TrendForce's latest investigations find that the global foundry industry recorded 1Q25 revenue of US$36.4 billion—a 5.4% QoQ decline. The downturn was softened by last-minute rush orders from clients ahead of the U.S. reciprocal tariff exemption deadline, as well as continued momentum from China's 2024 consumer subsidy program. These factors help offset the typical seasonal slump.

Looking ahead to Q2, the effects of tariff-driven early procurement are expected to fade and lead to a general slowdown. However, continued demand from China's subsidy program, along with pre-launch inventory builds for new smartphone models and stable AI HPC demand, are expected to support capacity utilization and drive a revenue rebound for the top 10 foundries.

According to the Financial Times, new restrictions—affecting the supply of Electronic Design Automation (EDA) software suites to Chinese companies—will cause major disruption within the domestic chip design industry. The US government's latest trade amendments are expected to impact Xiaomi and its freshly launched flagship XRING mobile chip family; the first iteration is a compelling first-party effort. Unlike many Chinese tech firms, the popular smartphone specialist can access pretty advanced TSMC node processes. Xiaomi's CEO—Lei Jun—announced his team's 3 nm design during pre-launch preview events.

Days later, closer analysis indicated a selection of TSMC's "N3E" node process. Digital Chat Station—a noted smartphone industry expert—summarized an uncertain future: "under this ban (of EDA tools), XRING chips will not be breaking through a 2 nm barrier, and can only revolve around the (current 'N3E') 3 nm node for a long time. XRING O1 will also be the only time in recent years that it can be on par with current-gen (proprietary) Apple, Qualcomm, and MediaTek mobile chipsets." Crucially, EDA software plays an important role in creating Gate All Around Field Effect Transistor (GAAFET) structures. TSMC's upcoming 2 nm node process is a GAA product. Tom's Hardware believes that several big Chinese tech players, including Huawei, are in the process of developing in-house EDA tools. Not long after unveiling their XRING flagship, Xiaomi outlined an extended Qualcomm chip deal.

Around mid-April, an unusual custom GeForce RTX 5090D design received international news coverage. Under normal circumstances, NVIDIA's board partners have not equipped modern gaming graphics cards with blower-style cooling solutions—typically, this type of treatment is reserved for workstation-grade products (operating at lower TDPs). Team Green's GB202 "Blackwell" GPU is in high demand, due to its AI-crunching prowess—even in slightly nerfed form. Smaller Chinese AI firms and well-heeled hobbyists seemed to be snapping up sanction-adjusted flagship gaming GPUs that are/were coupled with very unofficial blower-type coolers. A month and a half later, Olrak29 has placed a spotlight on a curious batch of bootleg-esque "RTX 5090 32G D7 Turbo" models. As implied by this identifier, these offerings seem to leverage the unpasteurized GeForce RTX 5090 (non-D) GPU.

Photo evidence was scraped from Goofish; a second-hand trading platform owned and operated by Alibaba. As of last month, industry whispers suggested another downgrade of the Chinese market-exclusive GeForce RTX 5090D design—following the complete cutoff of GB202 die shipments into the region. Given current global tensions and export restrictions, NVIDIA and Leadtek's Blackwell PRO W card series faces an uncertain future in China. The Goofish photo uploads indicate an impressive volume of unbranded "RTX 5090 32G D7 Turbo" stock. Stacks of boxes in the background are labelled with "NVIDIA RTX 4090 24G AIB BLOWER" or "NVIDIA RTX 4070 12G AIB BLOWER" stickers—perhaps as diversion tactics. According to VideoCardz and Tom's Hardware, Team Green is not expected to play detective—the North American corporation will probably not provide in-depth comments about elaborate "GPU smuggling" channels.

Greetings soon-to-be-awakened, today, just about 72 hours before the floodgates open, we can finally share with you that rentable private servers will be available from head start launch on June 5th! We've previously communicated that private servers are for post-launch, but we're happy to share that progress has been faster than expected. We do, however, want to manage expectations about how private servers work in Dune: Awakening. As you know, this is not your typical survival game.

Why private servers work differently in Dune: Awakening
Dune: Awakening is powered by a unique server and world structure, something we went in-depth on in a recent blog post. In short: each server belongs to a World consisting of several other servers, and each of those share the same social hubs and Deep Desert. This allows us to retain a neighborhood-like feel to the Hagga Basin and provide persistent, freeform building, and other server-demanding mechanics you typically see in survival games. We combine this with the large-scale multiplayer mechanics you would expect to find in MMOs where hundreds of players meet each other in social hubs and the Deep Desert to engage in social activities, trade, conflict, and more.

