On June 10, Taiwan's Ministry of Economic Affairs expanded its list of strategic export-controlled customers to include Huawei and Semiconductor Manufacturing International Corporation (SMIC). The ministry announced that this decision followed a review meeting focused on preventing the proliferation of arms and other national security concerns. Moving forward, any Taiwanese exporter must obtain formal government approval before shipping semiconductors, lithography machines, or related equipment to Huawei or SMIC. TSMC immediately confirmed its full compliance. Company representatives reminded stakeholders that no orders have been fulfilled for Huawei since September 2020 and pledged to enhance internal verification procedures to block any unauthorized transactions. These steps build on one‑billion‑dollar penalty, imposed after investigators determined that two million advanced AI chiplets had been supplied for Huawei's Ascend 910B accelerator without proper clearance.

For Huawei and SMIC, this latest measure compounds the challenges created by existing US export controls, which prohibit both companies from sourcing many US-origin technologies and designs. The two Chinese giants will accelerate efforts to develop domestic alternatives, yet true semiconductor independence remains a distant goal. Designing and building reliable extreme ultraviolet lithography systems demands years of specialized research and highly precise manufacturing capabilities. Scaling production without foreign expertise could introduce costly delays. In response, Chinese research institutes report that the country's first homegrown EUV lithography machines are slated to enter trial production in the third quarter of 2025. Meanwhile, state‑backed partners are racing to develop advanced packaging tools to rival those offered by ASML. Despite these initiatives, experts warn that catching up with global leaders will require substantial time and continued investment.

Mid-way through last month, Tencent's President—Martin Lau—confirmed that this company had stockpiled a huge quantity of NVIDIA H20 AI GPUs, prior to new trade restrictions coming into effect. According to earlier reports, China's largest tech firms have collectively spent $16 billion on hardware acquisitions in Q1'25. Team Green engineers are likely engaged in the creation of "nerfed" enterprise-grade chip designs—potentially ready for deployment later on in 2025. Huawei leadership is likely keen to take advantage of this situation, although it will be difficult to compete with the sheer volume of accumulated H20 units. The Shenzhen, Guangdong-based giant's Ascend AI accelerator family is considered to be a valid alternative to equivalent "sanction-conformant" NVIDIA products.

The controversial 910C model and a successor seem to be worthy candidates; as demonstrated by preliminary performance data, but fresh industry murmurs suggest teething problems. The Information has picked up inside track chatter from unnamed moles at ByteDance and Alibaba. During test runs, staffers noted the overheating of Huawei Ascend 910C trial samples. Additionally, they highlighted limitations within the Huawei Compute Architecture for Neural Networks (CANN) software platform. NVIDIA's extremely mature CUDA ecosystem holds a significant advantage here. Several of China's prime AI players—including DeepSeek—are reportedly pursuing in-house AI chip development projects; therefore positioning themselves as competing with Huawei, in a future scenario.

According to the Financial Times, gathering data from satellite images and industry intelligence, Huawei has endeavored to develop a domestic AI supply chain to bypass foreign tech restriction influence. In Guanlan, China, Huawei started developing a complete facility for manufacturing semiconductors on 7 nm technology for its custom processors. Out of frustration with SMIC's low output capacity, Huawei has secured the entire silicon production, from sourcing materials, chemicals, and wafer fab equipment to chip-making equipment to actual chip design. According to Dylan Patel of SemiAnalysis, "Huawei has embarked on an unprecedented effort to develop every part of the AI supply chain domestically from wafer fabrication equipment to model building," adding, "We have never seen one company attempt to do everything before."

It is also reported that Huawei's rivals in silicon manufacturing, SMIC and SMEE, have deployed engineers to help Huawei develop its own manufacturing flow. A few companies, all backed by Huawei with funding and research, are the backbone of this operation. SiCarrier, which we reported on back in March, supplies optical and X-ray inspection tools, atomic force microscopes, and alignment systems for metrology; gas-based and atomic layer deposition tools for film coating; plasma etchers for patterning; rapid thermal processors for material tuning; and electrical testing platforms for reliability screening. SwaySure and Fujian Jinhua supply memory chips, Si'En and Pehgjin supply power chips, and PWX and PST deal with logic.

