The Biden administration has introduced a directive aimed at maintaining the United States’ leadership in artificial intelligence (AI) technology by enforcing restrictions on AI chip exports. Countries are now classified into three tiers: Tier 1 nations (e.g., Japan, the UK) have unrestricted chip access, Tier 2 nations (e.g., Malaysia, Singapore) face limits, and Tier 3 nations (e.g., China, Russia) are entirely barred from accessing US chip technology.
This move addresses concerns over Tier 3 nations bypassing existing restrictions by leveraging cloud computing services in other regions. Notably, US companies are limited to deploying only 50% of their computing power abroad, with stringent caps on allocations to Tier 2 countries.
Malaysia, positioned as a Tier 2 country, could experience both opportunities and challenges. That said, its aim to be a regional data hub will largely remain unaffected, bolstered by partnerships with global technology leaders. While its semiconductor and AI sectors may face limitations on GPU access, the country is strategically positioned to benefit from the expansion of AI data centres by global technology giants like Microsoft, Google and Amazon, who are exempt from computing power limits and permitted to establish AI data centres in Tier 2 countries without restrictions.
Image: Bloomberg
However, industry reactions to the new policy are mixed. Malaysian Semiconductor Industry Association (MSIA) President Wong Siew Hai criticised the directive for creating confusion, citing conflicting compliance requirements for GPU purchases. In response to the proposed restrictions, the Ministry of Science, Technology, and Innovation (Mosti) announced plans to collaborate with the Ministry of Investment, Trade, and Industry (Miti) to prepare for potential impacts.
Meanwhile, Malaysian firms such as YTL Power remain optimistic. YTL, partnering with NVIDIA to deploy cutting-edge AI data centres in Asia, asserted that its operations remain unaffected due to exemptions granted to US hyperscalers.
NVIDIA Blackwell AI Chip (Image: NVIDIA)
In contrast, the Semiconductor Industry Association (SIA) and NVIDIA expressed concerns over the broader implications of the export curbs. SIA President John Neuffer warned that the policy might harm US competitiveness, stifle innovation, and cede strategic markets to global rivals. NVIDIA’s Ned Finkle echoed these sentiments, cautioning that the “AI Diffusion” rule could derail economic growth and undermine the US’ technological leadership.
USA: The Biden administration has unveiled a bold new directive aimed at preserving the United States’ dominance in artificial intelligence (AI) technology by tightening restrictions on AI chip exports.
According to an article published by Lowyat, the move introduces a tiered system for classifying nations, each with distinct access to US semiconductor technology, to curb the spread of sensitive AI capabilities to geopolitical rivals.
Three-tiered system for AI chip access
Under the new framework, countries are classified into three categories based on their access to advanced AI chips.
Tier 1 nations, including close US allies like Japan and the UK, enjoy unrestricted access to AI chips, positioning them as key partners in advancing the technology.
Tier 2 nations, such as Malaysia and Singapore, will face restrictions, limiting their access to certain high-performance chips.
Meanwhile, Tier 3 nations, including China and Russia, are completely barred from receiving AI chips, a move aimed at preventing these countries from exploiting the technology for military or economic gains.
One of the key concerns driving these new restrictions is the possibility of Tier 3 nations bypassing export controls by using cloud computing services hosted in third-party regions.
To counter this, the administration has imposed a cap on the amount of computing power that US companies can deploy abroad, limiting operations to just 50% of their total computing capacity and imposing stringent controls on allocations to Tier 2 nations.
Malaysia’s strategic position – opportunities and challenges
For Malaysia, placed in Tier 2, the policy presents both opportunities and challenges.
While the country’s ambition to become a regional data hub remains intact, it faces potential limitations in its semiconductor and AI sectors due to restricted access to crucial AI chips.
Despite these hurdles, Malaysia stands to benefit from the continued growth of AI data centers in the region, with global tech giants like Microsoft, Google, and Amazon being exempt from the computing power restrictions.
These companies are allowed to build AI data centers in Malaysia without limitations, reinforcing the nation’s standing as a hub for AI infrastructure in Southeast Asia.
However, the policy has sparked a mixed reaction from industry stakeholders.
