Major PC OEMs are beginning to secure deals with Chinese memory maker CXMT, as the supply chain has been depleted of available DRAM from the big three memory makers—SK hynix, Samsung, and Micron. According to a report from DigiTimes, CXMT's memory capacity has been booked through the end of 2027, leaving little room for new customers to source DRAM from China's largest memory producer. The report notes that large PC vendors such as Dell, HP, and Apple are the only ones that managed to receive allocations, while smaller brands with less presence in the desktop and laptop market have been left with nearly no allocation.

An important aspect of this situation is the negotiation leverage that larger PC vendors have. As they command more volume than smaller companies, Dell, HP, and Apple are willing to sign bigger contracts with CXMT to secure a larger DRAM supply, even at prices higher than historical levels. For CXMT, this is a positive indication that capacity expansion is well-received and a signal to continue building new fabs. DigiTimes also notes that the United States government has not yet made a firm decision on whether to keep CXMT on the barred entity list, so PC OEMs are taking advantage of the opportunity to stockpile any available DRAM. Future policy changes might impact this, but for now, long-term supply agreements are being made to secure any DDR5/LPDDR5X memory possible.

Dynabook Americas, Inc., a Sharp Company, today announced the Portégé Z40L-P, a premium Copilot+ PC, the second generation of Dynabook's advanced Portégé Z40L platform. Built for the demands of today's AI-enabled workplace, the 14-inch Portégé Z40L-P combines next-generation Intel Core Ultra Series 3 processor options, enhanced on-device AI capabilities, enterprise-grade security, Intel Wi-Fi 7 and user replaceable 56Wh Li-Polymer battery. Pricing for the Portégé Z40L-P starts at $2,170 MSRP.

Expanding the platform beyond premium business productivity, select Portégé Z40L-P configurations feature Intel Core Ultra X9 and X7 processors with Intel Arc Pro B390 graphics. With up to 12 Xe graphics cores, these models deliver discrete-class performance for technical and engineering workflows in a slim mobile workstation built for AI. Purpose-built to accelerate technical drafting, engineering design and computer-aided design applications, the Portégé Z40L-P provides smooth visuals, responsive performance and professional-grade capability in a thin-and-light form factor.

Chinese memory giant CXMT has been expanding its capacity so rapidly that the company is now able to compete with one of the largest memory manufacturers in the world—Micron. According to analysis from Citrini Research and their proprietary data, CXMT is projected to produce about 350,000 DRAM wafers per month by the end of 2026, assuming all current projects are completed on schedule. This is an impressive figure, especially considering that Micron, one of the "big three" DRAM manufacturers, is currently estimated to produce about 375,000 DRAM wafers per month by the end of 2026, assuming their capacity expansions are fulfilled. Micron has been a memory supplier for decades, while CXMT is a relatively new player that emerged to meet China's domestic needs.

A particularly surprising discovery is that CXMT is constructing its cleanrooms—where the actual silicon is produced—in just around 12 months, whereas the rest of the DRAM manufacturing industry typically takes 21-24 months, nearly a full year longer. The biggest challenge for Chinese memory makers like CXMT is acquiring advanced lithography solutions from around the world, as the United States government has imposed sanctions on the company and placed it on a blacklist, preventing access to modern EUV scanners. Instead, CXMT relies on older DUV lithography tools and uses multi-patterning techniques to work around the lack of EUV technology. Despite this, CXMT is matching Micron's output this year with nearly the same number of DRAM wafers, which is impressive for such a new market entrant.

Lenovo has hopped onto the AI trend, launching the diminutive Yoga AI mini PC (via IT Home), a small form factor PC seemingly designed to fit in seamlessly in a minimalist productivity desk setup and deliver both CPU, GPU, and NPU performance for a variety of workloads. The mini PC is powered by the Intel Core Ultra X7 358H CPU, which is a 25 W Panther Lake CPU with four P-Cores, eight E-Cores, and four LP E-Cores, and the Intel Arc B390 iGPU that is more than capable of running a few games when it's not busy with productivity work. The Yoga AI Mini PC also features 64 GB of LPDDR5X memory and a 1 TB PCIe 4.0 SSD. Lenovo claims that the mini PC's specs lend it up to 180 TOPS of AI performance between the CPU, GPU, and NPU, and the memory allows for models with up to 80 billion parameters.

