Chinese hardware enthusiasts and AI researchers are rolling out custom versions of NVIDIA's latest flagship GPU, the GeForce RTX 5090D "Blackwell" variant for China. Originally intended as a special model that complies with US export regulations, the RTX 5090D is proving to be an excellent choice for smaller AI labs. Building on the trend set by specialist firms modifying the RTX 3090 and RTX 4090, these teams have taken the GB202 chip and mounted it on new PCBs with blower‑style coolers better suited to AI server racks than gaming rigs. Instead of large fans and complex heatsinks, each card uses a dual‑slot blower design that pushes hot air straight out the back of the chassis.

The power connector has also been moved to the rear bracket, making it easier to stack cards in clusters without cables getting in the way. Each unit still packs 32 GB of GDDR7 memory and a PCIe 5.0 x16 interface, although the maximum power draw remains unconfirmed. While previous "D" series cards topped out around 450 W, with Blackwell, this figure might approach 575 W, similar to the regular RTX 5090. This blower cooler could be undersized for a near‑600 W TDP, but with undervolting and underclocking, it may operate with stability over long workloads. Given that Chinese AI labs are struggling to acquire enough GPU capacity, some smaller labs and university researchers are possibly far away from getting enough compute resources for their work. So, repackaging these GPUs into workstation bodies is the only choice to get some acceleration in the hands of the masses.

The unannounced Radeon RX 9070 GRE 12 GB graphics card model seems to be next in line within AMD's RDNA 4 range. Despite official presentation material teasing a Q2'25 launch of Radeon RX 9060 Series cards, insiders believe that Team Red will debut an in-between option—possibly before the arrival of Radeon RX 9060 XT 16 GB and 8 GB SKUs. Rumored new-generation "Great Radeon Edition" (GRE) cards are expected to launch as Chinese market exclusives; potentially as a "strategic" gap fill. Local board partner moles have whispered about almost zero replenishments of Radeon RX 9070 (non-XT) 16 GB stock in the region. As reported earlier today, some of VideoCardz's inside sources insist that Radeon RX 9070 GRE cards will—eventually—replace Radeon RX 9070 options. Initial leaks suggested fundamental "step-down" specification pillars: 12 GB of VRAM, a 192-bit memory interface, and a "reduced" "Navi 48" GPU die.

According to fresh claims, the Radeon RX 9070 GRE model could utilize a "Navi 48 XL" GPU variant. VideoCardz reckons that a quarter of the original GPU core count has been shut off; resulting in a total of 3072 stream processors. Comparatively, the Radeon 9070 XT arrived with 4096 SPs. The Radeon 9070 launched with 3584 units. The report delved into alleged core frequency details: "the RX 9070 GRE's clocks will be higher than the RX 9070, reportedly at 2.79 GHz boost, resulting in around 17.1 TFLOPS compared to 18 TFLOPS on the RX 9070...Some custom variants we know of will approach a 3.0 GHz boost clock, so there is definitely room for overclocking." The latest spec sheet leak confirms a 12 GB pool of VRAM, in GDDR6 form—VideoCardz weighed in with some embellishments: "the memory will not be clocked at 20 Gbps, as on (already launched) RX 9070 (XT) and (incoming) RX 9060 XT, but at 18 Gbps. This means that the memory bandwidth will be about 1/3 lower than the RX 9070 (XT) at 432 GB/s." Certain industry observers reckon that AMD will continue to rely on AIBs to produce an all-custom lineup of forthcoming RDNA 4 products. So-called "reference designs" (MBA) have turned up in China, but only in very limited numbers—sold via hazy avenues.

The Taiwan media is reporting that Foxconn has been busy moving more and more of its production from China to India, with the main focus being on Apple's iPhone and iPad products for now. Foxconn has been producing iPhones for the Indian market in India since 2019, but the story here is that China has banned Foxconn from sending Chinese staff to India to train the local staff and oversee the production. This has resulted in Foxconn having to send staff from Taiwan to India instead, but they might not be as familiar with the entire production process as the ones that are based in the factories in China.

