All published memory-market news.

Phison's Computex 2026 exhibit included the X3 controller, a PCIe 6.0 SSD platform that the company says can reach 28 GB/s of sequential bandwidth and 6.8 million IOPS in random workloads. The controller is also described as supporting drives as large as 2 PB, which places the design squarely in the high-capacity enterprise storage discussion. The company also highlighted the E37T for PCIe 5.0 SSDs, with a stated 4.5 W power profile. That lower-power angle matters for client and compact systems where PCIe 5.0 storage has sometimes been constrained by thermals and efficiency. For the memory market, the announcement is more of a technology roadmap signal than a direct pricing event. It keeps pressure on SSD controller vendors to prepare for PCIe 6.0 while also improving PCIe 5.0 efficiency, but the available report does not include shipment timing, controller pricing, NAND commitments, or customer design wins.

Apacer unveiled GraTherX at Computex 2026 as a memory cooling technology designed for industrial DDR5 modules operating in fanless and space-constrained environments. The company is positioning the product around a practical issue in newer memory deployments: higher thermal density and power draw can undermine stability as DDR5 speeds increase. For RamTrend, the announcement is relevant because it highlights how memory performance gains are creating secondary demand for better thermal engineering, especially in AI-adjacent systems and embedded hardware where airflow is limited. While this does not directly change DRAM supply, it reinforces the view that higher-spec DDR5 deployments are pushing system-level design requirements upward.

Tom's Hardware reports that Intel discussed memory-price pressure at Computex 2026 and said it will keep products in the market that can support older memory technologies. The report specifically connects that position to Raptor Lake and DDR4 platforms. The market signal is that high memory prices are influencing platform strategy, not just component purchasing. DDR4 support can help PC buyers and system vendors manage cost when newer memory remains expensive, while also extending demand for older DRAM generations. This does not resolve the pricing issue. Instead, it reinforces that memory costs are high enough to affect product planning and platform positioning across the PC market.

Samsung said at Computex 2026 that it is preparing a broader AI memory roadmap centered on HBM5, thermal management and advanced packaging. The company also pointed to closer alignment across its memory, foundry, logic and packaging activities as it gets ready for future AI platforms. For the memory market, the announcement reinforces that competition in AI memory is moving beyond chip density alone. Suppliers increasingly need to coordinate memory design, heat management and packaging technology to deliver usable bandwidth at system scale. The update does not provide shipment targets or pricing, but it signals that Samsung is positioning for the next wave of premium AI memory demand.

SK hynix's newsroom item says Chey Tae-won and Jensen Huang met at Computex Taipei 2026 and toured the SK hynix exhibition booth. The available content highlights AI memory technologies and a collaborative product display connected to the two companies' work. The item does not provide new product specifications, capacity commitments, or pricing details. Its market value is mainly as relationship and demand context: SK hynix continues to frame its memory roadmap around AI infrastructure, while Nvidia remains the most visible demand driver for advanced AI memory. For RamTrend, this should be treated as a soft AI memory signal, not a standalone pricing catalyst. It reinforces the strategic alignment between a leading AI accelerator vendor and a leading AI memory supplier, but does not change supply-demand assumptions by itself.

Lexar announced a collaboration with ASUS around next-generation AI client devices. The company plans to show AI-oriented storage with an ASUS NUC 15 Pro Mini PC and says local AI workloads are turning storage into a performance-sensitive part of model loading, data movement, latency, and sustained operation. The available report also mentions future AI storage format discussions, including Storage Stick concepts, along with handheld gaming and high-performance gaming memory positioning. For RamTrend, this is another signal that client storage vendors are trying to attach SSD value to AI PC use cases rather than treating SSDs only as capacity devices. It supports the narrative that local AI could raise expectations for faster and more specialized client SSDs, though no shipments, pricing, or bill-of-material changes are provided.

Phison's Computex display highlighted two important controller directions. The X3 series is an eight-channel PCIe Gen6 SSD controller with NVMe 2.3 support and computational storage capabilities. The report says Gen6 SSDs based on X3 would typically target roughly double current Gen5 sequential transfer rates. For near-term client SSDs, the E37T is positioned as a DRAMless PCIe Gen5 controller that can reach performance levels usually associated with higher-end controllers using DRAM cache. Phison also showed reference SSDs and new Pascari-branded enterprise SSDs. The market angle is that controller improvements can keep high-speed SSD performance moving even in DRAMless designs, while Gen6 prepares the enterprise and enthusiast storage roadmap beyond the current platform cycle. This is more of a technology roadmap signal than an immediate NAND pricing catalyst.