DeepSeek, a company that took the AI world by storm with its R1 model, is preparing a new and reportedly much improved DeepSeek R2 model release, according to a well-known AI insider @iruletheworldmo on X. Powered by Huawei's Ascend 910B chip clusters, a possible Huawei Atlas 900, and DeepSeek's in-house distributed training framework, R2 pushes these accelerators to an impressive 82% utilization, translating to 512 PetaFLOPS of FP16 performance—half an exaFLOP in computing power. According to Huawei lab data, that's roughly 91% of what NVIDIA's older A100 clusters deliver, yet DeepSeek claims it cuts per-unit training costs by a remarkable 97.3%. Behind DeepSeek R2 is a carefully cultivated ecosystem of partners. Tuowei Information, a leading OEM in the Ascend family, manages over half of DeepSeek's supercomputing hardware orders, while Sugon provides liquid-cooled server racks capable of handling up to 40 kW per unit. To keep power consumption in check, Innolight's silicon-photonics transceivers shave off another 35% compared to traditional solutions.

Geographically, operations are split across major hubs: Runjian Shares runs the South China supercomputing center under contracts exceeding ¥5 billion annually, and Zhongbei Communications maintains a 1,500-PetaFLOP reserve in the Northwest for peak demands. On the software side, DeepSeek R2 already supports private deployment and fine-tuning, powering smart-city initiatives in 15 provinces through the Yun Sai Zhilian platform. North China's node, overseen by Hongbo Shares' Yingbo Digital, adds another 3,000 PetaFLOPS to the mix. If computing power is scarce, Huawei is ready to deploy its CloudMatrix 384 system, which is positioned as a domestic alternative to NVIDIA's GB200 NVL72. It features 384 Ascend 910C accelerators to achieve 1.7× the overall petaFLOPS and 3.6× the total HBM capacity of the NVL72 cluster—yet it lags significantly in per-chip performance and consumes nearly four times more power. Nonetheless, the R2 model launch is expected to come smoothly online, and we are waiting for the official launch and benchmarks to see its performance.

TSMC has acknowledged fundamental visibility limitations in its semiconductor supply chain, stating in its latest annual report that it "inherently lacks visibility regarding the downstream use or user of final products." This disclosure relates to an incident where 7 nm chips manufactured for Sophgo were later identified in Huawei's Ascend 910B/C AI accelerators, whose hardware is subject to US export restrictions. The contract foundry outlined its standard process: receiving GDS files through intermediaries, validating technical specifications, creating photomasks, and fabricating wafers without insight into end applications. Subsequent analysis revealed that those very chips matched Huawei's specifications, providing components for approximately one million dual‑chiplet AI accelerator units, with two million dies shipped to Huawei.

The report warns that compliance violations by supply‑chain partners, such as failing to secure proper import, export or re‑export permits, could trigger regulatory investigations and penalties, even when TSMC adheres to its established protocols. US already proposed a $1 billion fine for TSMC. This visibility gap just shows that challenges in semiconductor manufacturing, where complex distribution networks obscure the path between fabrication and deployment, are not easily overcome. Foundries are facing increasing pressure to enhance tracking capabilities despite the inherent limitations of the contract manufacturing model. US sanctions on Chinese companies are growing their walls even higher, and this could mean that sanction-abiding companies might avoid doing business with Chinese entities altogether to avoid getting fined.

Huawei recently revealed that its CloudMatrix 384 AI super node cluster can outperform NVIDIA's GB200 NVL72 in standard benchmarks, even though it consumes more power per performance unit. That system relies on Huawei's current Ascend 910C accelerators, which deliver strong raw compute performance but lag behind in efficiency metrics. To address this gap, Huawei is preparing the Ascend 920 family, with the training‑focused Ascend 920C built on SMIC's 6 nm process. According to DigiTimes, each 920C card will deliver more than 900 TeraFLOPS of BF16 half-precision performance. It also upgrades memory to HBM3 modules, providing 4,000 GB/s of bandwidth, which is up from the Ascend 910C's eight HBM2E stacks and 3,200 GB/s.

The existing Ascend 910C peaks at 780 TeraFLOPS in BF16 operations and uses a chip‑to‑chip interconnect bandwidth of 400 GB/s. Its packaging limits that bandwidth, but it still supports high‑speed communication between nodes in ultra‑dense AI clusters. Huawei will retain the chiplet‑based design in the 920C and refine the tensor acceleration engines for Transformer and Mixture‑of‑Experts models. Internal projections estimate that overall training efficiency on the 920C will improve by 30-40 percent compared to the 910C. This should narrow the performance‑per‑watt difference against competitors' solutions. In terms of system integration, the Ascend 920C will support PCIe 5.0 and next‑generation high‑throughput interconnect protocols. These features aim to improve resource scheduling and reduce latency in super node deployments, where tight node‑to‑node synchronization is critical. Huawei has not announced a firm release date for the Ascend 920C, but DigiTimes sources claim that it will enter mass production in the second half of 2025, which could mean just a few months from now.

