A US research firm reported that Nvidia's next-generation Vera Rubin platform will consume more low-power DRAM (LPDDR) in 2027 than the combined global usage of Apple and Samsung, the two largest smartphone brands. According to a Citrini Research estimate cited by Wccftech, Rubin is expected to require more than 6 billion gigabytes of LPDDR in 2027, outstripping Apple's 2.966 billion gigabytes and Samsung's 2.724 billion gigabytes combined.

AI-driven memory demand has sent NAND prices soaring more than 20-fold, but flash memory and hard drive makers alike remain cautious about large-scale capacity expansion. SanDisk argues that the widening price gap between NAND and hard disk drives (HDDs) has further weakened the economic case for solid-state drives (SSDs) to replace HDDs in AI data centers.

Pradeep S. Shenoy, a computing technology expert at Texas Instruments's (TI) US headquarters, recently shared his views on 800V architectures for AI data centers in a media appearance. He said 800V solutions will not see broad adoption in AI data centers in 2026, but rollout will accelerate year by year from 2027.

Taiwan precision motion-control leader Hiwin Technologies is accelerating growth on the back of semiconductor automation and robotics demand, with robot-related products emerging as a rapidly expanding revenue driver as the company deepens its push into wafer-handling systems, logistics robots, and humanoid robot components.

Taiwan's Ministry of Digital Affairs (MODA) has revealed that three applicants have submitted bids for its AI compute center BOO (build-own-operate) program, which closed on May 14, 2026. The disclosure comes as Foxconn is widely reported to be highly interested in building a large AI compute center.

Asus is accelerating its push into the server market in South Korea, leveraging its manufacturing prowess, long-term ties with major memory makers, and flexible design and supply strategies to take advantage of mass procurements in South Korea's government-led sovereign AI infrastructure initiative and GPU server roll-outs at Hyundai Motor, telecom operators, and financial institutions in the country.

Kioxia Holdings is moving to capture a larger share of AI-related storage demand, as the shift from AI training to inference raises interest in high-speed, high-capacity NAND flash and solid-state drives.

Global server markets may shift as DIGITIMES analyst Luke Lin says Intel's revenue gains stem largely from price rises while AMD posts stronger shipment-led growth. TSMC plans another price increase as customers prioritize capacity over cost, developments that could affect cloud providers, vendors, and data center economics worldwide.

Ennoconn Technology held a quarterly investor briefing, during which executives said the first-quarter results were weighed down by deferred, high-margin orders for intelligent software and solutions, memory price fluctuations, and one-time costs tied to integrating Kontron, the German subsidiary. The firm reported consolidated first-quarter revenue of NT$37.11 billion (US$1.17 billion), a gross margin of 20.1%, operating profit of NT$1.99 billion, and net income attributable to the parent of NT$750 million, with earnings per share of NT$5.15.

Nvidia's GB300 is becoming the mainstream AI server product in the first quarter of 2026, boosting ODM revenue and profit worldwide; with Vera Rubin servers due to ramp in the third quarter and aggressive ASIC deployments by cloud providers, the global AI server supply chain expects sustained growth into the second half of 2026.

Adlink Technology said on May 13 that edge AI demand continued to rise, helping the industrial PC maker deliver one of its strongest profit performances in recent quarters in the first quarter of 2026. The company struck a cautiously optimistic tone for the second half and said long-term growth will center on expanding AI deployment across more real-world use cases.

Advantech is deepening the integration of hardware and software for edge AI and is moving to unify its global partner conference with exhibition activities at COMPUTEX Taipei 2026. The company said the reorganization and new strategic priorities are designed to accelerate the industrial deployment of physical AI.

Growing demand for AI is driving appetite for server components across cable connectors, cases, and rail kits. This wave has led most major Taiwanese players in the industry to record rising annual revenues, with some exceptions. Rail kit makers in particular have posted strong annual growth as AI servers claim a larger share of their business.

Taiwan-based Tripartite Therapeutics, Inc. (TTI) is positioning itself at the forefront of the global antibody-drug conjugate (ADC) race with a proprietary technology platform aimed at overcoming long-standing barriers in targeted cancer therapies. Ahead of its listing on the Taipei Exchange's Emerging Stock Board on May 20, the company says its integrated linker and payload technologies could help international drugmakers develop safer, more effective, and commercially scalable ADC treatments.

The global optical fiber market is shifting from a traditional telecom cycle into a new phase of structural shortage, driven by rapid AI infrastructure expansion and dense cloud data center construction. Prices and capacity are rising together as demand for fiber, a core material for data transmission, keeps climbing.

As AI server prices rise, original design manufacturers (ODMs) are hitting record revenue, but also coming under pressure on their gross margins. To defend profitability, manufacturers are cutting costs, negotiating with customers to switch from buy-sell procurement to consignment, and chasing ASIC and general-purpose server orders.

Biwin Storage Technology recently resubmitted its listing application to the Hong Kong Stock Exchange again, with Huatai Financial Holdings acting as the sole sponsor. The Chinese memory module maker has seen a rebound in profit as demand for memory continues to rise.

AAEON Technology reported sustained demand growth and a rising book-to-bill ratio, signaling stronger momentum into the second quarter and second half of 2026. The industrial PC maker expects double-digit growth for the full year, powered increasingly by artificial intelligence across its hardware and ecosystem strategy.

Market fears over Nvidia's next-generation Vera Rubin platform have eased after supply-chain sources said design issues tied to a cooling architecture change have largely been resolved. Nvidia has now confirmed a mass-production plan with ODMs and key suppliers, putting development back on track and pointing to a volume ramp from the third quarter of 2026.

Auras Technology reported record first-quarter results and said it expects revenue and profit to grow quarter by quarter in 2026, dismissing market talk that a design tweak on Nvidia's Vera Rubin server cooling system would harm its business. The cooling-module maker posted first-quarter revenue of NT$49.038 billion (approx. US$1.56 billion), up 110.17% year-over-year, gross profit of NT$14.597 billion, up 142.15%, operating profit of NT$12.019 billion, up 175.6%, and net profit after tax of NT$7.916 billion, or NT$20.17 per share.

The US has cleared around 10 Chinese companies to buy Nvidia's H200 artificial intelligence chips, but no deliveries have been made, suggesting that Washington's approval alone may not be enough to revive the company's high-end China business.

Quanta posted record revenue and earnings per share for the first quarter of 2026, but gross margin collapsed as the company absorbed costs tied to a rapid shift into high-priced GPU-based AI servers.

Hon Hai Precision Industry (Foxconn) reported first-quarter 2026 consolidated revenue of NT$2.11 trillion (approximately US$66.95 billion), up 29.68% year over year but down 19% from the fourth quarter of 2025 due to seasonal factors. Core profitability improved despite the slowdown.

Foxconn (Hon Hai) held an online investor briefing on May 14, where executives outlined that cloud and networking products accounted for nearly 50% of group revenue in the first quarter of 2026 and that AI server shipments will more than double year-over-year for 2026. The company said a shift toward consignment supply for some AI server transactions has depressed reported revenue sequentially but did not affect overall profitability and helped reduce working capital and improve capital efficiency.

Ubiqconn Technology reported consolidated revenue of NT$130 million in April 2026, down 25% month-over-month and down 20% year-over-year, and cumulative revenue for the first four months of 2026 totaled NT$678 million, a 12% decline from a year earlier. The firm said short-term revenue was affected by market demand adjustments, but ongoing product mix optimization had begun to drive structural, differentiated growth across its end markets.