This is only part of the issue though, as the news reports are also mentioning that China is preventing Foxconn from sending machines and equipment needed for the production to India. The same issue is said to be affecting other Taiwanese companies that are moving some of their production out of China to India, but they've had no issues when it comes to sending equipment to other nations in South East Asia. The equipment is apparently not possible to produce locally in India either, which means that these companies are more or less being forced to keep their production in China. To work around this issue, Foxconn is said to be sending half-finished phones and components to India, to do the final assembly there, to keep up production in India. So far, this is just rumours out of Taiwan, but it looks like China is intent on keeping the production on its shores for as long as it can.

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.

Last year, leakers posited that Xiaomi engineers were working on proprietary new-generation mobile chipset. The giant Chinese technology company has produced proprietary SoC designs in the past—most notably 2017's Pengpai/Surge S1 (blink, and you missed it)—but has largely relied on a wide swatch of Qualcomm solutions for deployment in smartphones. According to a fairly fresh ITHome news report (via a source at Sina Technology), Xiaomi has established a "chip platform department" under the umbrella of its mobile phone product development division. The exact nature of this newly-formed team remains semi-mysterious, and how it intermeshes with current (rumored) proprietary chip design efforts. The megacorporation's speculated "dedicated chip platform department" is said to be working on a "Xuanjie" SoC series.

Qin Muyun—a former senior director of product marketing at Qualcomm—is purported to be the new department's leader. Insiders believe that Muyun will answer directly to Xiaomi's CEO: Lei Jun. Over the past half decade, company engineers have accumulated chip designing experience in multiple lesser technology fields—ITHome lists: "imaging, fast charging, power management, communication, and display" aspects. Xiaomi leadership could be "shaking up" its first-party SoC development projects. Two weeks ago, rumors turned up online regarding a node process downgrade—from 3 nm to 4 nm. Smartphone industry watchdogs reckon that Xiaomi's forthcoming flagship chipset design will perform on the level of Qualcomm's first generation Snapdragon 8 mobile processor (from 2021).

AMD revealed today that it anticipates charges of up to $800 million following the Trump administration's latest crackdown on exports of advanced processors to China. The company confirmed that these new rules affect its MI308 product line. The U.S. Commerce Department declared on Tuesday that it was putting in place new licensing requirements impacting several high-performance chips, including NVIDIA's H20, AMD's MI308, and similar products. This regulatory move comes at a high price for AMD, as China stands as its second-biggest market in 2024 with about $6.23 billion in revenue, over 24% of the company's total sales.

AMD's regulatory filing reveals that the company faces $800 million in charges due to inventory issues, purchase commitments, and needed reserves. AMD "expects to apply for licenses but there is no assurance that licenses will be granted,". This uncertainty grows when we consider what Jefferies analysts pointed out on Tuesday: the U.S. has never agreed on licenses for graphics processor unit shipments to China. This news comes right after NVIDIA's announcement that it would take $5.5 billion in charges because of the same export rules. As a result, AMD and NVIDIA stocks were dropping by more than 5%.

President Trump's administration has announced that NVIDIA's H20 AI chip will require a special export license for any shipment to China, Hong Kong, or Macau for the indefinite future. The Commerce Department delivered the news to NVIDIA on April 14, 2025, citing worries that the H20 could be redirected into Chinese supercomputers with potential military applications. NVIDIA designed the H20 specifically to comply with earlier US curbs by scaling back performance from its flagship H100 model. The H20 features 96 GB of HBM3 memory running at up to 4.0 TB/s, delivers roughly 296 TeraFLOPS of mixed‑precision compute power, and offers a performance density of about 2.9 TeraFLOPS per die. Its single‑precision (FP32) throughput is around 74 TeraFLOPS, with FP16 performance reaching approximately 148 TeraFLOPS. In a regulatory filing on April 15, NVIDIA warned that it will record about $5.5 billion in writedowns this quarter related to H20 inventory and purchase commitments now blocked by the license requirement.

Shares of NVIDIA fell roughly 6 percent in after‑hours trading on April 15, triggering a wider sell‑off in semiconductor stocks from the US to Japan. South Korea's Samsung and SK Hynix each slid about 3 percent, while AMD also dropped on concerns about broader chip‑export curbs. Analysts at Bloomberg Intelligence project that, if the restrictions persist, NVIDIA's China‑related data center revenue could shrink to low‑ or mid‑single digits as a percentage of total sales, down from roughly 13 percent in fiscal 2024. Chinese AI players such as Huawei stand to gain as customers seek alternative inference accelerators. Commerce Secretary Howard Lutnick has pledged to maintain a tough stance on chip exports to China even as NVIDIA commits up to $500 billion in US AI infrastructure investments over the next four years. Everyone is now watching closely to see whether any H20 export licenses are approved and how long the ban might remain in place.