The impressive part of Lenovo's Yoga AI Mini PC is that the minimalist aluminium chassis is just 0.65 L and weighs in at 600 g. It also has built-in ARGB ring on the bottom of the chassis, which casts ambient light onto the desk. The downside of the small chassis is that connectivity is limited. While the wireless connectivity features both Bluetooth and Wi-Fi 7, the physical ports are limited to: 1x 3.5 mm audio combo jack, 2× Thunderbolt 4 ports, 1× USB Type-C 3.2 Gen 2 port, 1× USB PC port, 1x USB 3.2 Gen 2 Type-A port, 1× HDMI (unspecified revision), and 1× 2.5 G RJ45 LAN port. The mini PC is currently only available in China via JD.com at a price of RMB 17,999, which converts to roughly $2,650.

Apple has reportedly been negotiating with the U.S. government to get clearance to use sanctioned Chinese memory maker CXMT in its products. According to a report from the Financial Times, Apple is in talks with the U.S. government to clear the previously blacklisted Chinese entity. The U.S. government had earlier placed Chinese memory maker CXMT on a sanctions list, preventing American companies from purchasing products like the now highly sought-after DRAM. This has created a challenging situation given the severe supply shortage the entire industry is facing. Since the big three—Samsung, SK hynix, and Micron—are ramping up capacity at an insufficient rate, sourcing DRAM from another supplier would help ease the pressure on electronics makers by providing better pricing, more supply, and facilitating product shipments.

CXMT is capable of manufacturing the latest DDR5 and its low-power sibling LPDDR5X memory. Late last year, the company showcased DDR5-8000 and LPDDR5X-10667 DRAM modules that are in mass production. These memory modules are available in 12 Gb and 16 Gb LPDDR5X capacities, while DDR5 scales to 16 Gb and 24 Gb module formats. However, a significant problem remains. Even if Apple gets clearance from the U.S. government to use CXMT memory, the company likely can't satisfy domestic demand first before shipping the quantity that Apple would require. Some estimates suggest CXMT could reach over $50 billion in revenue by 2026, according to SemiAnalysis, indicating that domestic demand is so strong that CXMT is rapidly selling out wafers.

Yesterday, Qualcomm announced its High Bandwidth Compute (HBC) product, a memory-compute solution hybrid designed to replace traditional High Bandwidth Memory (HBM) and offer greater performance, efficiency, and throughput. HBC utilizes LPDDR memory, stacking them across multiple layers in a 3D vertical space and interconnecting them with through-silicon vias (TSVs). This approach provides better energy efficiency than traditional HBM, which stacks regular DDR in vertical layers, as the low-power DDR (LPDDR) chips consume less power while delivering similar bandwidth and capacity. Qualcomm has successfully stacked multiple LPDDR layers on top of each other, with the base of the HBC being a compute die that performs near-memory computation and offloads some processing from the main processor.

This technique is similar to what HBM4 memory currently does, with the base die being a logic die to better integrate compute solutions underneath for packet tracing and data preparation for entering and exiting HBM. However, Qualcomm aims to achieve higher efficiency and more optimized data movement, with a 6x increase in bandwidth per watt compared to the current HBM specification, likely HBM4. HBC Gen 1 achieved 133 TB/s of bandwidth on the AI250 accelerator card, a 18x increase over the LPDDR5X currently used in the AI200 card from Qualcomm. The company also claims there is a viable roadmap for using HBC across many Qualcomm AI accelerators, with HBC Gen 1 shipping with the AI250 AI accelerator in mid-2027. The next-generation HBC Gen 2 is expected to yield further significant gains in bandwidth.

AceMagic on Wednesday launched the Kron Mini K5, a compact mini PC powered by Intel Core 5 320 "Wildcat Lake" mobile processor. This chip features a 2P+4LPE core configuration, with a 4.60 GHz maximum P-core boost frequency, and 6 MB of L3 cache. AceMagic maxed out its cTDP to offer 15 W base power and 35 W maximum turbo power. The chip is paired with 16 GB of LPDDR5-6400 memory, and a 512 GB M.2 Gen 3 x4 NVMe SSD.