Earlier today, TechPowerUp covered the alleged performance prowess of Huawei's CloudMatrix 384 system super node. According to SemiAnalysis opinion, the system's Ascend 910C AI accelerators are a generation behind—in terms of chip performance—when compared to NVIDIA's GB200 "Blackwell" AI GPU design. SMIC seemed to be in the picture, as Huawei's main fabrication partner—possibly with an in-progress 5 nm node process. Instead, SemiAnalysis has surmised that the Ascend 910C is based on plenty of non-native technologies. Huawei's (current and prior) "aggressive skirting of export controls" has likely enabled the new-gen AI chip's better than expected performance stats. SemiAnalysis documented the early sample's origins: "while the Ascend chip can be fabricated at SMIC, we note that this is a global chip that has HBM from Korea (Samsung), primary wafer production from TSMC (Taiwan), and is fabricated by 10s of billions of wafer fabrication equipment from the US, Netherlands, and Japan...One common misconception is that Huawei's 910C is made in China. It is entirely designed there, but China still relies heavily on foreign production."

Despite China's premiere foundry business making pleasing in-roads with a theorized "7 nm N+2" manufacturing test line, Huawei has seemingly grown impatient with native immature production options. Today's SemiAnalysis article presents a decent dose of inside knowledge: "while SMIC does have 7 nm, the vast majority of Ascend 910B and 910C are made with TSMC's 7 nm. In fact, the US Government, TechInsights, and others have acquired Ascend 910B and 910C and every single one used TSMC dies. Huawei was able to circumvent the sanctions on them against TSMC by purchasing ~$500 million of 7 nm wafers through another company, Sophgo...It is rumored Huawei continues to receive wafers from TSMC via another 3rd party firm, but we cannot verify this rumor." Another (fabless) Chinese chip design firm—Xiaomi—appears to still have direct/unrestricted access to TSMC manufacturing lines, albeit not for enterprise-grade AI products.

TSMC is confronting a potential $1 billion-plus penalty from the US Commerce Department after inadvertently fabricating compute chiplets later integrated into Huawei's Ascend 910 AI processor. The fine, potentially reaching twice the value of unauthorized shipments, reflects the scale of components that circumvented export controls limiting Chinese access to advanced semiconductor technology. The regulatory breach originated in late 2023 when TSMC processed orders from Sophgo, a design partner of crypto-mining firm Bitmain. These chiplets, which are manufactured on advanced process nodes and contain tens of billions of transistors, were identified in TechInsights teardown analysis of Huawei Ascend 910 AI accelerator, revealing a supply chain vulnerability where TSMC lacked visibility into the components' end-use.

Upon discovery of the diversion, TSMC immediately halted Sophgo shipments and engaged in discussions with Commerce Department officials. By January, Sophgo had been added to the Entity List, limiting its access to US semiconductor technology. A Center for Strategic and International Studies report revealed that Huawei obtained approximately two million Ascend 910B logic dies through shell companies that misled TSMC. Huawei's preference for TSMC-made dies was due to manufacturing challenges in domestic chip production. This incident has forced TSMC to strengthen its customer vetting protocols, including terminating its relationship with Singapore-based PowerAIR following internal compliance reviews. The enforcement process typically begins with a proposed charging letter detailing violations and penalty calculations, followed by a 30-day response period. As Washington tightens restrictions on AI processor exports to Chinese entities, semiconductor manufacturers are under increased pressure to implement rigorous controls throughout multinational supply chains.

A consortium of over 50 Chinese companies, including names like Huawei, Hisense, and TCL, has unveiled a domestic alternative to HDMI that offers up to 192 Gbps bandwidth and 480 W of power delivery. This new standard, the General Purpose Media Interface (GPMI), supports next-generation multimedia devices, meeting the growing demands of 8K resolution, higher refresh rates, and simplified connectivity. There are two variants available: a smaller Type-C model providing 96 Gbps and 240 W and a larger Type-B model delivering the full 192 Gbps and 480 W. Developed as a third-generation audio and video interface, GPMI addresses the limitations of older standards such as DVI and VGA while vastly outperforming HDMI 2.1's 48 Gbps and DisplayPort 2.1's 80 Gbps in data throughput. Its design enables bidirectional communication, seamlessly transmitting video, audio, data, and power over a single cable.