Rumors of a forthcoming "Elder Scrolls IV: Oblivion Remake" launch popped up online last month; as predicted by NatetheHate. The veteran leaker—of video games industry inside info—proposed a public unveiling; scheduled in at some point within a March to April window. Since then, data miners have combed through various web presences—including Virtuos' online site. Going back to 2021, leaks suggested that this Chinese contract development house was deeply involved in the reimagining/modernizing of Bethesda's 2006 smash-hit.

Yesterday, Giant Bomb's Jeff Grubb posited that a formal introduction of the alleged "Elder Scrolls IV: Oblivion Remake" will happen "next week." Insiders reckon that the Unreal Engine 5-powered game will launch on "Xbox, PS5, PC, and Game Pass." Virtuos seems to be preparing itself for an imminent announcement; as evidenced by Elder Scrolls fans compiling leaked screenshots and promotional imagery within an Imgur album. Familiar locations and character designs are visible in gameplay captures. One visual "portfolio" grab compares the speculated new version to Bethesda's original build (2006 vs. 2025).

Vastarmor—a Chinese manufacturer—was introduced as an important board partner during AMD's special Radeon 9070 Series event. Despite having limited reach outside of its home base, the fairly young brand's custom designs have gained press traction around the globe. TechPowerUp covered Vastarmor's introduction of new Super Alloy series Radeon RX 9070 XT and RX 9070 models back in late February. The triple-fan ARGB-lit flagship model was positioned in good (rival) company, thanks to a maximum 3.1 GHz boost capability. As reported by VideoCardz, Vastarmor has added another custom Radeon RX 9070 XT option to its RDNA 4 product portfolio. A freshly-published product page has outlined specifications for the brand's new (non-Super) Alloy SKU.

Outside of Vastarmor's dedicated listing, a quick web search of the model's identifier code—AH-9070XT16TC1BR7N—reveals BIOS information, and little else. The Radeon RX 9070 XT Alloy seems to be a straightforward spin-off of its fancier sibling, albeit sans ARGB lighting zones. This barebone design sports non-transparent black fans, but preserves the flagship card's triple 8-pin power connector configuration. Vastarmor has downgraded its non-Super model's boost clock to a mere 3010 MHz (instead of 3100 MHz)—as configured on rival cards: ASUS PRIME RX 9070 XT OC, PowerColor Hellhound RX 9070 XT, and Sapphire PURE RX 9070 XT. Curiously, a white Alloy option is not present on Vastarmor's website—their current-gen Super Alloy range launched with pale alternatives.

GALAX did not expedite the development of a flagship GeForce RTX 5090D custom design; as evidenced by the absence of a top-flight Hall of Fame (HOF) option during launch week. "Cheaper" options—in the form of General and 1-Click models—were made available to the Chinese public on January 30. GALAX's fanciest new-gen "Blackwell" HOF model debuted with less potent hardware onboard: GeForce RTX 5080 OC LAB Plus-X Edition. Throughout early 2025, GALAX seeded early samples—of HOF GeForce RTX 5090D hardware—with tenured extreme overclocking organizations. Last month, almost finalized board designs became record breakers—achieved via liquid nitrogen cooled methodologies. Earlier prototypes sported a single 16-pin power connector, but newer Extreme Overclock (XOC) iterations were distributed with a second power delivery channel.

Yesterday, GALAX's social media channel teased the imminent launch of their air-cooled retail variant: "5090D HOF XOC Edition (Limited). Seems like it will be on China market soon…" Up until very recently, overclocking champions have shared photos of bare GALAX GeForce 5090D HOF cards—industry watchdogs reckoned that company engineers were still realizing a next-gen cooling solution. The Chinese manufacturer's promo material advertises the (presumably) very expensive forthcoming SKU as an option "only for better performance." Product renders showcase a familiar (optional) crown-adorned shroud, but it is not clear whether GALAX has revised its backplate design for this generation's XOC flagship.