Networking options include 802.11ax (Wi-Fi 6) + Bluetooth 5.2; along with a 2.5 GbE wired LAN. Display outputs include one each of HDMI 2.0, DisplayPort 1.4, and USB-C with DisplayPort 1.4 passthrough. Among the USB connections are four 5 Gbps USB 3.2 Gen 1 type-A, two 10 Gbps USB 3.2 Gen 2 type-A, and one 20 Gbps USB 3.2 Gen 2x2, which doubles up as a 100 W power input, although there is also a separate 2-pin DC input jack. The USB-C provides single-cable convenience if you want to pair it with docking stations that also put out USB-PD power to their upstream USB sources.

We previously covered the announcement of the GPD Box, a new Panther Lake mini PC that features an MCIO 8i port and dual USB4 v2.0 ports for expanded connectivity with eGPUs and other high-bandwidth applications. Now, the GPD Box has finally launched on Indiegogo with a starting price of HK$11,377 ($1,452.25 converted). The base configuration is powered by an Intel Core Ultra 7 356H with 32 GB LPDDR5-8533 RAM and a 1 TB M.2 2280 NVMe SSD. The next level up upgrades the CPU to the Core Ultra X7 358H for HK$12,020 ($1,534 converted), and the bundles with the Box mini PC and the G2 MCIO eGPU Dock come in at HK$14,394 ($1,837 converted) for the Core Ultra 7 356H version and HK$15,036 ($1,919 converted) for the X7 358H variant. The biggest difference between the Core Ultra X7 358H and the Ultra 7 356H is the significantly more powerful Arc B390 iGPU in the former—importantly, there is no MCIO connection on the Core Ultra X7 358H version of the Box, leaving it with only USB4 for eGPU connections.

Despite its compact chassis, at just 175 × 134 × 39.5 mm, the GPD Box has decent I/O capabilities, with the highlight of the show being the MCIO 8i port that allows for eGPU expansion with minimal bandwidth loss compared to USB4 and Thunderbolt connections. According to GPD, the maximum bidirectional bandwidth of the GPD Box's MCIO port is 512 Gbps, which effectively matches a native PCIe 5.0 x8 cable connection. Notably, while the G2 eGPU dock does include a PSU with an 800 W ATX 3.1 PSU with two GPU power cables, you will need to bring your own GPU if you buy the bundle with the dock. The G2 dock can also be purchased separately via Indiegogo at a starting price of $385. The GPD Box itself also features dual 2.5 GbE RJ45 ports with link aggregation support. For wireless support, there is Bluetooth 5.3 and Wi-Fi 6E, while the physical I/O includes 2× USB4 V2 Type-C ports, 2× USB 3.2 Gen 2 Type-A, and a 3.5 mm audio combo jack on the front. The rear I/O has a DP 2.1 UHBR20 port, 1× HDMI 2.1 FRL ports, 2× RJ45 ports with 2.5 Gbps speeds, 2× USB 3.2 Gen 2 Type-A ports, and the MCIO 8i port.

Following the unveiling of the new Intel Core processors (Series 3) (formerly Wildcat Lake) in April, AAEON's UP brand has announced its development timeline for four new platforms that will feature the processor line, with both developer boards and edge systems included. Expected to enter mass production in late Q3, 2026, the developer boards included in AAEON's roadmap are the UP WCL and UP Nexus WCL. The first of these is a credit card-sized board consistent with the original UP board's form factor, while the UP Nexus WCL represents a branding shift for the UP Squared Pro form factor, measuring 101.6 mm × 101.6 mm. Both boards will also offer a fully embedded edge system version, the UP WCL Edge and UP Nexus WCL Edge, designed to appeal to customers requiring plug-and-play systems.

All four platforms will be available with the option of Intel Core 7 processor 350 and Intel Core 5 processor 320 CPUs from the Wildcat Lake family, while the UP WCL will also provide SKUs with the Intel Core 3 processor 304. Delivering up to 40 TOPS of AI performance via integrated GPU and NPU but with a CPU prioritizing low power E-cores, the Intel Core series' first integrated heterogeneous compute architecture sees AAEON position its upcoming products as suitable for building advanced applications across market segments, but with the benefit of a lower power footprint than platforms tailored to high-performance computing needs.