The standard's architecture includes a primary data link that can be split into various configurations—such as 6+2 or 1+7 channels—to adapt to different usage scenarios. In addition to its high-bandwidth data channels, GPMI features auxiliary links for device management, cable information, and a limited USB 2.0 connection. The Type-C variant, which has received approval from the USB Association, ensures compatibility with the USB-C ecosystem, helping with the integration for smart TVs and other connected devices. Primarily developed for the domestic market, GPMI also aims to reduce China's dependence on Western-controlled standards and licensing regimes.

SiCarrier, a Huawei-backed Chinese semiconductor tool manufacturer, has launched a comprehensive suite of semiconductor manufacturing tools at this year's Semicon China. These tools are strategically essential to China's semiconductor self-sufficiency and a major step towards competitive nodes from the mainland. The new lineup spans multiple categories: optical inspection, deposition, etch, metrology, and electrical performance testing. Until now, Chinese chipmakers often depended on older-generation foreign equipment, but SiCarrier's new lineup promises domestic alternatives tailored to modern manufacturing processes. The tools address every stage of semiconductor fabrication, from inspecting microscopic defects to etching intricate circuits.

For quality control, SiCarrier's Color Mountain series functions like a high-powered microscope, using intense lighting and advanced imaging algorithms to examine both sides of silicon wafers for flaws as small as dust particles. Complementing this, the Sky Mountain series ensures the perfect alignment of circuit layers, which need perfect stacking, using diffraction-based measurements (analyzing light patterns) and direct image comparisons. The New Mountain suite combines specialized tools to analyze materials at the atomic level. One standout is the atomic force microscope (AFM), which maps surface topography with a nanoscale "finger," while X-ray techniques (XPS, XRD, XRF) act like forensic tools, revealing the chemical composition, crystal structure, and elemental makeup.

A mysterious HiSilicon Kirin X90 processor was included in a Chinese state report; the official assessment document seems to cover the topics of native CPU reliability and security. Jukanlosreve—a keen watcher of semiconductor industry inside tracks—highlighted the unannounced technology on social media. They alerted TP Huang (aka tphuang) to this discovery, possibly in reaction to the latter's reporting of a wholly Huawei-designed AI laptop. Last week, we heard about a speculated portable enterprise PC series powered by Kunpeng-920 mobile processors. HiSilicon is a Chinese fabless semiconductor firm, operating under Huawei ownership—normally, their Kirin processors are designated as smartphone-based solutions. Industry watchdogs believe that company leadership is paving the way for a new generation of personal and server processors—their current Kunpeng-900 series debuted back in 2019, so natural successors are very likely in the development pipeline. Early last year, insiders proposed that Huawei had prioritized its Ascend 910B AI accelerator chip—despite these rumors, Kirin-related leaks continued to trickle out.

According to industry moles, HiSilicon tends to spread its custom Arm-based Taishan core design across modern Kirin and Kunpeng processor families. Based on Jukanlosreve's initial detective work, Tom's Hardware proposed a plausible outlook for the (leaked) chip's future. Their report theorized: "Huawei's server and PC divisions have been relatively quiet, as evidenced by the lack of new Kunpeng SoC designs. The new Kirin X90, despite its name, could be a possible successor, considering that Huawei is reportedly launching a new 'AI PC' with HarmonyOS next month. It's likely to be fabricated using SMIC's 7 nm process featuring Taishan V120 cores based on either the Armv8 or Armv9 architectures, which are not subject to the US trade ban." Instead of using an old hat Kunpeng-920 SoC, Huawei's forthcoming next-gen "Qingyun" AI laptop could be equipped with a Kirin X90 APU—Tom's Hardware foresees an Apple MacBook-esque "integration of custom hardware and software" with the Chinese tech firm's fully in-house developed model.

The rising prominence of Huawei's Ascend accelerator series has generated plenty of news articles over the past couple of months. Earlier in the week, reports suggested that the Chinese technology giant had circumvented international trade sanctions. Many observers will forget that parts of Huawei's portfolio contain "less ambitious" products; namely notebook PCs. Despite navigating tricky conditions in 2024, the manufacturer still has access to a supply of Intel-produced processors—but a contingency plan is reportedly in place. TP Huang (aka tphuang) reckons that a completely in-house formulated "domestic commercial PC" is due for release in China. The technology soothsayer's social media post mentioned a "7 nm Kunpeng-920 CPU" clocked at 2.6 GHz. Previously, Huawei's Kunpeng proprietary processor designs were deployed in server (2024) and desktop (2020) environments.