Recent trends in the NAS market and general tech market has shown a propensity to lock down their devices to prevent you from using the device you paid for as you see fit. We at ASUSTOR do not believe in locking down your NAS device. As such, all ASUSTOR NAS devices come in an unlocked state. Today, ASUSTOR is announcing its intention to continue to maintain its devices in an unlocked state for maximum freedom.

Owning Your Products!
In recent years, companies across the tech industry have announced tighter and tighter vendor lock-in requirements that are largely arbitrary in nature. This has been in the form of putting software locks that disallow the use of unauthorized software or peripherals on devices you purchase with your hard earned money. Examples of this include companies forcing use of in-house branded hardware like hard drives in a NAS, locking out third party operating systems and apps not approved by the vendor This is part of a general trend towards disposable technology, vendor lock-in, planned obsolescence, and anti-consumer practices all at an unnecessary expense to you, the consumer. At ASUSTOR, we believe that you own the hardware you purchase, and as such, ASUSTOR NAS devices come in an unlocked state. ASUSTOR NAS devices do not place arbitrary restrictions on your use of the device, while using a non-approved drive on NAS devices with vendor lock-in policies can cause error messages or loss of features like building a RAID pool or knowledge of drive health and more.

Last week Synology put out a press release on its European website, informing its customers that the Plus Series—which starts with simple two bay devices—will only support its own brand of hard drives, if you want to be able to take advantage of basic NAS features like creating a storage pool or do a lifespan analysis. More advanced features like volume-wide deduplication will also not be available. Synology has already done this on its XS Plus and rack mounted NAS appliances—which in itself wouldn't be acceptable to many considering getting a NAS—but at least these are pricier, high-end devices. Now the company has moved down to the $300 price bracket, which just doesn't sit right.

Admittedly this will only affect new customers, as the company will start to roll out the limitation with its 2025 SKUs and all other models will continue to work as is. The problem here is that Synology isn't a hard drive manufacturer and according to Arstechnica, the company is simply re-branding Toshiba hard drives and charges an extra $50 or so for their basic tier of drives, with higher-end tiers having a much higher premium. Synology obviously claims that this is for the benefit of their customers and you can find the full statement below. However, we doubt many customers will be happy to pay extra for something that was working just fine, until Synology figured out it could charge its customers extra for it. For now, there has been no indication that this will happen outside of the European market, but it's likely to be rolled out globally, as something like this is rarely limited to one market.

Last week, industry reports pointed to evidence of NVIDIA H20 AI GPU shortages in China—supply chain insiders expressed frustration about limited availability, and alleged price hikes. Days later, local media outlets have disclosed staggering sales figures. Two unnamed sources opine that the likes of Tencent, Alibaba and ByteDance have spent roughly US$16 billion on H20 purchases, across the first three months of 2025. Back in February, Reuters noted an extraordinary surge in orders for: "(Team Green's) H20 model—the most advanced AI processor legally available in China under U.S. export controls—driven by booming demand for Chinese startup DeepSeek's low-cost AI models." The unprecedented rush—to secure precious AI-crunching hardware—was likely motivated by whispers of elevated restrictions; coming from across the Pacific Ocean.

Curiously, local government bodies have allegedly "advised" a stoppage of H20 orders—a recent Financial Times article suggested that this message was directed at the nation's largest AI players (mentioned above). A few industry moles believe that NVIDIA's engineering department is working on another Chinese market exclusive AI chip, although it is not clear whether a new entrant will be designed to conform to recently introduced "not very strict" environmental regulations. Anonymous sources have made noise about an upgraded H20 variant; sporting fancy HBM3E modules.

Yangtze Memory Technologies Co. has begun shipping its fifth-generation 3D NAND memory chips, featuring 294 total layers and 232 active layers, according to an analysis from TechInsights. The new chips achieve bit density near 20 Gb per square millimeter (19.8 Gb/mm²), comparable to current offerings from SK Hynix and approaching Kioxia/Western Digital's latest products. YMTC's design uses hybrid bonding to connect the memory array with logic components, suggesting that Chinese memory makers are up to speed on the packaging standards that their non-sanctioned competitors use. The company has opted for a quiet release without formal announcements, a departure from typical industry practice. This low-key approach may be related to ongoing US trade restrictions on Chinese semiconductor companies.