MSI revealed the new Intel Arc-powered Claw 8 EX AI+ gaming handhelds shortly ahead of Computex 2026, but pricing was still a mystery at that point in time, with availability and pricing announcements slated for June 23. However, it appears as though MSI has listed the new gaming handheld ahead of the official announcement. According to the apparent premature listing for the Claw 8 EX AI+ (CG3EM-024US) will cost $1,799, which isn't far off the €1,500 predictions from early leaks and rumors.

The MSI online store listing also confirms that the Claw 8 EX AI+ will start shipping on June 23, suggesting that component and memory shortages have not severely hampered the brand's ability to produce at least an initial production run for the launch. Hardware specifications are also unchanged from what MSI showed off at Computex, with the handheld still featuring 32 GB of LPDDR5x-8533 RAM, a 1 TB PCIe SSD, and an 80 Wh battery, alongside an 8-inch, 1080p, 120 Hz display with VRR and 500 nits peak brightness. The MSI store listing confirms that the Claw 8 EX AI+ will launch with Intel's Arc G3 Extreme, and there doesn't appear to be any word on a version of the new MSI handheld powered by the regular Intel Arc G3, which has an Intel Arc B370 iGPU, which has 10 Xe3 cores instead of the 12 found in the G3 Extreme.

As a global leader in the Mini PC industry, GEEKOM continues to deliver reliable computing solutions recognized for outstanding quality and stable performance. Expanding its product portfolio, GEEKOM has officially introduced its first high-performance business laptop—the GeekBook M16. Priced at around $800, the new device combines the flagship Intel Core Ultra 9 processor with up to 24 hours of battery life, setting a new benchmark for productivity laptops by achieving an exceptional balance between performance and endurance.

Powered by the Intel Core Ultra 9 processor, the GeekBook M16 is further enhanced by GEEKOM's proprietary GeekBook Dynamic Tuning Technology (DTT), enabling up to 55 W of performance output while maintaining stable operation under demanding workloads. Paired with high-speed LPDDR5 memory and a Kingston SSD, the laptop delivers fast system responsiveness, efficient multitasking, and smooth large-file processing for both business and professional applications.

A cleverly-named memory brand, AGI, unveiled its latest PC memory products at the 2026 Computex, along with some prebuilt exhibits. We begin our tour with the AGI microSDXC A2 V30 card that's rated for high endurance applications, such as surveillance cameras and FPV drones, with an assured 30 MB/s write speed. The TF138 Supreme Pro is a Switch-compatible microSDXC UHS-I card that comes in 128 GB and 256 GB capacities, with read speeds of 100 MB/s. Next up, is the RD238, a DDR5 RDIMM that comes in 64 GB and 32 GB per-module densities, and speeds of DDR5-6400 and DDR5-5600, JEDEC specifications.

The CA238 is a DDR5 CAMM2 module offering dual-channel capacities of 32 GB and 64 GB; along with speeds of DDR5-7200 and DDR5-6400. The LC238 is an LPCAMM2 that comes in 16 GB and 32 GB densities, with speeds of LPDDR5-6400 and LPDDR5-7200. The UD858 Turbojet is a gamer-grade PC memory with RGB LED lighting. It comes in 16 GB x2 dual-channel kits, with DDR5-6000 speeds. The company also showcased a prebuilt with a high-speed Dan Da Dan variant of the U858 with DDR5-10000 speeds.

From June 2 to 5, COMPUTEX 2026 in Taipei runs under the theme "AI Together." Longsys showcases its full-stack edge AI storage portfolio under "Edge AI Storage - Integrated Implementation," together with its high-end consumer storage brand Lexar, to optimize local LLM performance and accelerate AI ecosystem collaboration.

Global Launch: AIDIMM & AILPBGA for Edge AI Inference
AIDIMM delivers up to 128 GB capacity, 256-bit bus, and 307.2 GB/s bandwidth in a compact form factor. It resolves memory, thermal, and upgrade pain points of AI agent hosts—a single module reliably and stably runs edge LLMs with over 70 billion parameters. Dynamic voltage scaling (0.9 V-1.05 V) and FDVFS power tuning enable high-performance, low-power mass deployment.