According to Huawei Central, company representatives have disclosed that the mysterious Kunpeng-920-powered AI laptop will be added to their "Qingyun" enterprise-oriented product range. TP Huang elaborated with additional leaked details—the forthcoming model is tipped to feature: "fully domestic power management integrated circuits (PMIC); including Unisoc CT100 replacing EPSON in clock generator. It is expected to run on HarmonyOS PC version and a localized dev tool chain." An extension of Huawei and DeepSeek's current relationship is anticipated to trickle down to portable devices—insiders believe that the new AI notebook will support DeepSeek's model on edge. TP Huang predicts an April launch window.

According to a recent Center for Strategic and International Studies report, Huawei got its hand on approximately two million Ascend 910B logic dies through shell companies that misled TSMC. This acquisition violates US export controls designed to restrict China's access to advanced semiconductor technology. The report details how Huawei leveraged intermediaries to procure chiplets for its AI accelerators before TSMC discovered the deception and halted shipments. These components are critical for Huawei's AI hardware roadmap, which progressed from the original Ascend 910 (manufactured by TSMC on N7+ until 2020) to the domestically produced Ascend 910B and 910C chips fabricated at SMIC using first and second-generation 7 nm-class technologies, respectively. Huawei reportedly wanted TSMC-made dies because of manufacturing challenges in domestic chip production. The Ascend 910B and 910C reportedly suffer from poor yields, with approximately 25% of units failing during the advanced packaging process that combines compute dies with HBM memory.

Despite these challenges, the performance gap with market-leading solutions still remains but has narrowed considerably, with the Ascend 910C reportedly delivering 60% of NVIDIA H100's performance. Huawei has executed a strategic stockpiling initiative, particularly for high-bandwidth memory components. The company likely acquired substantial HBM inventory between August and December 2024, when restrictions on advanced memory sales to China were announced but not yet implemented. The semiconductor supply chain breach shows that enforcing technology export controls is challenging, and third parties can still purchase silicon for restricted companies. While Huawei continues building AI infrastructure for both internal projects and external customers, manufacturing constraints may limit its ability to scale deployments against competitors with access to more advanced manufacturing processes. Perhaps a future domestic EUV-based silicon manufacturing flow will allow Huawei to gain access to more advanced domestic production, completely circumventing US-imposed restrictions.

China's domestic extreme ultraviolet (EUV) lithography development is far from a distant dream. The newest system, now undergoing testing at Huawei's Dongguan facility, leverages laser-induced discharge plasma (LDP) technology, representing a potentially disruptive approach to EUV light generation. The system is scheduled for trial production in Q3 2025, with mass manufacturing targeted for 2026, potentially positioning China to break ASML's technical monopoly in advanced lithography. The LDP approach employed in the Chinese system generates 13.5 nm EUV radiation by vaporizing tin between electrodes and converting it to plasma via high-voltage discharge, where electron-ion collisions produce the required wavelength. This methodology offers several technical advantages over ASML's laser-produced plasma (LPP) technique, including simplified architecture, reduced footprint, improved energy efficiency, and potentially lower production costs.

The LPP method relies on high-energy lasers and complex FPGA-based real-time control electronics to achieve the same result. While ASML has refined its LPP-based systems over decades, the inherent efficiency advantages of the LDP approach could accelerate China's catch-up timeline in this critical semiconductor manufacturing technology. When the US imposed sanctions on EUV shipments to Chinese companies, the Chinese semiconductor development was basically limited as standard deep ultraviolet (DUV) wave lithography systems utilize 248 nm (KrF) and 193 nm (ArF) wavelengths for semiconductor patterning, with 193 nm immersion technology representing the most advanced pre-EUV production technique. These longer wavelengths contrast with EUV's 13.5 nm radiation, requiring multiple patterning techniques to achieve advanced nodes.

Huawei is likely celebrating milestones on multiple fronts—as reported earlier this month, the Chinese technology manufacture has pulled in record revenues and experienced consistent growth. Additionally, industry insiders believe that things are going well within the company's production pipeline. According to a Financial Times report, Huawei's next-generation AI accelerator model is on the way—the unannounced "Ascend 910C" is touted to directly compete with NVIDIA's H100 AI GPU. Industry moles believe that Huawei has partnered with SMIC for the manufacture of in-house accelerator designs. Whispers suggest a selection of the foundry's 7 nm N+2 process.