Since teasing its next-gen flagship motherboard design—at CES 2025—the ASUS Republic of Gamers (ROG) team has spent time finalizing an option that is even more extravagant than their hardcore overclocking-oriented Crosshair X870E Apex model. Naturally, the premium sub-brand did not reveal pricing during preview events—instead, premature European e-tail listings suggested a €1200 (~$1350 USD) tag. Tony Wu and colleagues at ASUS China introduced an impressive swath of brand-new products at an official in-person showcase last week, in Changsha.

During proceedings, company representatives introduced their top-flight E-ATX format board with a hefty day one price: 9999 RMB ($1360 USD). The ASUS ROG Crosshair X870E Extreme is expected to launch later this month; likely starting off with the Chinese domestic market. At the time of writing, the manufacturer's various global branches have not disclosed localized details—be it pricing or availability. Last week's presser did not produce any surprises, since plenty of specification details (and promo shots) leaked out late last month. One interesting and very over the top feature is the model's integration of an adjustable full-color 5-inch LCD screen. The primary M.2 storage solution will be cooled by an "innovative" 3D vapor chamber heatsink. Well-heeled customers will be relieved to know that ASUS has outfitted the AAA board design with their revamped Q-Release Slim mechanism.

Late last week, the Chinese ASUS branch started to tease a forthcoming ROG and TUF Gaming x Hatsune Miku collection. Eagle-eyed PC gaming hardware watchdogs spotted an intriguing custom GeForce RTX 5060 Ti 8 GB graphics card model within a messy pile of colorful brand-new products. The manufacturer has teamed up with Crypton Future Media; this collaboration is advertised as bringing "a cybernetic fusion to gaming." Initial announcements focused on peripherals, but shortly thereafter the partners unveiled a Hatsune Miku-themed pre-built system. VideoCardz spent its weekend investigating the mysterious new "RTX 5060 Ti O8G" design. A follow-up article pulled information and imagery from Tony Wu's "world premiere" bilibili video—the general manager of ASUS China showcased various upcoming products, including a relevant white triple-fan graphics card.

According to reports, Wu did not disclose an exact model name—NVIDIA and board partners are expected to make new entrants "official" at some point this week. Wu's demonstration unit was visually matched up to the teased "RTX 5060 Ti O8G" card. Additionally, Wu and colleagues presented the entire Hatsune Miku product line during a major press conference; last Saturday (April 12) in Changsha. Many items have already launched to market, through JD.com. Naturally, pricing and availability details were not mentioned during their "GeForce RTX 5060 Ti 8 GB OC" segment. The factory overclocked "RTX 50-series" card's unique white design is outfitted with a single 8-pin power connector; suggesting a tentative entry level price bracket. The backplate is decorated with mascot illustrations—TechPowerUp and other media outlets have noticed an uptick in "cute girl" drawings turning up on housings and retail packaging. ZOTAC joined in on the fun very recently; as reported last week.

The long-promised gaming performance uplift for "Arrow Lake" processors is here, but for now, it is only available in China. Called Intel Performance Optimizations (IPO), this feature aims at system integrators rather than end users and offers a balanced approach between stock settings and full manual overclocking, all while preserving warranty coverage. IPO works by applying optimized profiles that adjust a range of CPU and memory parameters. On the CPU side, it fine‑tunes P‑core and E‑core frequencies, ring‑bus speeds, the UPI interconnect, D2D links between tiles, and both PL1 and PL2 power limits. For RAM, IPO raises transfer rates and tightens timings, pushing modules beyond their factory XMP or EXPO profiles. Early results from Chinese OEM Maxsun show that IPO can boost core clocks by about 200 MHz and elevate DDR5‑8000 kits to DDR5‑8400. In gaming tests provided by Maxsun, this translated to roughly a 10 percent uplift in frame rates, an encouraging figure given Arrow Lake's mild launch performance in gaming.

At debut, some Arrow Lake SKUs trailed their "Raptor Lake" predecessors because the chiplet‑based memory controller introduced extra latency and ring‑bus clocks ran slower. Supply‑chain constraints and tariffs have limited IPO's rollout to China so far, where pre‑built systems from vendors like Maxsun are shipping with the feature enabled. Warranty handling stays with the OEM, so users can enjoy the extra headroom without risking hardware support. Behind the scenes, Intel has been issuing firmware and microcode patches since December, most recently microcode 0x114 paired with CSME firmware 19.0.0.1854v2.2, to improve Arrow Lake's efficiency. IPO represents the next step in that effort and could become a model for global "opt‑in" BIOS presets if the pilot proves successful. For now, China is serving as the testing ground for Intel's latest attempt to squeeze more performance out of Arrow Lake. We have to wait and see if Western markets follow soon.