AMD today released Ryzen AI Halo, a compact AI development computer designed by AMD to serve as a nearly full-stack machine for AI development, inferencing, and generative AI content creation. Ryzen AI Halo is designed to strike a unique balance of price, performance, and capabilities that rival Apple Mac Mini and NVIDIA DGX Spark. It is a compact computer that can be used as a standalone PC, or in small clusters talking to your workstation. An advantage it offers over DGX Spark and Mac Mini is that it's based on x86-64, and fully supports Windows. AMD is pricing the Ryzen AI Halo at $3,999. Pre-orders open in June 2026.

At the heart of the Ryzen AI Halo is the Ryzen AI Max+ 395 "Strix Halo" processor that combines 16 "Zen 5" CPU cores with an oversized iGPU powered by the RDNA 3.5 graphics architecture, featuring compute capabilities. There's also a nimble 50 TOPS NPU. Perhaps the most interesting aspect that brings this whole device together, is memory. The processor features a quad-channel (256-bit wide) LPDDR5x memory interface, and AMD has crammed 128 GB of memory into this box. Also helping things is a spacious 2 TB NVMe SSD. The large memory size, helps the processor run AI models with up to 200B parameters, with the right quantization. For reference, the current Mac Mini can only be configured with up to 64 GB of memory. Hardware is less than half the story, the bulk of AMD's engineering effort with the Ryzen AI Halo is on the software side. The company innovated a comprehensive software platform compatible with both Windows and Linux, which helps developers quickly deploy one or more Ryzen AI Halo boxes faster than it takes to set up DGX Spark, AMD says.

China appears to be the first in line to receive laptops powered by Intel's "Wildcat Lake" Core 300 series processors, and the initial models have already arrived. The most affordable option is the CHUWI UniBook, which features a five-core Intel Core 3 304 processor. This is the lowest in the series and the only five-core "Wildcat Lake" processor available. The laptop comes with 8 GB of LPDDR5X memory running at up to 7,467 MT/s and 256 GB of SSD storage. All of this is housed in a 14-inch body with a 1920x1200 IPS display. Interestingly, this model poses a direct challenge to Apple's MacBook Neo, operating on Windows 11 Pro and a different x86 platform.

Additionally, other laptop manufacturers have started listing their "Wildcat Lake" models on the Chinese retailer JD, including ASUS, HP, and Honor. These offerings include the ASUS Fearless 14SE 2026, HP OmniBook 3, and Honor Notebook X14 2026 Combat Edition. All three models are powered by the Intel Core 5 320 processor, which features a six-core design. They also come with 16 GB of LPDDR5X memory and 512 GB of SSD storage. The prices for these models are $597 for the ASUS, $571 for the Honor, and $662 for the HP OmniBook 3. These laptops offer twice the storage and memory compared to the popular Apple MacBook Neo, aiming to surpass Apple in the entry-level PC market. Interestingly, sales tax is not included in these prices.

Surging mobile DRAM contract prices in 2Q26 continue to place even greater cost pressure on smartphone brands, according to TrendForce's latest memory market research. The two major Korean suppliers may have adopted different pricing strategies: Samsung Electronics is pursuing a more aggressive one-step pricing approach with sharper increases, while preliminary quotations from SK hynix indicate relatively milder hikes through a gradual price-raising strategy, with final pricing expected to be completed in late May. Overall, TrendForce estimates that the ASP of LPDDR4X solutions will jump by at least 70-75% QoQ for 2Q26, while the ASP of LPDDR5X solutions will see a QoQ surge of 78-83%.

Notably, several consecutive quarters of steep price increases have significantly intensified cost burdens for smartphone vendors. Under such conditions, brands are expected not only to reduce total smartphone production in 2026, but may also struggle to fulfill the bit procurement volumes stipulated in the LTAs (or MoUs) arranged with suppliers at the end of last year.