The alleged doubling of production yields (within a year)—from 20% to 40%—signals a significant achievement. As reported by FT, this milestone indicates Huawei's Ascend chip production line becoming profitable for the very first time. Two inside sources propose that Huawei and SMIC are targeting a 60% yield goal in the near future. In 2025, leaked plans suggest production tallies of roughly 100,00 Ascend 910C processors, and 300,000 of the current-gen Ascend 910B chip.

China's dominance in semiconductor equipment procurement is expected to face its first setback since 2021, with spending projected to decrease from $41 billion to $38 billion in 2025, according to semiconductor research firm TechInsights. This 6% decline marks a significant shift for the world's largest buyer of wafer fabrication equipment, whose purchases represented 40% of global sales in 2024. The downturn reflects mounting pressures from both market dynamics and geopolitical constraints. US export controls targeting advanced semiconductor capabilities have intensified while domestic chipmakers grapple with overcapacity in mature node segments. SMIC, China's leading foundry, has already signaled concerns about oversupply risks in this sector, where Chinese manufacturers have rapidly expanded their market share against Taiwanese competitors.

Despite these headwinds, Chinese equipment manufacturers have notably advanced domestic capability development. Naura Technology Group has emerged as the seventh-largest global equipment manufacturer, while AMEC continues to expand its international presence. However, critical gaps persist in China's semiconductor equipment ecosystem, particularly in lithography systems, where dependence on foreign suppliers like ASML remains high. TechInsights data reveals that Chinese companies supplied only 17% of testing tools and 10% of domestic assembly equipment in 2023. The spending reduction comes after a period of aggressive stockpiling prompted by US sanctions to limit Beijing's access to advanced chipmaking capabilities, especially those applicable to artificial intelligence and military applications. However, Chinese manufacturers have demonstrated resilience, with SMIC and Huawei successfully producing advanced chips through alternative, albeit more costly, manufacturing methods.

Huawei Technologies reported a robust 22% year-over-year revenue increase for 2024, reaching 860 billion yuan ($118.27 billion), demonstrating remarkable resilience amid continued US-imposed trade restrictions. The Chinese tech giant's resurgence was primarily driven by its revitalized smartphone division, which captured 16% of China's domestic market share, overtaking Apple in regional sales. This achievement was notably accomplished by deploying domestically produced chipsets, marking a significant milestone for the company. In collaboration with Chinese SMIC, Huawei delivers in-house silicon solutions to integrate with HarmonyOS for complete vertical integration. The company's strategic diversification into automotive technology has emerged as a crucial growth vector, with its smart car solutions unit delivering autonomous driving software and specialized chips to Chinese EV manufacturers.

In parallel, Huawei's Ascend AI 910B/C platform recently announced compatibility with DeepSeek's R1 large language model and announced availability on Chinese AI cloud providers like SiliconFlow. Through a strategic partnership with AI infrastructure startup SiliconFlow, Huawei is enhancing its Ascend cloud service capabilities, further strengthening its competitive position in the global AI hardware market despite ongoing international trade challenges. Even if the company can't compete on performance versus the latest solutions from NVIDIA and AMD due to the lack of advanced manufacturing required for AI accelerators, it can compete on costs and deliver solutions that are typically much more competitive with the price/performance ratio. Huawei's Ascend AI solutions deliver modest performance. Still, the pricing makes AI model inference very cheap, with API costs of around one Yaun per million input tokens and four Yuan per one million output tokens on DeepSeek R1.

When High-Flyer, the hedge fund behind DeepSeek, debuted its flagship model, DeepSeek R1, the tech world went downward. No one expected Chinese AI companies can produce high-quality AI model that rivals the best from OpenAI and Anthropic. While there are rumors that DeepSeek has access to 50,000 NVIDIA "Hopper" GPUs, including H100, H800, and H20, it seems like Huawei is ready to power Chinese AI infrastructure with its AI accelerators. According to the South China Morning Post, Chinese cloud providers like SiliconFlow.cn are offering DeepSeek AI models for inference on Huawei Ascend 910B accelerators. For the price of only one Yuan for one million input tokens, and four Yuan for one million output tokens, this economic model of AI hosting is fundamentally undercutting competition like US-based cloud providers that offer DeepSeek R1 for $7 per million tokens.