AMD's Radeon RX 9070 XT and RX 9070 "MBA" graphics cards are no longer "best kept secrets"—as demonstrated recently by the "leaking out" of supposed reference models through black market/back alley channels in China. Late last month, a dual-fan non-XT specimen was snapped up by Chiphell forum member—alleged benchmark results were soon shared within that community. A few days later, a "Made-by-AMD" Radeon RX 9070 XT sample was dissected and compared to Sapphire's PULSE Radeon RX 9070 XT 16 GB SKU. An additional MBA XT example emerged last week, courtesy of another in-depth Chiphell thread. A "bored" enthusiast happened upon a 5499 RMB (~$748 USD) when idly browsing through Xianyu listings (Taobao's Ebay equivalent platform). Their buying experience was described as follows: "(I) found a 'public version' Radeon RX 9070 XT in Tianjin. It was said to be manufactured by (an) OEM, so I bought it without hesitation...I made an appointment to meet today and got it successfully. I don't have to wait for a graphics card anymore."

For unknown reasons, AMD decided to launch its first wave of RDNA 4 gaming graphics cards sans first-party designs. Leaked specimens have attracted much attention in China; with owners bragging about their respective ownerships of reasonably priced rarities. The latest back channel customer expressed satisfaction when inspecting Team Red's all-black flagship Navi 48 GPU-based solution: "after I got it, I have to say that it is not very heavy and is quite light. But the appearance is really what I like." After initial tests, they discovered that VRAM temperatures were not up to snuff—as alluded to (pre-launch) by other Chiphell figures. Their personal DIY improvements were described: "(I) took it apart to measure the thickness of the thermal grease pad. Everyone said the temperature of the video memory was high, so I decided to change to something better....(with) original silicone grease FurMark 2K resolution for 20 minutes: maximum core temperature was 62 degrees, maximum hot spot temperature was 84 degrees, maximum memory temperature was 88 degrees, maximum power was consumption 346 W."

ZOTAC started teasing a refresh of its Apocalypse product line earlier in the year. Two months later, fairly concrete details of the (still) upcoming GeForce RTX 5080 16 GB variant emerged via the NVIDIA board partner's Weibo blog. Unfortunately, ZOTAC's ultra premium 3.5-slot thick/ARGB-lit behemoth design is expected to remain exclusive to the Chinese PC hardware market. Western hardcore gaming enthusiasts are best served by the manufacturer's alternative flagship triple-slotter: GeForce RTX 5080 AMP Extreme INFINITY ULTRA. ZOTAC's mainland China and Hong Kong offices have declared the arrival of brand-new Apocalypse SKUs at retail; utilizing NVIDIA's "Blackwell" GeForce RTX 5080 and RTX 5070 Ti GPUs.

Yesterday's Weibo bulletin commenced with: "when mecha aesthetics collide with technology, and gaming passion merges with extreme performance, the ZOTAC GeForce RTX 50 Apocalypse series graphics cards are born! After (our) continuous R&D, improvement, testing and adjustment—today, newly upgraded flagship graphics cards are officially launched!" The brand has advertised the return of an apparently much-missed product line mascot: "Apocalypse Princess is back with a new look, starting a game/AI exploration journey with you." Promotional imagery and box art feature a prominent illustration of ZOTAC's flagship series heroine—this "mecha artwork" demonstrates a serious sci-fi aesthetic, albeit with a cute female protagonist leading the way. A rival AIB specializes in this type of "marketing"—Yeston's similar-ish presentation language concentrates on enchanting fantasy characters.

Today, the European Commission unveiled its AI Continent Action Plan, establishing a framework to enhance the EU's artificial intelligence computing infrastructure. The plan centers on developing five "AI Gigafactories," each housing approximately 100,000 specialized AI accelerator chips, quadrupling the training throughput capacity of current infrastructure projects. The €20 billion commitment from the EU's InvestAI fund will finance data center construction and semiconductor procurement, supplementing the €10 billion allocated to thirteen smaller AI factories scheduled to become operational by 2026. Site selection remains pending, though Germany's incoming administration under Chancellor designate Friedrich Merz intends to secure a facility within German territory. The Action Plan addresses data infrastructure through Data Labs tasked with standardizing datasets from research institutions and industry partners.