While not setting records for active layers, SK Hynix's upcoming 321-layer chip leads in that metric, and YMTC's achievement in total layer count demonstrates continued technical progress. The chip uses string stacking technology, though the specific configuration of the layer arrays remains unclear. Other specifications show the new chip employs YMTC's Xtacking 4.0 architecture and triple-level cell (TLC) design. This matches major competitors' architecture type, though detailed performance metrics such as interface speeds have not been disclosed. The 5th generation NAND focuses on getting density on the right track. However, we expect YMTC to continue development at the same speed, match SK Hynix with 321-layer chips, and surpass it with Xtacking 5.0 in the near future.

C.C. Wei, TSMC CEO and Chairman, has shared his latest views regarding his company's North American manufacturing center—Reuters cornered him for comment during a mid-week appearance at a National Taiwan University-held event. The Taiwanese government has recently lowered its "silicon shield"—following much (reported) deliberation over "legal restrictions on transferring leading-edge process technology overseas." This relaxation of rules has TSMC considering a new set of investments for operations outside of Taiwan—with an expansion into advanced node process manufacturing. Currently, 2 nm (N2) is a home turf-speciality—industry experts estimate an expenditure of $28-30 (USD) billion to bring this production technology over to the States. TSMC's CEO has described additional challenges—on top of (and impacting) finances—local bureaucracy is a big one.

Wei stated: "every step requires a permit, and after the permit is approved, it takes at least twice as long as in Taiwan." According Reuters, he reckons that it would be difficult for their North American sites to access the latest technologies ahead of teams in Taiwan. He detailed his company's recruitment of several experts—tasked with talking to local government; about regulatory issues. This was not a cheap undertaking: "we ended up establishing 18,000 rules, which cost us $35 million." TSMC's Arizona production hub will (eventually) consist of three large factories—despite long-term teething problems, Fab 21 is reported to be churning out the first wave of "Made in America" product for a very important client: Apple. Wei expressed positives views when asked about the USA site's prospects—during an earnings conference (Jan 16)—he believes that it will eventually produce the "same quality of chips as in Taiwan," through a "smooth ramp-up process."

AI agents are poised to transform productivity for the world's billion knowledge workers with "knowledge robots" that can accomplish a variety of tasks. To develop AI agents, enterprises need to address critical concerns like trust, safety, security and compliance. New NVIDIA NIM microservices for AI guardrails—part of the NVIDIA NeMo Guardrails collection of software tools—are portable, optimized inference microservices that help companies improve the safety, precision and scalability of their generative AI applications.

Central to the orchestration of the microservices is NeMo Guardrails, part of the NVIDIA NeMo platform for curating, customizing and guardrailing AI. NeMo Guardrails helps developers integrate and manage AI guardrails in large language model (LLM) applications. Industry leaders Amdocs, Cerence AI and Lowe's are among those using NeMo Guardrails to safeguard AI applications. Developers can use the NIM microservices to build more secure, trustworthy AI agents that provide safe, appropriate responses within context-specific guidelines and are bolstered against jailbreak attempts. Deployed in customer service across industries like automotive, finance, healthcare, manufacturing and retail, the agents can boost customer satisfaction and trust.

The Biden administration is rolling out a third major export control package aimed at China's semiconductor industry, as per a report from Reuters. Estimated to affect 140 companies, including China's chip equipment maker Naura Technology Group, Piotek, and Huawei Technologies, the effort aims to limit China's access to advanced chip making technology. In particular, technology that could be used in military products and artificial intelligence. Important sanctions include export controls to specific chip equipment manufacturers, blocking the delivery of high-performance memory chips and the addition of several semiconductor investment companies to the list of export-restricted entities.

The package expands U.S. regulatory authority through foreign direct product rules. It regulates chip manufacturing equipment manufactured around the world with U.S. technology, Japan and the Netherlands are exempt. However, the rules could have an impact on manufacturers outside U.S. such as those based in Israel, Malaysia, Singapore, South Korea, Taiwan and non-U.S. firms (i.e. ASML) due to the complexity of the technological and supply chain. This continues the Biden administration's strategy to limit China's semiconductor capabilities and comes just weeks before the Trump administration made changes. When asked about US new restrictions Chinese Foreign Ministry spokesperson Lin Jian said at a regular press conference on Monday that such behavior undermines the international economic and trade system, and disrupts global supply chains. China will take measures to protect companies' rights and interests.