Wong Siew Hai, President of the Malaysian Semiconductor Industry Association (MSIA), has expressed concerns over the confusion created by the directive, particularly regarding compliance requirements for GPU purchases.
In response, Malaysia’s Ministry of Science, Technology, and Innovation (Mosti) has announced plans to work closely with the Ministry of Investment, Trade, and Industry (Miti) to prepare for the potential economic impact of the restrictions.
On the other hand, some Malaysian companies, such as YTL Power, remain optimistic.
The company, which is working with NVIDIA to deploy state-of-the-art AI data centers in Asia, believes its operations will remain unaffected due to exemptions granted to major US tech firms.
Innovation and US competitiveness at risk
While Malaysia navigates its position within the new export framework, global industry voices have raised alarms about the broader impact of the export restrictions.
The Semiconductor Industry Association (SIA) has cautioned that the policy could stifle innovation and erode US competitiveness in the global tech race.
SIA President John Neuffer warned that by limiting access to advanced AI chips, the US risks ceding critical markets to global competitors, undermining its leadership in the fast-evolving AI sector.
NVIDIA, a leading provider of AI chips, echoed these concerns.
Ned Finkle, NVIDIA’s Vice President of Global Public Policy, argued that the new export restrictions—especially the “AI Diffusion” rule—could derail economic growth and hinder the development of cutting-edge technologies, ultimately weakening the US position as the global leader in AI and semiconductor innovation.
As the US moves forward with these restrictions, the industry will continue to grapple with the delicate balance between national security concerns and the need to foster innovation and maintain a competitive edge in the global technology arena.
KUALA LUMPUR (Jan 13): The Ministry of Science, Technology and Innovation (Mosti) is in discussions with the Ministry of Investment, Trade and Industry (Miti) and other ministries to prepare for any potential development following reports that the United States is considering further export restrictions on artificial intelligence (AI) chips.
Mosti Minister Chang Lih Kang acknowledged that the potential impact could be significant for Malaysia but clarified that the reports are still unconfirmed.
“If they really implement (export restrictions), of course there will be impact not only to Malaysia but also to the entire global community. It will definitely impact Malaysia, especially as we are looking at growing AI in a big way. So without those AI chips, I think we will have problems.
“But as of now, everything is still hearsay. We do not know what kind of category and what countries in what categories (are affected), but of course, we’ve heard that 85% of the countries are in Tier 2 (category), which is not the most permissive but also not the most restrictive,” he told the media after launching the Teater LogTech research laboratory and TalentBridge programme at the Faculty of Computer Science & Information Technology, Universiti Malaya, on Monday.
According to reports, under the new US restriction, companies can apply for blanket permission to ship chips to data centres in most parts of the world, provided that no more than a quarter of their total computing power is located outside of Tier 1 countries, and no more than seven per cent in any one Tier 2 country.
The outgoing administration of President Joe Biden is reportedly planning a new round of export controls on advanced chips used in data centres, targeting both countries and companies. Bloomberg, citing anonymous sources, first reported the potential move.
YTL Power International Bhd managing director Datuk Seri Yeoh Seok Hong reiterated that the group is the first non-US company selected in Asia to partner Nvidia Corp to deploy and manage a supercomputer driven by the GB200 Blackwell GPUs or graphic processing units in the region. (Photo by Zahid Izzani/The Edge)
KUALA LUMPUR (Jan 13): YTL Power International Bhd (KL:YTLPOWR), which is in the midst of setting up its own artificial intelligence (AI) data centres using the latest chips by US chip giant Nvidia Corp, does not expect to be impacted by the proposed US rules on advanced chip exports into countries like Malaysia.
Speaking to The Edge in response to a query on the impact of the potential export curbs, YTL Power managing director Datuk Seri Yeoh Seok Hong reiterated that the group is the first non-US company selected in Asia to partner Nvidia to deploy and manage a supercomputer driven by the GB200 Blackwell GPUs or graphic processing units in the region.
“We are confident that we will not be affected by the latest announcement on the limitation on the export of GPUs,” said Yeoh.