An upcoming Data Union Strategy will establish protocols for cross-border information sharing, creating a unified market for AI-ready data resources across member states. With only 13.5 percent of EU enterprises currently using AI in production environments, the Commission will implement an "Apply AI" initiative focusing on deployment in strategic sectors. This program will utilize European Digital Innovation Hubs to provide implementation support. For talent development, the plan includes fellowship programs, visa pathways for specialized non-EU professionals, and an AI Skills Academy offering training in generative models and machine learning operations. An AI Act Service Desk will provide technical guidance on regulatory compliance. The AI arms race is currently being fought on the front between the US and China, where AI labs are acquiring more accelerators and outputting better models almost weekly. In the EU, the goal is to have AI development on par with the two superpowers, leading to more competition and advancements.

Since its summer 2023 launch, Apple's pricey Vision Pro mixed reality headset has not exactly attracted a mainstream audience. Roughly a year later, rumors of a (then) recently canceled successor appeared online—insiders posited that company engineers had pivoted onto the development of a cheaper alternative model. Vision Pro "Version 1.0" arrived with an intimidating $3499 price tag; thus eliminating interest from a wide swath of potential AR/VR headset enthusiast customers. Industry insiders reckon that Apple had "abruptly reduced production" of the current-gen model last October, with further whispers suggesting a complete cessation of manufacturing activities by the end of 2024. Yesterday, an ITHome article cited compelling claims made by supply chain insiders—the initiation of mass production for a speculated second generation "Apple XR/Vision Pro" device.

The online report stated that: "multiple independent sources (have) confirmed that the panels, shells and other key components of the second-generation Apple XR headset are already in production." Very specific leaked information indicates Lens Technology being the exclusive supplier of glass panel pieces, and Changying Precision tasked with the making of the next-gen model's casing. Additionally, several contract circuit manufacturers are supposedly "rushing to complete orders." Secretive figures posit that Apple will release its sequel mixed reality headset later on in 2025. Differing "expert opinions" have not determined whether this incoming set of fancy goggles will be the predicted "cheaper" model, or a proper "M5 SoC-powered" successor.

Over the past couple of months, Yeston has attracted plenty of media attention—in particular, with its recently launched Sakura Atlantis card design. The Chinese manufacturer produces custom AMD and NVIDIA gaming graphics cards—for its native market—but their unique Radeon RX 9070 Series offerings have made the most noise around the globe. Yeston did unveil GeForce RTX 50 Series Sakura and Game Ace designs around mid-January, but this announcement did not describe the exact nature of upcoming SKUs. As reported by VideoCardz earlier today, the company has distributed "new" GeForce RTX 5080, RTX 5070 Ti, and RTX 5070 models in China—e-tail listings have appeared on JD.com. Their rollout of rather sober looking affairs will disappoint many "waifu" illustration and bright color palette enthusiasts.

Yeston's Deluxe lineup was swiftly identified as a somewhat lazy rebadging of Gainward's GeForce RTX 50 Series Phoenix design. TechPowerUp's W1zzard reviewed the latter's GeForce RTX 5080 Phoenix GS model, just over two months ago. VideoCardz noted that Yeston has a short history of rebranding Gainward products; going back to an almost identical strategy with GeForce RTX 40 "Ada Lovelace" options (Phoenix => Deluxe). Three fan stickers differentiate the board partner brands—everything else is identical; including prominent "Phoenix" text on backplates and on top-mounted ARGB lighting zones. Yeston started to tease its new-gen Deluxe lineup around mid-February; as seen on their Weibo channel. At the time, this official account asked: "is this the card you've been waiting for?" We suspect that the majority of potential customers are preparing credit cards for the purchase of more elaborate options, in the near future.

Lonely City Hardware has discovered an intriguing new ASUS ROG Strix gaming laptop, equipped with a mildly mysterious AMD Ryzen 9 8940HX APU and a familiar NVIDIA GeForce RTX 5070 Ti Mobile 12 GB graphics solution. Potential customers—in China—can part ways with 13999 RMB (~$1907 USD) for the pleasure of ownership. A Weibo bulletin alluded to Team Red's secretive approach with the quiet release of Ryzen 8000HX mobile chipsets. Lonely City Hardware could not find this specific SKU within official company listings/web presences.