TSMC and Samsung are reportedly in talks with the United Arab Emirates (UAE) to establish chip factories in the Gulf nation. As reported by the Wall Street Journal, this "desert dream" aligns with the UAE's ambitious plans to diversify its economy beyond oil and become a key player in the AI sector by building chips for AI domestically. The UAE and neighboring Saudi Arabia plan to leverage their oil wealth to invest in cutting-edge manufacturing, with AI emerging as a primary focus due to its high computational demands. Successful implementation of chip factories could significantly boost the region's AI capabilities and impact the global semiconductor supply chain. However, the project faces substantial challenges. Previous attempts to establish semiconductor manufacturing in the Gulf, such as the GlobalFoundries initiative over a decade ago, have yet to progress beyond initial planning.

The current proposal faces even greater obstacles, with estimated costs exceeding $100 billion for a state-of-the-art facility and necessary infrastructure. Geopolitical concerns add another layer of complexity. Recent US export restrictions of certain chips to the Gulf region may complicate the transfer of advanced manufacturing processes to the UAE. Despite these hurdles, the potential benefits are significant. For the UAE, success would represent a major step towards economic diversification and technological leadership. TSMC and Samsung could gain a strategic presence in a region eager for technological advancement. TSMC noted that the company focuses on current expansion projects in the US, Japan, and Germany, while Samsung declined to comment.

According to sources familiar with the matter and reported by Bloomberg, the Biden administration is considering imposing further export controls to limit China's ability to acquire advanced semiconductor technologies crucial for developing AI systems. Gate-all-around (GAA) transistor technology and high-bandwidth memory (HBM) chips are at the center of the proposed restrictions. These cutting-edge components play a pivotal role in creating powerful AI accelerators. GAA transistors, a key feature in next-generation chips, promise substantial improvements in power efficiency and processing speeds. Meanwhile, HBM chips enable high-speed data transfer between a processor and memory. While existing sanctions prevent American firms from supplying Chinese companies with equipment for manufacturing leading-edge chips, concerns persist that China could still attain advanced capabilities through other means.

For instance, China's leading chipmaker, SMIC, could potentially integrate GAA transistors into its existing 7 nm process node, markedly enhancing performance. Access to HBM would further augment China's ability to develop AI accelerators on par with cutting-edge offerings from US firms. The reflections within the Biden administration show a strategic effort to preserve America's technological edge by denying China access to key semiconductor innovations. However, implementing such stringent export controls is a delicate balancing act, as it risks heightening tensions and prompting Chinese retaliation. No final decision has been made, and officials continue weighing the proposed restrictions' pros and cons. Nonetheless, the discussions highlight the pivotal role that semiconductor technology plays in the great power rivalry between the US and China, especially in the AI era.

A recent Reuters investigation has uncovered evidence suggesting Chinese universities and research institutes may have circumvented US sanctions on high-performance NVIDIA GPUs by purchasing servers containing the restricted chips. The sanctions tightened on November 17, 2023, prohibit the export of advanced NVIDIA GPUs like the consumer GeForce RTX 4090 to China. Despite these restrictions, Reuters found that at least ten China-based organizations acquired servers equipped with the sanctioned NVIDIA GPUs between November 20, 2023, and February 28, 2024. These servers were purchased from major vendors such as Dell, Gigabyte, and Supermicro, raising concerns about potential sanctions evasion. When contacted by Reuters, the companies provided varying responses.

Dell stated that it had not observed any instances of servers with restricted chips being shipped to China and expressed willingness to terminate relationships with resellers found to be violating export control regulations. Gigabyte, on the other hand, stated that it adheres to Taiwanese laws and international regulations. Notably, the sale and purchase of the sanctioned GPUs are not illegal in China. This raises the possibility that the restricted NVIDIA chips may have already been present in the country before the sanctions took effect on November 17, 2023. The findings highlight the challenges in enforcing export controls on advanced technologies, particularly in the realm of high-performance computing hardware. As tensions between the US and China continue to rise, the potential for further tightening of export restrictions on cutting-edge technologies remains a possibility.