“We understand that American hyperscalers, which include Nvidia, are not subject to the limits. As the only non-US company selected by Nvidia in Asia to deploy its latest chips on its DGX Cloud AI platform, there should be no issue with our roll-out,” he said.
Yeoh also said the sanction will not affect YTL Power’s customer pipeline for its AI data centres.
From their Jan 8 peak of RM4.51, shares of YTL Power have fallen as much as 11.3% to RM4 at Monday’s market close, dragged by concerns over the impact of the potential US chip export curbs on YTL Power's prospects of securing the latest, highly sought-after Nvidia chips made public since late 2023.
Shares of YTL Corp Bhd (KL:YTL), which owns a 48.43% direct stake in YTL Power, declined 14% over the same period.
YTL was mentioned by Nvidia, when the latter revealed the Blackwell chips in March 2024. In a statement, Nvidia said “sovereign AI clouds” such as YTL Power’s will provide Blackwell-based cloud services and infrastructure.
To be sure, under the alleged proposed regulation, firms headquartered in the US can apply for blanket US government permission to ship chips to data centres in most other parts of the world, based on certain volumes.
Separately, Nvidia DGX Cloud vice-president Alexis Black Bjorlin also named YTL Corp as one of five partners, alongside Amazon Web Services, Google Cloud, Microsoft Azure, and Oracle Cloud, which will “bring cutting-edge Nvidia technology to their clouds”.
The supercomputer combines 72 Blackwell GPUs and 36 Grace CPUs interconnected by the fifth-generation NVLink communications link to accelerate the development of AI models on DGX Cloud.
Analysts’ views on the prospects of proxies in the AI race have been mixed due to uncertainties raised from any cuts in chip supply.
A fund manager, when contacted, believes that the policy will be positive for YTL Power in view that “even less [parties] will get access to GB200 [GPUs] in its vicinity, while it has preferential treatment”.
Up in the air is whether companies like YTL Power plan to secure the validated end user (VEU) status — which requires adherence to US government security requirements — to secure higher limits for the GPUs, and bypass the nationwide volume limits that may be imposed on countries like Malaysia.
YTL Power first announced in December 2023 that it was going to collaborate with Nvidia to build an AI infrastructure that will be powered by the American chip firm’s technology. The collaboration was announced when Nvidia founder and chief executive officer Jensen Huang made his maiden visit to Malaysia.
Aside from YTL Power, companies like Amazon, Google, Microsoft, and Oracle — the same list of five Nvidia partners mentioned by Bjorlin — are also setting up data centres in Malaysia.
Analysts covering YTL Power, which has set aside 100MW of its 500MW Kulai data centre capacity for AI infrastructure, have previously said that the first 20MW of the AI portion could begin operating from mid-2025 if the chipsets arrive on time.
At RM4, YTL Power is trading at a trailing price-earnings ratio (PER) of 10.7 times and a forward PER of 10.9 times, compared with forward PERs of peers such as Tenaga Nasional Bhd (KL:TENAGA) at 19.8 times, Malakoff Corp Bhd (KL:MALAKOF) at 14.4 times, and Mega First Corp Bhd (KL:MFCB) at 9.7 times.
There are currently 13 ‘buy’ calls and one ‘neutral’ rating among 14 research houses covering the power, water and telco firm, with target prices ranging from RM3.30 to RM7, and an average TP of RM5.31. At RM4, the group has a market capitalisation of RM32.85 billion.
KUALA LUMPUR (Jan 13): US-based Semiconductor Industry Association (SIA) and Nvidia Corp, two prominent voices in the global chip industry, have criticised the US government’s new export curbs on artificial intelligence (AI) chips.
The outgoing Joe Biden administration’s sweeping restrictions on advanced semiconductors and AI technology, billed as a national security measure, have drawn sharp criticism for potentially stifling innovation, harming economic growth, and eroding the US’s competitive edge in AI.
In a statement on Monday, the SIA president and CEO John Neuffer highlighted that the Biden administration’s decision to publish an interim final rule titled, “Export control framework for artificial intelligence diffusion”, would impose global restrictions and onerous licensing requirements on US exports of advanced integrated circuits (ICs).
This comes at a time when several regulations have already been implemented in recent years to control and restrict access to advanced semiconductors.