Patriot's Computex display focused on Viper Gaming Steel 5 Infinite RGB DDR5 memory. The lineup in the available report includes DDR5-6000, DDR5-6400, and DDR5-8000 speed bins, with kit capacities of 32 GB, 48 GB, and 64 GB across two modules. The company also showed portable storage accessories and USB flash drives, including an M.2 NVMe enclosure with 40 Gbps USB4 support and flash drives using USB-C and high-speed USB interfaces. For the memory market, this is a consumer-channel product signal rather than a supply-side event. It shows that DDR5 vendors continue to push both speed and capacity options in gaming memory, with 2x24 GB and 2x32 GB kit formats remaining visible alongside standard 2x16 GB kits.

Biwin's Computex lineup spans both PC memory and storage. The company showed Origin Code enthusiast DDR5, including a Vortex quad-rank module designed to increase module capacity by using four ranks across a single module's two sub-channels. The booth also included a DDR5-6200 CL28 demonstration on an AMD and ASUS platform, plus DW100 DDR5 memory at DDR5-6000 speeds. On the storage side, Biwin showed the M560 Gen5 NVMe SSD in 1 TB and 2 TB capacities using a DRAMless controller. It also showed the BL130 single-chip SSD for gaming handhelds, with 512 GB, 1 TB, and 2 TB capacity points over a PCIe Gen5 x2 interface. For RamTrend, the clearest signal is product segmentation. Biwin is addressing high-capacity enthusiast DDR5, mainstream DDR5, high-speed DRAMless client SSDs, and ultra-compact handheld storage. This supports ongoing competition in consumer memory and SSD channels, but the article does not provide volume, pricing, or availability details.

South Korea's industry ministry announced a regulatory change for extreme ultraviolet equipment used in semiconductor fabrication. The available DigiTimes summary says the approval path will be shortened by reclassifying EUV tools under a new equipment category, cutting installation lead times from 34 days to 9 days. The item matters for memory because Samsung and SK Hynix are named in the expansion context, and both rely on advanced process tooling for leading-edge DRAM and broader semiconductor production. Faster tool approvals can reduce friction in capacity buildouts, even if the article does not specify which fabs, nodes, or memory products will benefit first. The market effect is more medium-term than immediate. Smoother EUV installation can support advanced production expansion, which may eventually ease supply constraints in high-end memory, but it does not change near-term pricing by itself.

Phison Electronics announced a collaboration with Intel around aiDAPTIV memory extension technology for Intel AI PC platforms running Core Ultra processors. The available summary says the effort is intended to enable larger on-device AI workloads and reduce dependence on cloud execution. The work also includes OpenVINO toolkit support, ISV validation, technology demonstrations, and development of workloads optimized for local execution. For RamTrend, the key angle is the attempt to stretch AI PC memory resources through a platform-level extension approach rather than relying only on raw DRAM capacity increases. This is an early ecosystem and software-enablement signal, not a direct memory pricing event. If local AI PC workloads become more practical, they could still support demand for systems with higher DRAM and fast local storage configurations over time.

DigiTimes reports that SK Hynix is preparing a new 200-layer-class NAND generation at its Dalian Phase 2 site in China. The technology is described as using floating-gate architecture, with the target market focused on enterprise SSDs for AI data centers. The move is strategically important because AI storage demand is becoming a clearer NAND growth driver alongside HBM-led DRAM strength. Enterprise SSDs require high-capacity NAND, controller integration, validation, and sustained supply, so a Dalian ramp would strengthen SK Hynix's ability to compete in the AI infrastructure storage cycle. The pricing effect is mixed. New capacity can eventually add supply, but the reason for the ramp is strong eSSD demand. Near term, the article reinforces a constructive demand signal for NAND and enterprise SSDs rather than an immediate easing signal.

At COMPUTEX 2026 in Taipei, SK hynix centered its booth around AI-focused memory and storage products, with the strongest attention on its high-bandwidth memory lineup. The company displayed 12-layer 36GB HBM3E, 16-layer 48GB HBM4 and 12-layer 48GB HBM4E alongside server DRAM products such as a 256GB 3DS RDIMM, a 64GB RDIMM built on its sixth-generation 1c 10nm-class process, and a 128GB DDR5 MRDIMM aimed at AI workloads. The portfolio also included enterprise SSDs and several forward-looking products tied to AI infrastructure, including High Bandwidth Flash, ZUFS 4.1, LPCAMM2 and a CXL-based CMM-DDR5 concept. For RamTrend, the announcement matters less as an immediate supply event and more as a signal of where vendor investment and customer attention are concentrating: high-bandwidth DRAM, higher-capacity server memory, and storage products designed for AI systems. The presentation supports the view that AI-driven demand remains a major force shaping premium memory product development, though the article does not provide new pricing, shipment or capacity guidance.