Microsoft has so far been mum on the prospect of new Surface hardware featuring Intel's Panther Lake CPUs, although leakers have started reporting that that's exactly what is on the horizon, mere weeks ahead of the supposed launch date. Now, we have two Geekbench listings backing up those early leak and revealing two versions of an upcoming Microsoft Surface Pro 12 and a new Surface Laptop 8 (via Notebookcheck). The two Surface Pro 12 variants, which will be thin-and-light convertible devices, be available with both the Intel Core Ultra 5 325 and Core Ultra 5 335 CPUs, which are both eight-core chips with a small 100 MHz difference in clock speed. It also looks as though the Pro 12 will be available with up to 32 GB of memory.

According to the leaks, the new Surface Laptop 8 will follow the standard 13-inch laptop form factor we've come to expect from Microsoft and come with both the Intel Core Ultra X7 368H, with Intel's Arc B390 iGPU, and the Core Ultra 7 366H CPU, which features the much less powerful four-core iGPU. The now-removed Geekbench listing lists a single-core score of 2,833 points and a multicore score of 16,368 points, and the new Laptop 8 will feature 32 GB of LPDDR5-8532 memory. The same spec options are expected for the 15-inch version of the 8th-gen Surface Laptop 13. An exact release date has not yet been declared, but a mid-May launch seems likely, according to WinFuture.

JEDEC Solid State Technology Association, the global leader in the development of standards for the microelectronics industry, today previewed a set of new features planned for incorporation into the next version of its JESD209-6 LPDDR6 standard. Building on the foundational JESD209-6 published in July 2025, JEDEC's JC-42.6 Subcommittee has been working to enhance the next version of the standard to extend LPDDR6 beyond mobile platforms to support selected data center and accelerated computing workloads seeking a power-efficient, high-capacity memory platform.

Planned features for the upcoming LPDDR6 update include:

• Narrower per-die interface (x6) enables higher capacities: With the move to a non-binary interface width - from x16 to x24, the inclusion of x12 and an additional x6 sub-channel mode, allows more die per package and higher memory capacities per component and per channel, a critical enabler for AI-scale memory footprints.
• Flexible metadata carve-out intended to minimize impact to peak data throughput, giving data center customers the option to balance user capacity and metadata needs according to their specific reliability requirements.
• 512 GB density on the horizon: LPDDR6 is expected to unlock densities beyond the current LPDDR5/5X maximum, a capability designed to address the ever-growing memory capacity requirements of AI training and inference workloads.
• LPDDR6 SOCAMM2 module standard in development: JEDEC is actively working on an LPDDR6-based SOCAMM2 module standard, which is being designed to carry the compact, serviceable module form factor forward and offer a clear upgrade path from today's LPDDR5X SOCAMM2 modules.

Intel has officially unveiled its "Wildcat Lake" Core 300 series of processors for the entry-level PC segment. These CPUs are designed for value-oriented buyers looking for good CPU performance and basic GPU output in small form-factor commercial and edge AI PCs. Officially rated at 40 TOPS, these processors are capable of Copilot+ AI PC certification, meaning Intel has managed to provide basic local AI processing functionality to entry-level buyers. As part of the Core 300 series non-Ultra family, Intel's main selling point is all-day battery life in laptops with local AI processing. There are three segments, each featuring a hybrid core configuration, pairing two "Cougar Cove" P-cores with four LPE "Darkmont" cores.

The actual SoC package integrates two dies. The first and most important one, built on the 18A internal Intel node, features a 6-core CPU configuration, NPU 5 with 40 TOPS of INT8 data, Xe display and media engine, and a GPU that includes up to two Xe3 cores. This die also contains the memory controller and cache pool, supporting LPDDR5X memory running at 7,467 MT/s or up to 6,400 MT/s for DDR5. The memory-side cache includes 4 MB to help tasks "buffer" into this cache before and after accessing memory. Intel dedicates the second die to I/O handling, as the platform supports 6 PCIe Gen 4 lanes, two Thunderbolt 4 connections, two USB 3.2 connectors, and up to eight USB 2.0 connectors. For Wi-Fi, there is the Wi-Fi 7 standard, and Intel also includes Bluetooth 6.0.