Not only is running on the Huawei Ascend 910B cheaper for cloud providers, but we also reported that it is cheaper for DeepSeek itself, which serves its chat app on the Huawei Ascend 910C. Using domestic accelerators lowers the total cost of ownership, with savings passed down to users. If Western clients prefer AI inference to be served by Western companies, they will have to pay a heftier price tag, often backed by the high prices of GPUs like NVIDIA H100, B100, and AMD Instinct MI300X.

Huawei's Ascend 910C AI chip was positioned as one of the better Chinese-developed alternatives to NVIDIA's H100 accelerator—reports from last autumn suggested that samples were being sent to highly important customers. The likes of Alibaba, Baidu, and Tencent have long relied on Team Green enterprise hardware for all manner of AI crunching, but trade sanctions have severely limited the supply and potency of Western-developed AI chips. NVIDIA's region-specific B20 "Blackwell" accelerator is due for release this year, but industry watchdogs reckon that the Ascend 910C AI GPU is a strong rival. The latest online rumblings have pointed to another major Huawei customer—DeepSeek—having Ascend silicon in their back pockets.

DeepSeek's recent unveiling of its R1 open-source large language model has disrupted international AI markets. A lot of press attention has focused on DeepSeek's CEO stating that his team can access up to 50,000 NVIDIA H100 GPUs, but many have not looked into the company's (alleged) pool of natively-made chips. Yesterday, Alexander Doria—an LLM enthusiast—shared an interesting insight: "I feel this should be a much bigger story—DeepSeek has trained on NVIDIA H800, but is running inference on the new home Chinese chips made by Huawei, the 910C." Experts believe that there will be a plentiful supply of Ascend 910C GPUs—estimates from last September posit that 70,000 chips (worth around $2 billion) were in the mass production pipeline. Additionally, industry whispers suggest that Huawei is already working on a—presumably, even more powerful—successor.

A South Korean online retailer has listed Huawei's eKitStore Extreme 200E M.2 NVMe SSD for sale at 47,500 Won ($32) for the 1 TB model. This marks Huawei's entry into consumer storage products after previously only manufacturing server SSDs. The PCIe 4.0 drive's specifications show read speeds of 7,400 MB/s and write speeds of 6,700 MB/s. Beyond confirming the use of QLC memory without a DRAM cache, Huawei has not disclosed the drive's internal components. Due to trade restrictions, both the controller and memory chips likely come from Chinese manufacturers, though specific suppliers remain unknown. Chinese makers have been recorded to use Silicon Motion's SSD controllers, but domestic controllers are also an option here.

Manufacturing costs for similar drives typically exceed $32, raising questions about the pricing strategy. The drive's components could match those used in other brands' SSDs, as multiple storage companies often use identical parts in their products. For NAND Flash, Huawei likely sourced its chips from YMTC, whose Xtacking 4.0 proves to be quite a success. The drive has appeared only at one retailer so far, and Huawei has not announced plans for sales in other regions. This price is notably lower than comparable PCIe 4.0 drives. Common retail prices for 1 TB PCIe 4.0 SSDs are typically more than double this amount. Whether this represents a temporary price or a long-term strategy remains unclear. Perhaps Huawei is operating at a net loss to gain some customers, with possible plans for more SSDs in the future, along with the PCIe 5.0 version.

The Biden administration is set to announce new, targeted restrictions on China's semiconductor industry, focusing primarily on emerging chip manufacturing equipment companies rather than broad industry-wide limitations. According to Bloomberg, these new restrictions are supposed to take effect on Monday. The new rules will specifically target two manufacturing facilities owned by Semiconductor Manufacturing International Corp. (SMIC) and will add select companies to the US Entity List, restricting their access to American technology. However, most of Huawei's suppliers can continue their operations, suggesting a more mild strategy. The restrictions will focus on over 100 emerging Chinese semiconductor equipment manufacturers, many of which receive government funding. These companies are developing tools intended to replace those currently supplied by industry leaders such as ASML, Applied Materials, and Tokyo Electron.

The moderated approach comes after significant lobbying efforts from American semiconductor companies, who argued that stricter restrictions could disadvantage them against international competitors. Major firms like Applied Materials, KLA, and Lam Research voiced concerns about losing market share to companies in Japan and the Netherlands, where similar but less stringent export controls are in place. Notably, Japanese companies like SUMCO are already seeing the revenue impacts of Chinese independence. Lastly, the restrictions will have a limited effect on China's memory chip sector. The new measures will not directly affect ChangXin Memory Technologies (CXMT), a significant Chinese DRAM manufacturer capable of producing high-bandwidth memory for AI applications.