On March 29th, the United States announced another round of updates to its export controls, targeting advanced computing, supercomputers, semiconductor end-uses, and semiconductor manufacturing products. These new regulations, which took effect on April 4th, are designed to prevent certain countries and businesses from circumventing U.S. restrictions to access sensitive chip technologies and equipment. Despite these tighter controls, TrendForce believes the practical impact on the industry will be minimal.

The latest updates aim to refine the language and parameters of previous regulations, tightening the criteria for exports to Macau and D:5 countries (China, North Korea, Russia, Iran, etc.). They require a detailed examination of all technology products' Total Processing Performance (TPP) and Performance Density (PD). If a product exceeds certain computing power thresholds, it must undergo a case-by-case review. Nevertheless, a new provision, Advanced Computing Authorized (ACA), allows for specific exports and re-exports among selected countries, including the transshipment of particular products between Macau and D:5 countries.

SMIC, China's largest semiconductor manufacturer, is reportedly assembling a dedicated team to develop 3 nm semiconductor node technology, following reports of the company setting up 5 nm chip production for Huawei later this year. This move is part of SMIC's efforts to achieve independence from foreign companies and reduce its reliance on US technology. According to a report from Joongang, SMIC's initial goal is to commence operations of its 5 nm production line, which will mass-produce Huawei chipsets for various products, including AI silicon. However, SMIC is already looking beyond the 5 nm node. The company has assembled an internal research and development team to begin work on the next-generation 3 nm node.

The Chinese manufacturer is expected to accomplish this using existing DUV machinery, as ASML, the sole supplier of advanced EUV technology, is prohibited from providing equipment to Chinese companies due to US restrictions. It is reported that one of the biggest challenges facing SMIC is the potential for low yields and high production costs. The company is seeking substantial subsidies from the Chinese government to overcome these obstacles. Receiving government subsidies will be crucial for SMIC, especially considering that its 5 nm chips are expected to be up to 50 percent more expensive than TSMC's due to the use of older DUV equipment. The first 3 nm wafers from SMIC are not expected to roll out for several years, as the company will prioritize the commercialization of Huawei's 5 nm chips. This ambitious undertaking by SMIC represents a significant challenge for the company as it strives to reduce its dependence on foreign semiconductor technology and establish itself as an essential player in the global manufacturing industry.

ZOTAC debuted a massive flagship GeForce RTX 4090 24 GB custom design graphics card last summer—the Prime Gamer Force (PGF) OC edition model was released as a China exclusive product. ZOTAC's PGF shroud design remains the largest on the market—381 mm (L) x 154 mm (W) x 74 mm (D)—even with downgraded silicon beneath the surface. NVIDIA's China-specific GeForce RTX 4090D GPU was introduced last December, as a sanction conformant substitute for the full-fat version—naturally, ZOTAC has prepared a revised PGF model. This week, Expreview has published an in-depth review of the GeForce RTX 4090 D PGF OC edition graphics card. They found that ZOTAC's cooling system—three 11 cm fans and a vapor chamber—offered: "high-frequency stability...comparable to that of water-cooled (solutions)."

The Chinese publication reviewed the GALAX RTX 4090 D Metal Master model in January—at the time, software restrictions prevented the implementation of significant overclocks. It was theorized that future updates or community workarounds could bypass limitations, but the latest review—of ZOTAC's "super luxurious" PGF edition—indicates that this GeForce RTX 4090 D GPU's OC potential is still constricted. VideoCardz has pulled out essential details from the Expreview article: "(The PGF) has high maximum TGP (530 W) and a powerful 28-phase power PCB design. Despite the technological headroom, the card struggles to offer much of the overclocking potential. The team from Expreview only managed to squeeze 3.7% more performance from this card. That's despite 24.7% more power theoretically available." An underwhelming overclocking aspect is counterbalanced by the premium-tier card's impressive performance stability—the review also praised ZOTAC's quiet cooling solution and usage of high-end "heat dissipation materials."

The publication of ASML's 2023 Annual Report has revealed some interesting insights into how the photolithography producer remains diplomatic in times of global tension. Peter Wennink (President, Chief Executive Officer and Chair of the Board of Management) discussed his company's carefully considered tightrope act—here is his message to stakeholders: "In 2023, demand for our DUV systems continued to be strong, particularly in China. During the previous two years, our Chinese customers had received significantly fewer systems than they had ordered, due to global demand for our systems exceeding supply. However, the shifts in demand timing from other customers that we experienced in 2023 meant that we had the opportunity to backfill these orders for mature and midcritical nodes to China, while of course complying with export regulations." ASML is seemingly keen to continue doing business with Chinese customers, despite having to juggle with strict international trade rulings—as revealed in their financial report, trade in this region accounts for "26.3% of our 2023 total net sales." This places China in a second tier position, just behind Taiwan (29.3% of 2023 total net sale).

