had recently raised 1 billion dollars in a funding round earlier this year, in which AMD also participated. Dr. Fei-Fei Li, a renowned AI researcher known for her pioneering work in computer vision, will join AMD as executive vice president and chief scientist once the acquisition process is complete. She will report to AMD CEO Lisa Su, and the deal is expected to close by the end of 2026.
HKTDC raises Hong Kong 2026 export forecast to 42-47% on AI demand tailwind
The Hong Kong Trade Development Council, or HKTDC, on Monday, September 28, raised its forecast for Hong Kong's merchandise export growth in 2026 to 42-47%, after global demand for artificial intelligence-related technology proved far stronger than expected. Meanwhile, the HKTDC's exporter confidence index for the third quarter of 2026, released the same day, showed the current performance index at 51.8 and the expectation index at 51.3, reflecting that exporter sentiment remains broadly sound even as geopolitical uncertainty persists. Bruce Pang, HKTDC's director of research, said Hong Kong's strong export growth this year has been driven mainly by the rapid acceleration of global demand for AI-related goods and infrastructure, with demand for semiconductors, memory chips, computer components, telecommunications equipment and other advanced electronics expanding much faster than expected, providing significant support to Hong Kong's exports.
Alibaba Nears Chinese Approval to Buy Nvidia RTX PRO 5500 Chips
Alibaba Group Holding is expected to secure Chinese approval to purchase Nvidia RTX PRO 5500 chips under current export rules. Regulators in Beijing are reviewing the RTX PRO 5500 orders as a category of AI hardware seen as compliant with US curbs. The prospective approval would expand Alibaba's pool of Nvidia processors for use in its cloud services and in-house AI projects, supporting the company's RMB 380 billion AI and cloud buildout over three years alongside its Zhenwu V900 chips. The likely green light comes as Alibaba competes with Tencent and Huawei to serve enterprise AI workloads at scale, while analysts warn that relying on both in-house accelerators and imported Nvidia gear could keep capital intensity high and leave profitability under strain if usage lags capacity.
9988.HK · Regulation · Positive Expected to secure Chinese regulatory approval to purchase Nvidia RTX PRO 5500 chips, expanding its AI processor pool.
9988.HK · Capital · Negative Analysts warn dual reliance on in-house and imported Nvidia gear keeps capital intensity high and pressures profitability if usage lags capacity.
NVDA · Demand · Positive Alibaba expected to secure Chinese approval to buy Nvidia RTX PRO 5500 chips, expanding orders for Nvidia AI hardware.
0700.HK · Competition · Neutral Named as a competitor to Alibaba in serving enterprise AI workloads at scale.
Huawei · Competition · Neutral Named as a competitor to Alibaba in serving enterprise AI workloads at scale.
NVIDIA Boosts Buyback Authorization by $1 Billion, Fulfilling Jim Cramer's Wish
NVIDIA Corporation increased its share repurchase authorization by $1 billion on Monday, a move CNBC host Jim Cramer said he had pushed for throughout 2026. Cramer argued the buyback could change the stock's trajectory if the company is active in the market every day, comparing it to the repurchases Apple executed under former CFO Luca Maestri. The announcement comes as NVIDIA guides 70% fiscal year 2028 revenue growth after growing revenue 106% annually in the second quarter, though higher memory prices and tight semiconductor manufacturing capacity remain risks. The company guided third quarter gross margin of 75%, a full point lower than Q2's 74%, and said fiscal Q4 gross margin could range between 71% and 72% before recovering. NVIDIA's long term debt jumped to $32.3 billion in the second quarter from $7.5 billion a year earlier, and the stock trades at a forward P/E ratio of 24.9, below Intel's 63.29 and AMD's 39.53.
QNX and neueHCT Win German Automaker Program for Smart Camera Solution
A major German automaker has selected neueHCT's QNX-powered HCT Luna High-Performance Smart Camera Solution for a global passenger vehicle platform launching in China, with production set to begin in 2027 and projected volumes of several million units within the first three years. The program, announced by QNX, a division of BlackBerry Limited, and neueHCT, the intelligent driving company jointly established by Horizon Robotics and AUMOVIO, extends across Australia, New Zealand, Japan and South Korea, establishing a foundation for further global expansion. The win marks the first major global program for the HCT Luna High-Performance Smart Camera Solution powered by QNX, moving the product from individual vehicle applications to large-scale deployment across multiple platforms and models. Built on Horizon Robotics' Journey 6B automotive compute platform and powered by QNX OS for Safety 8.0, the solution will be deployed in vehicles sold in China and multiple international markets. Simon Zhang, COO of neueHCT, said securing the win demonstrates that the company's solution has met the technical standards required by global OEMs, while John Wall, President at QNX, said the program shows the role QNX OS for Safety 8.0 can play in scaling advanced driver assistance capabilities across platforms, models and markets.
