Leadtek Research Inc (麗臺), a distributor of graphics cards, and artificial intelligence (AI) workstations and servers, yesterday said that it expects revenue growth next year to match the growth pace of Nvidia Corp, as supply constraints ease.
The company in August reported delayed shipments of Nvidia Corp’s RTX 50-series graphics cards, which are widely used for gaming, video editing and professional applications.
Shipments of the RTX PRO 6000 series — typically used for professional purposes such as AI model training and medical imaging — were also delayed, Leadtek said.
Photo courtesy of Leadtek Research Inc
As supply constraints for both products have since eased, Leadtek expects next year’s growth to be driven by those products, Richard Chou (周世偉), general manager of Leadtek’s computer business unit, told reporters on the sidelines of an earnings conference.
Nvidia’s AI supercomputer, the NVIDIA DGX Spark, would also help drive revenue next year, Chou said.
The firm has shifted its focus from consumer-oriented GeForce graphics cards — including the RTX 50 series — to higher-end AI application workstations and servers, with next year’s emphasis largely on the RTX Pro series, Chou said.
Specifically, the RTX Pro 6000 would be Leadtek’s key product next year, as enterprises, government agencies and research institutions have expressed interest in self-built large language models, driving demand for graphics cards with greater memory capacity, he said.
A surge in memorychip prices had a limited effect on the company’s operations, as most of the memory chips used in its products are GDDR — mainly for graphics cards — rather than DDR4 or DDR5 used in systems, which have borne the brunt of the price increases, the company said.
The effects have also been limited because most of the products shipped to Nvidia come with complete memory configurations, Chou said.
The company expects China to account for about 70 percent of its total revenue next year, Leadtek president Jerry Liang (梁傳杰) said.
It plans to expand into other markets, including Thailand and South Korea, aiming to eventually split revenue evenly between China and non-China markets, Liang said.
The company also makes its own-brand graphics cards for professional work stations and AI solutions, he said.
Overall, China accounted for about 70 percent of the company’s total revenue this year, while Taiwan made up about 30 percent, he said.
US tariffs had limited effect on its shipments to the US, mostly virtual PC products, the company said.
However, it is closely monitoring the outcome of an investigation under Section 232 of the US Trade Expansion Act, it added.
Most of the company’s virtual PC products fall under the US government-promoted “zero trust” cybersecurity model, but the suspension of procurements has partially affected its shipments to the US market, Liang said.
Leadtek reported net income of NT$7.38 million (US$234,092) in the first three quarters of this year, compared with a net loss of NT$309.8 million in the same period last year.
Earnings per share in the period rose to NT$0.09 from losses per share of NT$3.71 in the same period last year.
Revenue rose 22.05 percent year-on-year to NT$3.51 billion in the first three quarters of this year.
HORMUZ ISSUE: The US president said he expected crude prices to drop at the end of the war, which he called a ‘minor excursion’ that could continue ‘for a little while’ The United Arab Emirates (UAE) and Kuwait started reducing oil production, as the near-closure of the crucial Strait of Hormuz ripples through energy markets and affects global supply. Abu Dhabi National Oil Co (ADNOC) is “managing offshore production levels to address storage requirements,” the company said in a statement, without giving details. Kuwait Petroleum Corp said it was lowering production at its oil fields and refineries after “Iranian threats against safe passage of ships through the Strait of Hormuz.” The war in the Middle East has all but closed Hormuz, the narrow waterway linking the Persian Gulf to the open seas,
Nanya Technology Corp (南亞科技) yesterday said the DRAM supply crunch could extend through 2028, as the artificial intelligence (AI) boom has led the world’s major memory makers to dramatically reduce production of standard DRAM and allocate a significant portion of their capacity for high-bandwidth memory (HBM) chips. The most severe supply constraints would stretch to the first half of next year due to “very limited” increases in new DRAM capacity worldwide, Nanya Technology president Lee Pei-ing (李培瑛) told a news briefing. The company plans to increase monthly 12-inch wafer capacity to 20,000 in the first half of 2028 after a
Taiwan has enough crude oil reserves for more than 100 days and sufficient natural gas reserves for more than 11 days, both above the regulatory safety requirement, Minister of Economic Affairs Kung Ming-hsin (龔明鑫) said yesterday, adding that the government would prioritize domestic price stability as conflicts in the Middle East continue. Overall, energy supply for this month is secure, and the government is continuing efforts to ensure sufficient supply for next month, Kung told reporters after meeting with representatives from business groups at the ministry in Taipei. The ministry has been holding daily cross-ministry meetings at the Executive Yuan to ensure
RATIONING: The proposal would give the Trump administration ample leverage to negotiate investments in the US as it decides how many chips to give each country US officials are debating a new regulatory framework for exporting artificial intelligence (AI) chips and are considering requiring foreign nations to invest in US AI data centers or security guarantees as a condition for granting exports of 200,000 chips or more, according to a document seen by Reuters. The rules are not yet final and could change. They would be the first attempt to regulate the flow of AI chips to US allies and partners since US President Donald Trump’s administration said it rescinded its predecessor’s so-called AI diffusion rules. Those rules sought to keep a significant amount of AI