StorageReview's review frames the Radeon RX 9070 GRE as an upper-mainstream RDNA 4 card that sits below the RX 9070 and targets 1440p gaming rather than flagship performance. For memory-market tracking, the relevant point is the continued use of a 12GB graphics-memory configuration in a globally released midrange-to-upper-mainstream GPU. The review explicitly notes that buyers in this segment are weighing whether 12GB of memory will remain sufficient for newer games over the next few years. This is not a broad pricing catalyst on its own. It is a product-level signal that graphics memory capacity remains a visible buying criterion in consumer GPUs, especially as AI-assisted rendering, higher-refresh displays, and modern game workloads become part of mainstream purchase decisions.

China's softer holiday demand has not been enough to break the upward trend in memory pricing, according to the available DigiTimes summary. The report says spot prices recovered modestly by the end of May after sharp April and May swings, while the industry's tight supply-demand balance remains in place. The key pricing point is the expected path for contract prices. DigiTimes says third-quarter memory contract price increases should slow from a high base, but still remain on an upward trajectory, with likely quarterly gains in the double-digit 10% to 20% range. For RamTrend, this is a direct pricing signal. It suggests consumer demand pressure in China may cap the pace of hikes, but it has not removed the supply tightness that is supporting DRAM and broader memory contract pricing into Q3 2026.

Yangtze Memory Technology is using Taiwan as the starting point for a broader Asia-Pacific retail SSD push, with Singapore and South Korea named as later targets in the available report summary. The market context is more important than the geography alone. The source describes memory supply as tight and prices as still elevated, while AI, gaming, and PC demand continue to shape storage conditions. That combination makes YMTC's regional retail expansion a notable competitive signal for consumer SSD channels. For pricing, the article points to a market where new retail availability does not automatically imply lower prices. If demand from AI infrastructure, gaming systems, and PCs continues to absorb NAND and SSD supply, YMTC's entry may add channel competition while the broader supply-demand balance remains supportive of firm SSD pricing.

Micron's announcement centers on a broad AI hardware portfolio rather than a single product launch. The company highlighted HBM, SOCAMM2, high-capacity DDR5 RDIMMs, data center SSDs, LPCAMM2, GDDR7, LPDDR5X, client SSDs, and UFS for automotive applications. For the memory market, the important signal is the breadth of demand Micron is trying to address. The named products cover AI accelerators, server memory, graphics memory, notebook memory modules, client storage, mobile-class low-power DRAM, and automotive storage. That reinforces the view that AI-related demand is no longer limited to HBM capacity, even though HBM remains the highest-profile segment. The release does not provide shipment volumes, pricing, customer names, or capacity allocations. Its near-term pricing read is therefore directional rather than precise: Micron is emphasizing active production across multiple premium memory and storage categories, which supports a constructive demand backdrop for advanced DRAM and SSD products.

EE Times Asia reports that Rambus launched a DDR5-9600 Client Memory Module Chipset for high-performance clocked client modules. The chipset includes a Gen2 client clock driver, PMIC, and SPD hub for modules operating up to 9600 MT/s. For RamTrend, this is a client-memory technology signal. AI PCs are pushing higher bandwidth and capacity requirements, while DDR5 speeds above mainstream levels need more module-level signal and power management. Rambus is positioning its chipset to support the transition toward CUDIMM and CSODIMM architectures in desktops, notebooks, and workstations.

EE Times Asia reports that Counterpoint tracked an 80% sequential jump in global DRAM revenue during 1Q 2026, reaching a new high. The report links the increase to AI data-center demand, record memory pricing, HBM, and rising LPDDR5 content. The supplier details matter for RamTrend. Samsung is reported at 38% share, with SK Hynix behind it, while Micron held share and CXMT's revenue grew sharply as China demand expanded. Counterpoint also expects DRAM average prices to rise further in 2Q 2026, including both HBM and commodity DRAM. This is one of the clearest positive pricing signals in the current queue.

DigiTimes says Nvidia has continued to support TSMC's higher pricing for advanced process and supply-chain services, while major semiconductor suppliers broaden collaboration around mature-node foundry services and advanced packaging. For RamTrend, the story matters because AI memory is increasingly tied to packaging capacity and cross-company execution. HBM stacks, logic die, and advanced substrates all depend on tight integration between memory suppliers, foundries, and system companies. The article is not a spot-price update, but it reinforces why packaging constraints can remain price-supportive for AI-linked memory.