Robotics & Physical AI › Autonomous Vehicles & Robotaxi Technology
neueHCT · Demand · Positive neueHCT's HCT Luna smart camera solution won a major German automaker program with production in 2027 and millions of units.
BB · Demand · Positive QNX OS for Safety 8.0 selected for a major German automaker's global smart camera program with millions of units projected.
9660.HK · Demand · Positive The HCT Luna solution is built on Horizon Robotics' Journey 6B automotive compute platform, gaining a multi-million-unit automaker program.
Arista Networks Touts AI Networking Moat as Revenue Tops $3 Billion
Arista Networks is positioning its EOS software and AI networking expertise as a durable competitive moat as AI data centers scale up. The company's second-quarter revenue jumped 37.7% year over year to $3.04 billion, its first quarter above $3 billion, and management raised its 2026 revenue-growth target to roughly 40%, or about $12.6 billion. Management says more than 100 customers now use its Etherlink AI fabric products, up from only four or five a couple of years ago, and the technology supports multiple accelerator architectures including AMD's accelerators and Google's TPUs. The stock is up 54.08% year-to-date, and shares trade at roughly 39.2x forward earnings, though gross margin fell year over year in the second quarter amid tight supplies and rising component costs. Competitors including Cisco and Broadcom are also chasing AI infrastructure spending, while hedge funds tracked by Insider Monkey increased their holdings from 85 to 91 between the end of Q1 and the end of Q2 2026, with position value rising from about $5.21 billion to $6.77 billion.
Broadcom Lifts Fiscal 2027 AI-Chip Revenue Forecast to About $115 Billion
Broadcom raised its fiscal 2027 AI-chip revenue forecast to about $115 billion, a target that sets a high bar for converting demand into shipped chips. Shares slipped roughly 0.7% to $350.10 around 11:51 a.m. ET on September 28. The latest quarterly results give that target some weight: AI-chip revenue hit $16.7 billion, up 54% from the previous quarter, while Broadcom expects $21.7 billion in the fourth quarter. Free cash flow reached $13.7 billion, or 46% of total revenue. Even at the fourth-quarter forecast's annual pace of $86.8 billion, Broadcom would need another substantial step up to reach $115 billion next fiscal year. The $350.10 share price sits 15.48% below GuruFocus' $414.21 GF Value estimate, and the next test is whether customer demand becomes shipped chips and cash.
Artificial Intelligence › AI Compute & Accelerator Silicon ▲Demand
AVGO · Capital · Positive Quarterly results showed AI-chip revenue of $16.7 billion and free cash flow of $13.7 billion (46% of revenue), with a $21.7 billion Q4 guide.
AVGO · Demand · Positive Broadcom raised its fiscal 2027 AI-chip revenue forecast to about $115 billion on strong AI-chip demand, with AI revenue up 54% sequentially.
Marvell Begins Custom AI Chip Shipments to Google, Plans Higher AI Revenue Outlook
Marvell Technology has begun shipping custom Compute Express Link chips to Alphabet's Google and plans to lift its AI revenue outlook after deepening its custom silicon partnership with the company. Management expects large-scale custom AI inference silicon programs with Google to ramp through 2028, pointing to a multi-year engagement. The expanded Google engagement reinforces Marvell's bet on hyperscaler design wins and multi-year programs, and ties into its expanded SiGe and IC substrate agreements aimed at securing the manufacturing headroom those contracts require. The unresolved piece is concentration risk, as a bigger Google pipeline leans harder into dependence on a few large, lumpy projects and tight supply chains compared with more diversified rivals such as Broadcom or Intel.
Arm Shares Fall 7.6% as It Outlines Role in Nvidia Agent Security Platform
Arm Holdings outlined its role in Nvidia's new agent security platform as its shares fell 7.6% to $286.79 around 9:57 a.m. ET Monday. Nvidia's OpenShell can run on Arm's Vera CPU, and the approach can extend to platforms such as AWS Graviton, Google Axion and Microsoft Cobalt. Arm based infrastructure processors can also monitor an agent's activity and enforce security rules separately from the processor running it, putting Arm technology on both sides of the setup. The announcement described a possible route to more chip demand rather than a new customer order, and the valuation leaves little room for a story that stays theoretical, with a GF Value of $194.76 putting Monday's $286.79 share price 47.25% above that estimate. More agent deployments could mean more Arm powered processors, but investors still need evidence that those deployments are turning into licensing and royalty growth.