“We’re deeply disappointed that a policy shift of this magnitude and impact is being rushed out the door days before a presidential transition and without any meaningful input from industry. The new rule risks causing unintended and lasting damage to America’s economy and global competitiveness in semiconductors and AI by ceding strategic markets to our competitors.
The stakes are high, and the timing is fraught. We stand ready to work with leaders in Washington to chart a path forward that protects national security while allowing us to do what America does best — compete and win globally,” said Neuffer.
SIA members account for 99% of all US semiconductor industry sales, amounting to US$264 billion in 2023. Its charter members include Nvidia, Advanced Micro Devices Inc (AMD), Intel Corp, Broadcom Inc, Qualcomm Inc, Micron Technology Inc and Texas Instruments Inc.
Meanwhile, Nvidia is one of the most valuable companies in the world with a market capitalisation of US$3.26 trillion. The AI chip giant is specialising in accelerating computing and graphics processing unit (GPU).
Ned Finkle, the vice president of government affairs at Nvidia, pointed out that for decades, leadership in computing and software ecosystems has been “a cornerstone of American strength and influence worldwide”.
Therefore, the US federal government has wisely refrained from dictating the design, marketing and sale of mainstream computers and software — key drivers of innovation and economic growth.
“The first Trump Administration laid the foundation for America’s current strength and success in AI, fostering an environment where US industry could compete and win on merit without compromising national security.
“As a result, mainstream AI has become an integral part of every new application, driving economic growth, promoting US interests and ensuring American leadership in cutting-edge technology,” he said.
Finkle further said that today, companies, startups and universities around the world are tapping mainstream AI to advance healthcare, agriculture, manufacturing, education and countless other fields, driving economic growth and unlocking the potential of nations.
“Built on American technology, the adoption of AI around the world fuels growth and opportunity for industries at home and abroad.
“That global progress is now in jeopardy. The Biden administration now seeks to restrict access to mainstream computing applications with its unprecedented and misguided “AI Diffusion” rule, which threatens to derail innovation and economic growth worldwide,” he stressed.
Finkle said the Biden administration, in its last days in office, seeks to undermine America’s leadership with over 200 pages “regulatory morass”, “drafted in secret” and “without proper legislative review”.
According to him, this sweeping overreach would impose bureaucratic control over how America’s leading semiconductors, computers, systems and even software are designed and marketed globally.
“And by attempting to rig market outcomes and stifle competition — the lifeblood of innovation — the Biden administration’s new rule threatens to squander America’s hard-won technological advantage.
While cloaked in the guise of an “anti-China” measure, these rules would do nothing to enhance US security,” he warned.
Rather than mitigate any threat, Finkle opined that the new Biden rules would only weaken America’s global competitiveness, undermining the innovation that has kept the US ahead.
“Although the rule is not enforceable for 120 days, it is already undercutting US interests. As the First Trump Administration demonstrated, America wins through innovation, competition and by sharing our technologies with the world — not by retreating behind a wall of government overreach.
We look forward to a return to policies that strengthen American leadership, bolster our economy and preserve our competitive edge in AI and beyond,” he concluded.
KUALA LUMPUR (Jan 13): The Joe Biden administration's plans to introduce new export restrictions on artificial intelligence (AI) chips under a three-tiered system by country have sparked concerns in Malaysia.
According to news reports, Malaysia has been classified as a Tier 2 country, permitted to import only 50,000 graphics processing units (GPUs) over two years, with data centre operators restricted to deploying a maximum of 7% of their computing capacity in any single Tier 2 nation.
Malaysia Semiconductor Industry Association (MSIA) president Datuk Seri Wong Siew Hai told The Edge that the industry is grappling with understanding the directives from such a “blanket” statement.
“First of all, I think it is a very important thing that we must comply with the US requirements. But the purported [new] directives are causing a lot of confusion. It is like two requirements that are not jiving, one statement is based on one assumption, and the other is based on another assumption,” Wong said.
Wong explained that the current compliance requires buyers of GPUs to get approval from vendors like Nvidia Corp. But the new purported tiering system makes interpretation difficult.