According to VideoCardz's interpretation of events, AMD was expected to unveil "Dragon Range Refresh" processors at CES 2025—instead, fresher Ryzen 9000HX "Fire Range" options were prioritized. Going against the grain, MSI allegedly published press material that mentioned a new Ryzen 9 8945HX-powered laptop. Previous leaks have indicated that refreshed "Dragon Range" Zen 4 APUs sport slightly elevated boost frequencies. The Ryzen 9 8940HX can go up to 5.3 GHz; 100 MHz greater than its unmodified sibling; Ryzen 9 7940HX. Team Red's "Dragon Range" processors utilize older RDNA 2 integrated iGPUs, so laptop/notebook manufacturers will likely pair Ryzen 8000HX APUs with the latest-gen discrete graphics solutions.

According to reports from last year, Xiaomi was expected to unveil an oft-rumored proprietary mobile chipset design at some point in 2025. By October 2024, the Chinese technology giant allegedly reached the tape-out phase of its first 3 nm SoC—at the time, insiders posited that Xiaomi was seeking a manufacturing partner. Months earlier, a prototype design was linked to TSMC's 4 nm "N4P" node process—this rumor raised many smartphone watchdog eyebrows. Unlike many other Chinese firms, Xiaomi was reportedly allowed to select a fairly advanced manufacturing process at Taiwan's premier foundry service. In a past weekend news article, Wccftech outlined interesting circumstances: "(US) export controls have yet to affect Xiaomi, which is supposedly on track to launch its first in-house chipset later this year. However, while we reported last year that the company was scheduled to unveil its custom 3 nm SoC in 2025, we were disappointed to learn just the specifications of this version that will utilize TSMC's 'N4P' process. According to more details, this silicon will not sport any homegrown cores like Qualcomm has adopted for the Snapdragon 8 Elite."

Late last week, Jukanlosreve highlighted another leaker's prediction—regarding the technological foundations of Xiaomi's mystery flagship mobile processor. Fixed Focus Digital's Weibo post mentioned the "N4P" node, as well the utilization of current generation Arm Cortex-X925, Cortex-A725 and Cortex-A520 units. A speculated Imagination Technologies "IMG DXT 72-2304" integrated graphics solution is touted to outperform Qualcomm's Adreno 740 iGPU; as featured in their Snapdragon 8 Gen 2 (2022) SoC. As highlighted by Wccftech's report, one of the publication's associates has deemed Fixed Focus Digital to be an unreliable source of inside track info. In response to Jukanlosreve's tweeted question, Mochamad Farido Fanani opined: "that's right, how does Xiaomi use N4P in its new chipset? But this guy always guesses blindly." Older leaks—based on "N4P" rumors—projected performance levels roughly on par with Qualcomm's first generation Snapdragon 8 chip. This model was introduced at the tail end of 2021.

Earlier today, ZOTAC unveiled a special edition spin-off of its relatively new GeForce RTX 5070 X-Gaming 12 GB design. The manufacturer's Weibo channel outlined the very exclusive nature of this MMORPG-themed offering (via machine translation): "watch the finals and get a graphics card! ZOTAC and the first 'Jian Wang 3' competitive group hero joint customized graphics card, fine customization, supports NVIDIA DLSS 4 and Reflex 2 technology, excellent performance. On April 12-13, Nanjing Ledongli Jiangning International E-sports Center, watch the heroes aiming for the top!" According to an ITHome report, the Chinese brand has redecorated its overclocked GeForce RTX 5070 X-Gaming model with Jian Wang 3 characters. ZOTAC's promotional shots showcase a customized backplate, and the same illustration of Jian Wang 3's main cast adorning the card's retail box.

Late last month, Kuroutoshikou Japan introduced a less impressive Blade and Soul NEO-themed rebadge of their standard Radeon RX 7600 SKU—essentially, PowerColor's Fighter RX 7600 design. ITHome noted that ZOTAC has a short-ish history of readying very limited edition designs for past Jian Wang 3 Masters Tournaments. 2024's prize was based on the manufacturer's GeForce RTX 4070 Ti SUPER TRINITY OC White SKU. 2023 finalists were gifted with customized GeForce RTX 4070 Ti X-Gaming OC cards. Both of the previous gen tie-in designs sported Jian Wang 3 graphics or symbols on their shrouds, while the latest entry simply reuses its base SKU's "youthful graffiti" aesthetic.