Cybersecurity & Digital Trust › AI Security & Agent Guardrails Technology
Artificial Intelligence › EDA & Semiconductor IP Technology
Semiconductors › EDA & Semiconductor IP Technology
Artificial Intelligence › Agentic AI & Autonomous Workflows Technology
ARM · Technology · Neutral Arm outlined its role in Nvidia's OpenShell agent security platform, a possible route to more chip demand but no new customer order or licensing/royalty evidence.
NVDA · Technology · Neutral Nvidia's OpenShell agent security platform is the subject, but the article focuses on Arm's role and gives no Nvidia-specific financial or demand impact.
Alibaba Unveils 10-Trillion-Parameter AI Model and Zhenwu V900 Chip
Alibaba Group laid out plans for a five trillion to ten trillion parameter model and a new Zhenwu V900 chip at its September 22 conference. The V900 delivers three times the performance of its predecessor and can power clusters of up to 500,000 chips, while CEO Eddie Wu set a target for Alibaba Cloud to operate more than 20 gigawatts of data center capacity by 2032. Reuters reported that the announcement lifted Alibaba shares on September 22, but the U.S. shares slipped about 0.1% to $109.63 at 11.10am ET on September 28, showing that early excitement has since cooled. The investment case rests on turning that computing scale into paying cloud demand, since building its own models and chips could give Alibaba more control over costs but a 20 gigawatt buildout raises the stakes for spending and utilization. The shares sit $11.86 below the $121.49 GF Value estimate, a 9.76% gap that will matter only if the expansion earns an adequate return.
9988.HK · Technology · Positive Alibaba unveiled a 5-10 trillion parameter AI model and the Zhenwu V900 chip, which delivers 3x its predecessor's performance.
9988.HK · Capital · Neutral The 20GW data-center buildout raises spending/utilization stakes, and shares sit 9.76% below the GF Value estimate, hinging on adequate returns.
SiMa.ai, a startup developing chips and software that let robots, drones, cameras, and other devices run AI directly on the device, has raised a $150 million Series C at a $1.45 billion valuation. The round was co-led by Fidelity Management & Research Company and Amplify, with participation from Alter Venture Partners, Dell Technologies Capital, and StepStone Group. Founded in 2018 by Krishna Rangasayee, previously COO of chipmaker Groq, the company provides energy-efficient chips that eliminate the need to send data back and forth to the cloud. SiMa.ai hopes its low-latency performance and more affordable chips, compared to Nvidia's GPUs, will help it capture the growing market for physical AI devices, including humanoid robots. The new round brings SiMa.ai's total capital raised to over $500 million, and the startup was previously valued at $960 million after raising an $85 million Series B in July 2025, according to PitchBook.
Nvidia Adds $150 Billion to Buyback, Lifting Total Authorization to $235 Billion
Nvidia shares rose after its board authorized an additional $150 billion under the company's existing share-repurchase program, increasing the total remaining amount authorized to $235 billion, with the AI chip leader expecting to complete the program through fiscal 2028. In the mining sector, Australia's Northern Star Resources Ltd. rejected a takeover approach from South African rival Gold Fields Ltd. that could have created the second-largest gold miner, saying the A$38.7 billion ($27.1 billion) cash-and-shares offer undervalued its business. Gold Fields shares fell as much as 16% on the proposal, while precious-metal miners also slid as gold and silver dropped, with Barrick Gold down 4% and Freeport-McMoRan down about 3.5%. Roblox was cut to underperform from hold at Jefferies, which said the stock's 30% rally since the gaming company's second-quarter results in July reflects an overly optimistic view of bookings for the next 12 months; the shares fell 5% and are down 43% so far this year. Nvidia also rolled out a new double-layered AI security system that it says would have prevented the recent high-profile breach of Hugging Face by OpenAI's models, and China may allow Alibaba and ByteDance to buy Nvidia's new RTX Pro 5500 chips.
GFI · Capital · Negative Gold Fields shares fell as much as 16% after Northern Star rejected its A$38.7 billion takeover offer.
NVDA · Capital · Positive Nvidia's board authorized an additional $150 billion buyback, lifting total authorization to $235 billion.
NVDA · Technology · Positive Nvidia rolled out a new double-layered AI security system it says would have prevented the Hugging Face breach.
RBLX · Capital · Negative Jefferies cut Roblox to underperform, saying its 30% rally reflects overly optimistic bookings expectations.
Northern Star Resources Limited · Capital · Neutral Northern Star rejected Gold Fields' A$38.7 billion takeover offer, saying it undervalued the business.