Commenting on the AI chips tiering news, Kenanga Investment Bank head of research Peter Kong said the new development may induce companies to come under US security standards to become validated end users (VEUs), which would then not be limited by the chip country quota.
“It does imply that there are higher guardrails being put up for the end-owners of these chips themselves, but this should not preclude assemblers of AI-chip powered servers from being able to conduct business,” Kong said.
Currently, there is no confirmation that the US will implement this, even though Bloomberg reported it would be announced by last Friday (Jan 10), Kong added.
Rollout of data centres to remain unaffected
Meanwhile, industry players believe the rollout of data centre (DC) projects will largely remain intact.
MSIA’s Wong explained that DCs have already obtained approvals from GPU vendors to secure GPUs. “As of now, data that is presented suggests that business is as usual for the industry, unless they receive new instructions from the US government.”
When contacted, YTL Power International Bhd (KL:YTLPOWR) managing director Datuk Seri Yeoh Seok Hong remained confident that the group will not be affected by the latest announcements on the limitation on the export of GPUs.
“We understand that American hyperscalers, which include Nvidia, are not subject to the limits. As the only non-US company selected by Nvidia in Asia to deploy Nvidia’s latest chips on Nvidia’s DGX Cloud AI platform, there should be no issue with our roll out,” he told The Edge.
Nevertheless, YTL Power saw its stock price drop by 25 sen or 5.9% to close at RM4 on Monday, giving it a market capitalisation of RM33.07 billion.
Tradeview Capital chief investment officer Nixon Wong Gok Hey thinks DC projects will be unaffected in the near term. “Based on the guide of the 7% threshold, current DC projects which are mostly US-related are still below that threshold,” he explained.
Little impact on local semiconductor sector
Analysts believe the impact on the local semiconductor industry will be minimal. For one, Tradeview’s Wong said that while Malaysia might be under some restrictions, it will not be entirely cut off from accessing advanced AI chips.
“Our semiconductor sector is not too exposed to advanced AI chips, and our strength lies in legacy, matured tech as Malaysia is still a key player in outsourced semiconductor assembly and test (OSAT) and automatic test equipment (ATE) fields,” Wong explained.
The same view was shared by TA Securities senior analyst Tony Chan Mun Chun, who highlighted that the proposed new regulation will certainly disrupt the global supply chain, and he expects local equipment players may temporarily see some business slow down.
Tradeview’s Wong noted that companies leveraging Nvidia's chips for AI-driven research and development or manufacturing — such as in Internet of Things (IoT), automotive technology, or industrial automation — might face disruptions in sourcing cutting-edge technology.
“I believe that the extent of the impact depends on several factors but may be moderate and will hinge on how restrictive the Tier 2 classification becomes. Perhaps it might just slow down innovation in tech production,” he added.
Malaysia’s foreign investment appeal at risk
Economist Samirul Ariff Othman highlighted that the tiered approach may threaten Malaysia’s attractiveness to foreign investors, particularly US tech firms, who may view restricted chip access as a liability.
Samirul also believes that by categorising Malaysia in the second tier, the US has sent a signal of mistrust, potentially undermining its image as a reliable trade partner and tech hub.
“The spectre of a past Malaysian firm’s inclusion on the US Entity List serves as a stark reminder that export controls and compliance frameworks must be watertight,” he noted.
Meanwhile, MSIA’s Wong stated that all semiconductor manufacturers have been compliant with the US ruling on export controls for advanced chips.
“So far, there was only one violated case reported, but it was related to a distributor, not a manufacturer,” he noted. “The issue happened when the distributor sold it to a customer, then the customer sold it to another party. It’s quite hard to track after the next few layers. But at the first layer, we will definitely comply.”
As for Samirul, the path forward requires a delicate balancing act. Diplomatic engagement with the US to secure VEU status, coupled with robust internal governance to prevent technology misuse, will be crucial.
According to Science, Technology and Innovation Minister Chang Li Kang, Malaysia is bracing for the potential fallout from the US' proposed restrictions on chip exports. He cautioned that the measures, if enforced, would have far-reaching consequences, impacting not just Malaysia but the global technology ecosystem.