ASML was expecting to deliver a grand total of 600 DUV equipment units to Chinese customers by the end of 2025, but trade restriction adjustments nixed that avenue of business. The report's "Strategy and products" section highlights the company's concerns about narrowed lanes: "Geopolitical tensions may result in export control restrictions, trade sanctions, tariffs and more generally international trade regulations which may impact our ability to deliver our systems, technology, and services." China's leading foundry—Semiconductor Manufacturing International Corp (SMIC)—is reportedly targeting a 5 nm process node, although this would require a major readjustment of its existing collection of (older) lithography equipment. SMIC's flagship Shanghai location cannot upgrade to the most advanced DUV machinery in ASML's catalog, therefore workers are reliant on slightly antiquated gear (previously tasked with 7 nm manufacturing)—low yields and added expense are the anticipated headaches.

The first review of a Chinese-exclusive "RTX 4090D" GPU model hit the internet last week—Expreview received a sample GALAX RTX 4090 D Metal Master model not long ago, and their testing team proceeded to find out whether the nerfed version of NVIDIA's flagship gaming GPU was truly compromised in terms of performance. Effective October 2023, the US Federal Trade Commission placed restrictions on Team Green—thus blocking trade of units based on the "Ada Lovelace" AD102-300 GPU in China. In turn, a variant—AD102-250-A1—was prepared in order to confirm to new policies.

NVIDIA's China-specific GeForce RTX 4090D launched officially right at the end of 2023. Board partner GALAX seems to be leading the pack, with customized versions being sent out for evaluation. The GeForce RTX 4090D GPU arrives with a lesser configuration: 14,592 CUDA, 456 Tensor, and 114 RT cores—but the first review indicates that this only trails behind its uncompromised sibling by roughly 5 to 6% across sixteen games. It lags behind in Stable Diffusion benchmarks—an AI workload at 512x512 resolution shows a 10% difference, although the gaps narrows at 768x768 and 1024x1024.

A somewhat expected result of the November 17th export ban on RTX 4090 GPUs to China has been that AMD's top offerings; recently rumored to soon suffer the same fate, have been selling faster than AMD or its partners can make them. Predicting the worst, some OEMs such as Dell have allegedly already set their own self-imposed restrictions on the export of both Radeon and Instinct cards to China. Meanwhile insiders in board channels have not yet received any official warning that they can no longer sell RX 7900 XTXs or XTs to DIY markets, however vendors and consumers alike sense an oncoming storm and unable to get RTX 4090s are buying up as many of the high end Radeons as they can.

It's said that AMD's production capacity for this surge in sales has been underwhelming, and supply to Chinese board partners has been barely trickling in. This lack of incoming supply of the Radeon RX 7900 XTX and XT, restriction or not, is creating a shortage of cards in China that is expected to impact the remainder of Q4 2023 and much of Q1 2024. With no immediate restrictions expected for AMD's top cards it's possible that they could react and ramp up exports to cover the demand, however that could be risky if AMD is trying to avoid being targeted.

NVIDIA's most recent FY3Q24 financial reports reveal record-high revenue coming from its data center segment, driven by escalating demand for AI servers from major North American CSPs. However, TrendForce points out that recent US government sanctions targeting China have impacted NVIDIA's business in the region. Despite strong shipments of NVIDIA's high-end GPUs—and the rapid introduction of compliant products such as the H20, L20, and L2—Chinese cloud operators are still in the testing phase, making substantial revenue contributions to NVIDIA unlikely in Q4. Gradual shipments increases are expected from the first quarter of 2024.

The US ban continues to influence China's foundry market as Chinese CSPs' high-end AI server shipments potentially drop below 4% next year
TrendForce reports that North American CSPs like Microsoft, Google, and AWS will remain key drivers of high-end AI servers (including those with NVIDIA, AMD, or other high-end ASIC chips) from 2023 to 2024. Their estimated shipments are expected to be 24%, 18.6%, and 16.3%, respectively, for 2024. Chinese CSPs such as ByteDance, Baidu, Alibaba, and Tencent (BBAT) are projected to have a combined shipment share of approximately 6.3% in 2023. However, this could decrease to less than 4% in 2024, considering the current and potential future impacts of the ban.