B · Monetary · Negative Barrick Gold fell 4% as gold and silver prices dropped, pressuring precious-metal miners.
Amphenol Q2 2026 Orders Jump 94% as AI Datacom Drives Growth
Amphenol reported second-quarter 2026 orders of $10.7 billion, up 94% year over year and 14% sequentially, for a book-to-bill ratio of 1.23:1, with every end market posting a positive book-to-bill ratio. The adjusted operating margin expanded 420 basis points year over year and 250 basis points sequentially to 29.8%, and even excluding tariff recoveries the margin approached 29%. CommScope contributed more than $1.2 billion in second-quarter sales at a margin above 20%, and Amphenol raised its expected 2026 CommScope revenues to $4.6 billion from $4.1 billion while doubling expected adjusted EPS accretion to 30 cents from 15 cents. IT datacom represented 43% of second-quarter sales, with revenues rising 89% year over year and 63% organically, and adjusted EPS rose 67% year over year to $1.35 as operating and free cash flows reached $1.6 billion and $1.2 billion. Amphenol faces competition from Marvell Technology, which lifted its fiscal 2028 revenue outlook to $18 billion, and TE Connectivity, which reported record orders of $5.7 billion, up 27% year over year.
APH · Capital · Positive Adjusted operating margin expanded 420bp to 29.8% and adjusted EPS rose 67% YoY to $1.35, with raised CommScope revenue and EPS accretion guidance.
APH · Demand · Positive Q2 2026 orders jumped 94% YoY to $10.7B with 1.23:1 book-to-bill and IT datacom revenue up 89%, signaling strong end-customer demand.
MRVL · Competition · Neutral Named only as a competitor, with its fiscal 2028 revenue outlook lifted to $18B cited as context for Amphenol's competitive landscape.
TEL · Competition · Neutral Named only as a competitor reporting record orders of $5.7B, up 27% YoY, cited as context for Amphenol's competitive environment.
Piper Sandler Projects Google TPU Sales to Reach $104 Billion by 2028
Piper Sandler analyst Thomas Champion is projecting that Google's newly disclosed external chip business will generate $104 billion in TPU hardware revenue by 2028, a forecast the firm says the broader Street has yet to price in. Champion published the note after Google disclosed external hardware sales for the first time in its second-quarter 2026 earnings report, a milestone that triggered a formal reassessment of the revenue opportunity. The firm's model has TPU sales climbing from approximately $8 billion in 2026, or roughly 240,000 chips, to $51 billion in 2027, or approximately 2.4 million chips, and then to $104 billion in 2028, anchored to a facilities and power-capacity build-out of 3 gigawatts of run-rate capacity by the end of 2027 and 6GW by the end of 2028. Piper Sandler raised its price target on Alphabet to $400 from $395 and reiterated its Overweight rating, with its 2027 and 2028 Cloud revenue estimates now 15% and 17% above consensus and its Cloud EBIT estimates 18-19% above consensus. Champion also flagged that even these estimates may prove conservative, noting that Broadcom management recently spoke to 5GW of Anthropic demand in 2027 and 10GW in 2028, which Piper Sandler estimates as $105 billion and $210 billion in revenue if that capacity is achieved.
TSMC 2nm Output Seen Reaching 120,000 Wafers Monthly by End-2026
Taiwan Semiconductor Manufacturing is expanding 2-nanometer production faster than previously expected as major customers increase their commitments to the new technology. The company could reach monthly 2nm output of about 120,000 wafers by the end of 2026, compared with earlier estimates of 90,000 to 100,000, according to Taiwan's EDN. Apple, Nvidia, Advanced Micro Devices, Qualcomm and MediaTek are reportedly lifting 2nm orders by 10% to 20%, giving TSMC additional visibility as customers prepare next-generation processors and AI accelerators. In the second quarter, 2nm represented 3% of TSMC's wafer revenue, while technologies at 7nm and below accounted for 77%, and high-performance computing made up 66% of revenue, up from 60% a year earlier. TSMC has projected 2026 2nm capacity growth as it ramps multiple facilities in Taiwan.
Artificial Intelligence › Edge & On-device AI Silicon ▲Supply
2330.TW · Demand · Positive TSMC's 2nm output is seen reaching 120,000 wafers monthly by end-2026 as major customers raise orders 10-20%.
2454.TW · Demand · Positive MediaTek is reportedly lifting its 2nm orders by 10-20%, increasing demand for TSMC's advanced node as it prepares next-generation processors.
AAPL · Demand · Positive Apple is reportedly lifting its 2nm chip orders by 10-20%, signaling stronger demand for TSMC's advanced node.
AMD · Demand · Positive AMD is reportedly increasing 2nm orders by 10-20% as it prepares next-generation processors.