Elon Musk announced today on X that Tesla has completed the design of a new generation of its AI chips for Full Self-Driving (FSD). The CEO also mentioned earlier that the AI5 chip will offer performance comparable to NVIDIA's "Hopper" architecture, with two AI5 units matching the power of a single "Blackwell" processor. In late 2025, reports indicated that Samsung had achieved a significant win for its previously struggling foundry business, as Tesla decided to split the manufacturing of its new AI5 accelerator between Samsung and TSMC. The chips will be produced at Samsung's plant in Taylor, Texas, and TSMC's facility in Arizona. This decision is part of a strategy to keep the supply chain diversified and maintain chip supply under control for any demand scenario.

Samsung and TSMC are not the only partners in this project. Tesla is also sourcing DRAM chips from SK hynix, which appear to be LPDDR5X memory integrated into the package. On both the left and right sides, there are two rows of SK hynix LPDDR5X memory modules, each with three modules. This totals 12 LPDDR5X memory modules per AI5 chip. With 16 GB per module, this results in 192 GB of LPDDR5X memory per single AI5 SoC.

The latest in a recent flurry of AMD Ryzen AI Max+ 395 mini PCs is OneXPlayer's ONEXStation—a new AI mini PC that should be just as capable at everything from gaming to workstation and creative workflows. Aside from the already known benefits of the massive AMD APU, with its Radeon 8060S iGPU and 128 GB of LPDDR5X memory on a 256-bit bus, the new mini PC boasts a decent selection of I/O ports in an impressively compact aluminium chassis, measuring in at just 186 × 193 × 62 mm. By default, the ONEXStation will be equipped with a single 1 TB PCIe Gen 4×4 M.2 SSD, but there is a second M.2 slot easily accessible for up to 8 TB total storage.

With 13 ports in total, the ONEXStation's front I/O features dual USB 3.2 Gen 2 ports, a USB4 port, a UHS-II SD card slot, and a 3.5 mm audio combo jack. There's also a turbo button that presumably boosts the fan speed for improved thermal capabilities at the cost of noise. Around back, the ONEXStation has a DC power barrel plug, a 3.5 mm audio combo jack, a 2.5 GbE RJ45 port, one USB 3.2 Type-A port, a USB4 Type-C port, one DisplayPort, an HDMI port, and dual USB 2.0 ports—the latter presumably for peripherals. It also features both Wi-Fi 7 and Bluetooth 5.4 onboard. Cooling is handled by a custom solution with three heatpipes and a tri-fan ventilation system, and OneXPlayer says the TDP can be configured to 55 W, 85 W, or 120 W from within the mini PC's software. The launch price is $2,999, but the MSRP will seemingly increase to $3,599 after the launch offer. It's worth noting that, due to current market conditions, pricing and availability of systems like those based on Strix Halo and similarly powerful SoCs, like Intel's recent Panther Lake chips, is highly volatile, with Ayaneo already having suspended pre-orders for its Strix Halo-powered Next 2 handheld and Khadas having recently bumped up the price of its Panther Lake mini PC.

NVIDIA's upcoming N1 SoC, designed for gaming laptops, has officially appeared for the first time alongside its motherboard intended for a laptop. Featuring 128 GB of LPDDR5X memory running at 8,533 MT/s, the SoC is paired with SK hynix modules. NVIDIA's OEM partners for this launch, including Dell and Lenovo, have confirmed that the new NVIDIA laptop platform will be available this year, and we have seen early benchmark submissions of the N1 SoC. As NVIDIA prepares to enter the gaming laptop market, the company initially released its GB10 SoC for the DGX Spark mini workstation. This chip allowed NVIDIA to test its collaboration with MediaTek on developing new low-power SoCs that offer unusually high-performance graphics for AI applications, as well as very high RAM capacity.

In the latest images that appeared on Goofish, we see a laptop motherboard equipped with the NVIDIA N1 SoC and 128 GB of LPDDR5X memory closely packed in a specialized PCB area. This area might be designated for the cooler, which could potentially also cool the LPDDR5X memory modules. NVIDIA plans to introduce its N1/N1X SoCs with Arm-based CPU cores around Computex 2026. These will offer consumers 20 CPU cores, consisting of 10 Cortex-X925 and 10 Cortex-A725 CPU cores based on the Armv9.2 ISA, along with a "Blackwell" GPU optimized for low power settings with up to 6,144 CUDA cores in the higher-end form. You can also see the port selection on the side of the PCB, which seems to be USB-C or Thunderbolt 4/5, USB-A, and an HDMI connection. Since we don't know which laptop model this is, we have to wait a bit more before getting final details.