The U.S. Commerce Department seems to have thrown a big spanner into the NVIDIA machinery, by informing the company that some US$5 billion worth of AI chip orders for China falls under the latest US export restrictions. The orders are said to have been heading for Alibaba, ByteDance and Baidu, as well as possibly other major tech companies in China. This made NVIDIA's shares drop sharply when the market opened in the US earlier today, by close to five percent, dropping NVIDIA's market cap below the US$1 Trillion mark. The share price recovered somewhat in the afternoon, putting NVIDIA back in the trillion dollar club.

Based on a statement to Reuters, NVIDIA doesn't seem overly concerned, despite what appears to be huge loss in sales, with a company spokesperson issuing the following statement "These new export controls will not have a meaningful impact in the near term". The US government will implement these new export restrictions from November, which obviously didn't give NVIDIA much of a chance to avoid them and it looks as if the company is going to have to find new customers for the AI chips. Considering the current demand for NVIDIA's chips, this might not be too much of a challenge for the company though.

Phytium, a Chinese semiconductor company that faced U.S. government sanctions from 2021, has introduced its latest data center processor, the 64-core Feiteng Tengyun S2500. Designed for cloud and high-performance computing applications, this processor features a large-capacity shared L3 cache, enhanced security capabilities for cloud servers, and improved memory subsystem reliability. The Feiteng Tengyun S2500 features 64 FTC661 cores developed by Phytium, which are based on Armv8 ISA. Reportedly, the CPU features 64 MB of L3 cache and 512 KB of L2 per core, bringing the total to 96 MB of processor cache. Compared to the previous generation line, the S2500 brings an L3 cache and TDP of 150 Watts, up from 90 Watts of previous generation.

This is Phytium's first new CPU in several years, raising questions about its production capacity and access to foundries, given its sanctions-related restrictions. It is currently unknown which foundry will manufacture the Feiteng Tengyun S2500, and we expect to hear more about it as (if) units get shipped. So far only display units have made appearance. Nonetheless, the company has continued its hardware development efforts and garnered interest in collaborating with Huawei to unify hardware and software ecosystems, which has yet to come to fruition.

Today, the US authorities published the updated version of the advanced computing and semiconductor manufacturing equipment rule, imposing additional restrictions on export of advanced chip manufacturing technology. These regulations will become effective after a period of 30 days. Given the length and complexity of the regulations, ASML will need to carefully assess any potential implications. However, as to our business, from the information we received, it is our understanding that the new regulations will be applicable to a limited number of fabs in China related to advanced semiconductor manufacturing.

These export control measures will likely have an impact on the regional split of our systems sales in the medium to long term. However, we do not expect these measures to have a material impact on our financial outlook for 2023 and for our longer-term scenarios for 2025 and 2030, as communicated during our Investor Day in November 2022. ASML will seek further clarification from the US authorities on the scope of these new regulations. ASML is fully committed to comply with all applicable laws and regulations including export control legislation in the countries in which we operate.

The Chinese government introduced restrictions on the export of gallium and germanium (plus their chemical compounds)—both crucial materials in the computer chip manufacturing process—a couple of months ago. Big players within the semiconductor industry shrugged this off as a minor inconvenience, and simply shifted to more expensive sources. Prior to an August 1 implementation of new rulings, according to Reuters, China exported 36.48 metric tons of germanium, and 22.72 tons of gallium (starting January 2023). Customers were in a rush to acquire as much material as possible, before the "cut off" date—so 8.63 tons of germanium and 5.15 tons of gallium got shifted overseas throughout July.

Reuters has kept a watchful eye on the situation since then—its latest report states that "China's exports of germanium and gallium items plunged in August, the first month of the export controls, customs data showed on Wednesday (September 20)." A Ministry of Commerce spokesman, He Yadong, last week revealed that his department will be granting a limited number of export licenses to interested parties, on the condition that these local companies "meet relevant requirements." An undisclosed percentage of submitted applications have already received government approval. Signed paperwork reportedly gives the thumbs up to "dual use" purposes, implying that potential customers are in the military and civilian fields.