Huawei is advancing archive storage technology with the development of a new solution called the Magneto-Electric Disk (MED). Aiming to provide an alternative to traditional hard drives, this storage device merges tape and SSD technologies to deliver high capacity and low power consumption. The first-generation MED can hold up to 72 TB of data in a compact 7-inch housing and uses significantly less energy than standard HDDs. The MED was developed as part of Huawei's efforts to mitigate potential supply shortages of conventional hard drives, particularly given the US export restrictions to China. By combining tape storage, traditionally slow but reliable, with a high-speed SSD, Huawei has designed a hybrid storage solution that addresses both speed and efficiency needs.

Data is initially recorded on the SSD for quick access and is then written to the tape in sequential blocks for long-term storage. Frequently accessed data, or "warm" data, stays on the SSD for faster retrieval, while "cold" data, accessed less often, is stored on the tape. According to early specifications, a MED rack can store over 10 petabytes and consumes under 2,000 watts of power, significantly less than HDD-based storage. This efficiency could redefine data archiving, especially for large-scale data centers. Data throughput on the MED is projected to reach 8 GB/s, highlighting its potential to handle high-demand environments. The MED is expected to launch in 2025, with a second, more compact generation targeting a 3.5-inch form factor in 2026 or 2027.

According to a recent Reuters report, TSMC has decided to cut off Chinese firm Sophgo following the discovery of TSMC-manufactured components in Huawei's advanced AI processor. The suspension came after technology research firm TechInsights identified a TSMC-manufactured chip within Huawei's Ascend 910B processor during a detailed analysis. This discovery raised significant concerns, as Huawei has been restricted from accessing such technology under US export controls since 2020. TSMC promptly notified US authorities upon learning of the situation and launched an internal investigation. While being sanctioned by the US, Huawei needed to use a proxy firm to get access to high-end silicon manufacturing to produce its Ascend accelerators.

Sophgo, which has ties to cryptocurrency mining equipment manufacturer Bitmain, strongly denies any business relationship with Huawei. The company states it has provided TSMC with a detailed investigation report asserting its compliance with all applicable laws, saying: "SOPHGO has never been engaged in any direct or indirect business relationship with Huawei. SOPHGO has been conducting business in strict compliance with applicable laws and regulations, including but not limited to all the applicable US national export control laws and regulations, and has never been in violation of any of such laws and regulations. SOPHGO has provided detailed investigation report to TSMC to prove that SOPHGO is not related to the Huawei investigation."

Huawei has reportedly started shipping its Ascend 910C accelerator—the company's domestic alternative to NVIDIA's H100 accelerator for AI training and inference. As the report from China South Morning Post notes, Huawei is shipping samples of its accelerator to large NVIDIA customers. This includes companies like Alibaba, Baidu, and Tencent, which have ordered massive amounts of NVIDIA accelerators. However, Huawei is on track to deliver 70,000 chips, potentially worth $2 billion. With NVIDIA working on a B20 accelerator SKU that complies with US government export regulations, the Huawei Ascend 910C accelerator could potentially outperform NVIDIA's B20 processor, per some analyst expectations.

If the Ascend 910C receives positive results from Chinese tech giants, it could be the start of Huawei's expansion into data center accelerators, once hindered by the company's ability to manufacture advanced chips. Now, with foundries like SMIC printing 7 nm designs and possibly 5 nm coming soon, Huawei will leverage this technology to satisfy the domestic demand for more AI processing power. Competing on a global scale, though, remains a challenge. Companies like NVIDIA, AMD, and Intel have access to advanced nodes, which gives their AI accelerators more efficiency and performance.

A new forecast from the International Data Corporation (IDC) Worldwide Quarterly Personal Computing Device Tracker shows shipments of personal computing devices are expected to grow 2.6% year over year in 2024 to 398.9 million units. The traditional PC market will remain flat in 2024 with 261 million units shipped while the tablet market is forecast to grow 7.2% year over year as a refresh cycle and project investments are expected to drive the market.

For traditional PCs, the global market excluding China is expected to grow 2.8% in 2024 as China continues to suffer through a confluence of macroeconomic challenges, including high youth employment, deflation, and a tumultuous real estate market. However, China's economic concerns have largely impacted just the PC market as tablet demand has proven to be more resilient thanks to Huawei's efforts.