The Apple MacBook Neo has been a bit of a shock to system in the laptop world, thanks to its low barrier to entry and surprisingly solid performance for everyday tasks. However, one of the biggest criticisms levelled against the MacBook Neo is its mobile A18 Pro chip and limited 8 GB of RAM. According to recent rumors, though, Apple may upgrade the internals in the MacBook Neo in 2027, bumping it up to an A19 Pro, which is the SoC featured in the latest iPhone 17 Pro smartphones.

The A19 Pro is still a mobile SoC, and it is slated to use the same 5-core GPU configuration as the A18 Pro in the current MacBook Neo, but that SoC upgrade also means that the MacBook Neo would no longer be limited to just 8 GB of LPDDR5X. Instead, as the rumors suggest, the MacBook Neo 2027 will see a 50% memory increase to 12 GB total memory. So far, the MacBook Neo seems to have been such a hit in sales that the Cupertino giant has already started increasing production of the laptop.

Late last week, we reported on a new series of rowhammer bit-flip attacks targeting GDDR6-based NVIDIA GPUs. Most of these attacks can be mitigated by enabling IOMMU through the BIOS, which restricts the memory regions the GPU can access on the host system, thereby closing the primary attack path. However, researchers from the University of Toronto have introduced "GPUBreach," which can bypass IOMMU and enable CPU-side privilege escalation, unlike the previous "GDDRHammer" and "GeForge" attacks. In most typical server, workstation, and even PC configurations, IOMMU restricts the GPU's access to the CPU's physical addresses, preventing direct memory access. These are the typical DMA-based attacks that the Input-Output Memory Management Unit protects users from. However, the new "GPUBreach" operates differently.

For example, "GPUBreach" exploits memory-safe bugs in the actual GPU driver and corrupts them. When IOMMU confines the GPU's direct memory access to driver-assigned buffers, the new exploit corrupts metadata within these permitted buffers. This causes the driver, which has kernel privileges enabled on the CPU host, to perform out-of-band writes to the buffer, effectively bypassing any protection IOMMU can offer. This logic is built into the kernel by default, as the GPU driver is one of the most trusted components of the operating system. Hence, IOMMU bypass is possible when the metadata is corrupted. Since "GPUBreach" grants an attacker full root privilege escalation, the attack differs significantly from previous rowhammer attacks.

Before its global rollout, Lenovo first launched the Yoga Mini i mini PC in China, a few months after its introduction at CES 2026. The Yoga Mini i is built around the Intel Panther Lake platform, with configurations listed up to a Core Ultra X7 385H processor at a 45 W TDP. Graphics are handled by integrated Intel Arc B-series GPUs, with the top configuration reaching Arc B390. The system also includes an NPU rated at up to 50 TOPS, aligning it with Microsoft Copilot+ PC requirements. Memory goes up to 32 GB of LPDDR5x, paired with up to 2 TB of PCIe Gen 4 storage. Despite its compact footprint, measuring 130 x 130 x 48.5 mm and weighing around 600 g, the mini PC offers a relatively complete I/O setup. This includes multiple USB-C ports with Thunderbolt 4 and DisplayPort support, HDMI 2.1, USB-A, and 2.5 GbE, Wi-Fi 7 and Bluetooth 6. The system integrates basic audio hardware with a 2 W built-in speaker and dual microphones. Security features include a fingerprint reader built into the power button, Human Presence Detection, and Walk Away Lock. Power comes from a 100 W adapter.

At this moment, Lenovo is only offering a lower-tier configuration in China equipped with an Intel Core Ultra 5 325 processor with 16 GB of RAM and a 512 GB SSD. This model is priced at CNY 5,499 (around $800), indicating that earlier references to a $699 starting price will likely apply to similar entry-level SKUs rather than higher-end Core Ultra X7 variants. Lenovo still lists the Yoga Mini i as "coming soon" in other regions, with a broader rollout expected later this year, presumably even before this year's Computex, which is held on June 2-5.