Zephyr has presented a new Small Form Factor (SFF) graphics card design that houses previous-gen NVIDIA GPU hardware. The Chinese board partner showcased its GeForce RTX 4070-based "Sakura Snow X" via a 1-minute 42-second long teaser video; according to their bilibili channel this compact wonder is: "the world's first mass-produced CNC single-fan graphics card. A small space aesthetic benchmark—a balance of high performance and exquisite size." Zephyr's latest Mini-ITX offering seems to be a more premium—in terms of enclosure materials—alternative to last year's GeForce RTX 4070 12 GB "Sakura Blizzard" model—examples were placed next to each other in the AIB's demonstration video (aluminium versus plastic). This brand-new dual-slot entrant has arrived with a slightly larger cooling solution (see relevant screenshot below), and a fancier 105 mm airflow fan design—outfitted with more blades (11 vs. 9). VideoCardz has observed "Sakura Snow X" first batch listings on an undisclosed Chinese e-tail platform; launch pricing seems to be 4399 RMB (~$602 USD).

Unlike its pink-hued sibling, Zephyr's sober metal design integrates an I/O plate—giving off the impression of a unibody-esque setup. SFF enthusiasts will welcome this truly compact design, albeit with the disadvantage of playing host to old hat "Ada Lovelace" silicon. Throughout early 2025, Team Green and certain manufacturing partners have hyped up various "SFF-Ready" new-gen solutions. Last week, ZOTAC updated its "Blackwell" gaming portfolio with a dual-slot GeForce RTX 5070 Ti SOLID SFF card. This slimmed down triple-fan option is still a lengthy prospect (304.4 mm); thus quickly dismissed as unworthy of its moniker by compact graphics solution connoisseurs. Similarly, GIGABYTE sells 304 mm-long "SFF" variants—launched months ago in WINDFORCE and EAGLE guises.

Late last month, speculative Snapdragon 8 Elite Gen 2 (aka SM8850) technical details emerged online. Up until then, Digital Chat Station's Weibo channel has delivered all sorts of pre-release information—mostly concentrating on Qualcomm's alleged redeployment of a familiar "2 + 6" core configuration, and selection of TSMC's 3 nm "N3P" node process. Earlier today, the veteran leaker predicted a couple of new-gen improvements—first concentrating on the alleged mobile chipset's "Adreno 840" integrated graphics solution. Digital Chat Station (DCS) believes that the company's engineering team has increased their iGPU design's independent cache size "from 12 MB to 16 MB," leading to: "early setting performance increased by 30%±." It is not clear whether this rumored upgrade has affected the SoC's L2 or L3 cache provisions, but the current-gen Snapdragon 8 Elite mobile processor makes do with 12 MB of L2 cache per cluster. Additionally, DCS reckons that an enlargement of caches has resulted in performance of Qualcomm's "second-generation self-developed CPU architecture" increasing "by 25%±."

As interpreted by Wccftech: "the upcoming SoC will now feature 32 MB of L2 cache, with the L3 count currently unknown at this time. The initial test results revealed that the 'Snapdragon 8 Elite Gen 2' delivered a 30 percent performance increase, but it is unconfirmed if this delta exists between the Snapdragon 8 Elite or some other silicon." DCS has heard whispers about the upcoming chip supporting "new generations of high-speed LPDDR5X/LPDDR6" memory. The mentioning of LPDDR5X is not surprising; given that the current Snapdragon 8 Elite model is already up to snuff with this spec. Just over a year ago, JEDEC was reportedly working on the finalization of LPDDR6 standards for mobile platforms. Shortly thereafter, smartphone industry watchdogs started to theorize about the arrival of a so-called "Snapdragon 8 Gen 4" chip with LPDDR6 RAM in 2025. Fast-forward to the present day; fresh reports suggest that manufacturers will have the option to outfit next-gen flagship devices with "bog-standard" LPDDR5X or faster/more efficient LPDDR6 memory.

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.

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.