Nvidia Corp’s GB300 platform is expected to account for 70 to 80 percent of global artificial intelligence (AI) server rack shipments this year, while adoption of its next-generation Vera Rubin 200 platform is to gradually gain momentum after the third quarter of the year, TrendForce Corp (集邦科技) said.
Servers based on Nvidia’s GB300 chips entered mass production last quarter and they are expected to become the mainstay models for Taiwanese server manufacturers this year, Trendforce analyst Frank Kung (龔明德) said in an interview.
This year is expected to be a breakout year for AI servers based on a variety of chips, as shipments of graphics processing unit (GPU)-based rack systems — including Nvidia’s GB300 and Vera Rubin 200 platforms, as well as Advanced Micro Devices Inc’s MI400 — are set to accelerate, while cloud service providers (CSPs), such as Google Inc, Amazon Web Services and Meta Platforms Inc, are expected to step up use of application-specific integrated circuit (ASIC)-based AI infrastructure, he said.
Photo: AP
As development of GPU and ASIC-based servers accelerates this year, the threshold for system integration among local server manufacturers is rising, Kung said.
Hon Hai Precision Industry Co (鴻海精密), Quanta Computer Inc (廣達) and Wistron Corp (緯創) are some of the Taiwanese server system integration suppliers for overseas tech giants.
The challenge is becoming more pronounced as major CSPs continue to develop ASICs with increasingly advanced and customized specifications, a trend that further raises technical barriers to entry for manufacturers, he said.
GB300 is one of Nvidia’s Blackwell GPU series, with modest upgrades compared with the GB200 across connectors, substrates and thermal components, while the Vera Rubin 200 platform features a significant increase in GPU power consumption, analysts said.
High power consumption in AI servers based on Nvidia chips, coupled with the CSPs’ continued upscaling of AI data centers, results in a growing need for liquid cooling solutions this year, TrendForce analyst Fiona Chiu (邱珮雯) said.
As the market’s heat dissipation solutions remain concentrated on liquid-to-air designs — which serve as a transitional phase between traditional air cooling and full liquid cooling — liquid-to-air systems are still expected to dominate AI infrastructure deployments this year, she said.
Fully liquid-to-liquid cooling solutions are expected to become more prominent next year as data center power density continues to rise, she added.
The developments would benefit Taiwanese liquid-cooling solution providers, with heat dissipation specialists such as Auras Technology Co (雙鴻科技) and Asia Vital Components Co (奇鋐科技) poised to strengthen their positions in the market, Chiu said.
Power supply solution providers such as Delta Electronics Inc (台達電) and Lite-On Technology Corp (光寶科技) are also expected to gain market share with their power supply and infrastructure-related products, she added.
After several years flying high as Asia’s best Nvidia Corp proxy, Taiwan Semiconductor Manufacturing Co (TSMC, 台積電) is increasingly vying with other artificial intelligence (AI) stocks for investor attention. Stock traders are chasing a wider array of beneficiaries as mainstream usage of AI creates demand for hardware beyond the most-advanced chips TSMC makes for Nvidia. Subthemes from the deepening memory crunch to advances in robotics are also luring bids. At the same time, investment caps on single stocks are pushing funds to diversify, while retail investors long familiar with TSMC through its US depositary receipts are being offered a broader set of
Netherlands-based semiconductor equipment supplier ASML Holding NV yesterday said that it is planning to hire an additional 1,000 people in Taiwan this year in response to growing demand from clients. ASML had previously planned to recruit 600 people this year, but that the plan has been adjusted upward, ASML vice president and ASML Taiwan general manager Grace Wang (汪佳慧) told reporters. ASML has a workforce of more than 4,500 in Taiwan, accounting for about 10 percent of its global total, Wang said. This year’s recruitment campaign would focus on adding people in the customer support, manufacturing and supply chain domains to assist ASML
UNDER MICROSCOPE: Taiwan detained three people who allegedly conspired to buy servers in Taiwan and export them using fraudulent documentation, prosecutors said Nvidia Corp chief executive officer Jensen Huang (黃仁勳) on Saturday urged Super Micro Computer Inc to tighten up on compliance after Taiwan detained three people this week for allegedly making fraudulent declarations about artificial intelligence (AI) servers made by its US partner. The development marked the nation’s first crackdown on semiconductor smuggling, which grew after the US slapped restrictions on exports of high-end chips such as Nvidia AI accelerators to China. Nvidia is “rigorous” in explaining regulations to all of its partners, Huang told reporters after arriving in Taipei. “Ultimately Super Micro has to run their own company,” he said in response to
Nvidia Corp yesterday announced that CEO Jensen Huang (黃仁勳) would attend an employee meeting in Taipei tomorrow to celebrate the launch of the company’s Taiwan headquarters project. Huang would attend a gathering at the site of Nvidia’s planned headquarters in Beitou Shilin Technology Park (北投士林科技園區), the company said in a statement. After arriving in Taiwan on Saturday last week, Huang told reporters that he plans to meet with Quanta Computer Inc (廣達) chairman Barry Lam (林百里) and Taiwan Semiconductor Manufacturing Co (TSMC, 台積電) chairman C.C. Wei (魏哲家), and would attend the groundbreaking ceremony for Nvidia’s Taiwan headquarters tomorrow. Nvidia has not yet applied
RSS