"If the restriction is implemented, it will impact Malaysia, especially as we are looking to grow AI in a big way. Without AI chips, we will have problems. But as of now, everything is still hearsay," Chang reportedly said.
Biden unveils last round of AI chip curbs aimed at China, Russia
AI chip giant Nvidia has spoken out against the newest export curbs imposed by the Biden administration.
Justin Sullivan/Getty Images
Washington/Hong KongCNN —
The Biden administration has issued new restrictions on the export of US-developed computer chips that power artificial intelligence (AI) systems, in a final effort to prevent rivals like China from accessing the advanced technology, just a week before leaving office.
The fresh curbs, the culmination of years of attempts to block China from gaining ground in its military and industrial leadership efforts, are expected to further inflame tensions between Washington and Beijing ahead of the inauguration of President-elect Donald Trump next week. They’ve also triggered intense criticism from US tech giants like Nvidia and Oracle.
Speaking to reporters on Sunday, US Secretary of Commerce Gina Raimondo said the new rules were “designed to safeguard the most advanced AI technology and ensure that it stays out of the hands of our foreign adversaries, but also enabling the broad diffusion and sharing of the benefits with partner countries.”
The global export framework, announced Monday, creates three tiers of countries for exports of advanced AI chips and technology. There are no new restrictions for partners and allies like Australia, Japan, South Korea and Taiwan.
A second tier of countries including China and Russia, which are already blocked from buying advanced chips, will be newly subject to restrictions on the sale of the most powerful “closed” AI models, which refer to models whose underlying architectures are not released to the public.
The biggest changes will be faced by the third group, which comprises most of the world, which will soon have new caps on the amount of computing power that can be bought, although they will be able to apply for additional quotas subject to certain security requirements. Analysts have said this change is intended to prevent China from accessing AI chips through third countries, particularly in the Middle East.
The restrictions are being announced against a global backdrop of soaring demand for AI chips made by the likes of Nvidia, AMD and Intel. With days to go before Biden leaves office, the rules now enter a 120-day comment period but will take effect before that period is over.
“We hope that the next administration takes full advantage of those 120 days to listen to experts, industry, industry players, partner countries,” Raimondo said. “I fully expect the next administration may make changes as a result of that input.”
Senior Biden administration officials, speaking on background, were asked Sunday about the extent to which they consulted with the incoming Trump administration. Officials would only acknowledge there were “ongoing discussions about a range of issues.”
“We believe we’re in a critical window right now, particularly vis-a-vis China,” one official said. “Every minute counts from our perspective.”
The latest measures were announced just a month after the outgoing administration announced curbs on the sale of two dozen types of semiconductor-making equipment and restrictions on numerous Chinese companies from accessing American technology.
Since October 2022, the administration has announced several rounds of semiconductor export restrictions targeting Beijing. Chinese leader Xi Jinping has made self-sufficiency a major pillar of his economic strategy to make China a tech superpower.
‘In jeopardy’
Tech giants Nvidia, the world’s largest provider of processors that power AI, and Oracle as well as an influential semiconductor industry group blasted the newest restrictions, accusing the Biden administration of bureaucratic overreach and saying they would harm US competitiveness.
In a blog post published on Monday, Ned Finkle, Nvidia’s vice president of government affairs, wrote the adoption of AI around the world fuels growth and opportunity for industries at home and abroad.
“That global progress is now in jeopardy. The Biden Administration now seeks to restrict access to mainstream computing applications with its unprecedented and misguided ‘AI Diffusion’ rule which threatens to derail innovation and economic growth worldwide,” he wrote.
“While cloaked in the guise of an ‘anti-China’ measure, these rules would do nothing to enhance US security,” he added.
Oracle Executive Vice President Ken Glueck wrote last week the rule “does more to achieve extreme regulatory overreach than protect US interests and those of our partners and allies.”
“It practically enshrines the law of intended consequences and will cost the US critical technology leadership,” he said.
Last week, the Washington-based Semiconductor Industry Association said it was “deeply concerned by the unprecedented scope and complexity of this potential regulation, which was developed without industry input and could significantly undercut US leadership and competitiveness in semiconductor technology and advanced AI systems.”
没有评论:
发表评论