NVDA · Demand · Positive Nvidia is reportedly raising 2nm orders by 10-20% for next-generation AI accelerators.
QCOM · Demand · Positive Qualcomm is reportedly lifting 2nm orders by 10-20% for upcoming processors.
Volato's Alignment Engine Signs $1.2 Billion AI Infrastructure Orders
Volato Group's Alignment Engine subsidiary has entered into two four-year, take-or-pay orders with a large AI customer representing $1.17 billion in aggregate contract value, marking Alignment Engine's first major customer commitment for its AI infrastructure platform. The orders cover phased deployment of dedicated AMD GPU infrastructure at Alignment Engine's Ohio AI infrastructure campus, with the initial deployment using AMD MI355X GPUs and a contractual start date of December 31, 2026, and a second deployment using AMD MI455X GPUs starting June 30, 2027. The customer has committed to approximately $133 million in contractual prepayments, including approximately $40 million due following execution of the initial order, with the remaining prepayments scheduled in stages based on specified dates and equipment shipment milestones. The deployments will operate as dedicated, single-tenant GPU clusters, and the remaining contract value is payable over the four-year terms on a take-or-pay basis. Volato Chief Executive Officer Chris Ensey called the orders a transformational commercial milestone, and Chief Financial Officer Mark Heinen said they demonstrate a model of building infrastructure against long-term contracted demand and meaningful customer cash commitments.
SOAR · Demand · Positive Volato's Alignment Engine subsidiary signed two four-year take-or-pay orders worth $1.17B with a large AI customer, its first major AI infrastructure commitment.
Alignment Engine · Demand · Positive Alignment Engine secured $1.17B in take-or-pay AI infrastructure orders plus ~$133M in customer prepayments for its Ohio GPU campus.
AMD · Demand · Positive Volato's $1.17B AI infrastructure orders will deploy dedicated AMD MI355X and MI455X GPU clusters, driving demand for AMD's AI chips.
Nvidia releases safety tools to prevent AI agents from going rogue
Nvidia on the 28th released a suite of safety software for artificial intelligence agents. One of the tools, OpenShell, uses hardware features in Nvidia-made CPUs to keep agents confined within isolated environments, and the company is working with Arm Holdings and Intel to make it run on their CPUs as well. Another system, Sentry, combines a different Nvidia semiconductor with OpenShell to block out-of-control agents from trying to escape containers on the CPU. Justin Boitano, who heads enterprise computing, said the tools could have prevented the attack on Hugging Face made public this summer, and that the company will roll them out with dozens of partners, including Anthropic. Hugging Face was hit by an attack from a swarm of rogue OpenAI agents, and a few months later Nvidia acquired it for 13 billion dollars.
JPMorgan Expects AI-Linked Stocks to Rebound, Stays Bullish on Semiconductors
JPMorgan analysts said on the 28th that the recent correction in global AI-related trading has improved investors' positioning and made valuations more attractive, which could encourage investors to re-enter the market, particularly in semiconductor stocks. In a report, they noted that tech may not return to past peak levels, but the underlying scenario remains positive and there are many investment opportunities within the AI-related sector. They said investor positioning in the AI space has been cleaned up and valuations have fallen sharply across most areas, while capital expenditure is likely to remain solid despite recent slowdown concerns. On semiconductor stocks, they maintained a bullish outlook citing strong fundamentals, price increases through 2027, and supply-demand tightness expected to persist until 2028, while keeping a cautious stance on software stocks, assessing that AI growth continues to leave the sector's long-term outlook uncertain.
JPM · Capital · Neutral JPMorgan analysts issued a bullish report on AI/semiconductor stocks; the firm is the source of the call, not a subject of fundamental impact.
Singapore manufacturing output jumps 15.4% in August on AI and chip demand
The Singapore Economic Development Board (EDB) said on September 28 that Singapore's manufacturing output in August surged 15.4% year on year, driven by broad-based growth across most industry clusters. On a seasonally adjusted basis, manufacturing output in August fell 0.5% month on month. The report said output rose year on year in nearly every sector except chemicals. The precision engineering cluster soared 33.9%, led mainly by the production of semiconductor machinery and equipment. The electronics cluster rose 28.5% on the strength of server and related products, semiconductors, and data storage equipment, amid strong demand related to artificial intelligence. The transport engineering cluster expanded 9.5%, led by land vehicles and aircraft. The general manufacturing industries cluster grew 1.5%, and biomedical manufacturing edged up 0.4%. Chemicals was the only cluster to decline, with output falling 12.7%, weighed down by the petroleum and petrochemical segments.
