AMD unveils next-gen AI chips as it takes on Nvidia: 'For the first time, we...' CEO Lisa Su says
AMD
has unveiled details about its
Instinct MI400 series
, the company's next generation of AI chips set to ship next year. CEO Lisa Su presented the chips at a launch event in San Jose, California, emphasising their design for "rack-scale" systems crucial for powering the massive AI computations of today and tomorrow. Su also claimed that AMD's MI355X can outperform
Nvidia
's Blackwell chips.
The MI400 series chips are designed to be assembled into a full server rack, dubbed Helios, which AMD described as a unified system capable of tying thousands of chips together, as per a report by CNBC.
"For the first time, we architected every part of the rack as a unified system," Su explained, highlighting Helios as "really a rack that functions like a single, massive compute engine."
OpenAI CEO Sam Altman endorses AMD chips
A significant endorsement came from OpenAI CEO Sam Altman, who appeared on stage with Su. Altman expressed confidence in the new chips
"When you first started telling me about the specs, I was like, there's no way, that just sounds totally crazy. It's gonna be an amazing thing," said Altman, whose company is a customer of Nvidia chips. OpenAI also confirmed it will be integrating the AMD chips.
What is different in AMD's newest AI chips
This "rack-scale" approach is vital for hyperscale AI clusters that span entire data centers, catering to the enormous power demands of cloud providers and developers of large language models, the report said. Su directly compared Helios to Nvidia's upcoming Vera Rubin racks, signaling AMD's intent to directly challenge its main rival.
AMD's rack-scale technology aims to put its latest chips squarely in competition with Nvidia's Blackwell chips, which already offer configurations integrating 72 graphics-processing units. Nvidia currently holds a near-monopoly in the market for big data center GPUs, partly due to its early lead in developing essential AI software like CUDA.
OpenAI, notably a significant Nvidia customer, has been providing feedback to AMD on its MI400 roadmap. AMD is positioning the MI400 chips, along with this year's MI355X chips, as a more cost-effective alternative to Nvidia's offerings.
Su also claimed that AMD's MI355X can outperform Nvidia's Blackwell chips, despite Nvidia's proprietary CUDA software advantage, saying that AMD's "really strong hardware" and the "tremendous progress" made by open software frameworks.
AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
The Hindu
2 hours ago
- The Hindu
Meta's data broker
Alexandr Wang, a 28-year-old entrepreneur, has emerged as a key player in the rapidly evolving landscape of artificial intelligence. His company, Scale AI, has strategically positioned itself at the forefront of this transformation. Founded in 2016 as a data labelling company for autonomous vehicles, Scale AI quickly gained prominence. It secured partnerships with major tech giants such as OpenAI, Microsoft, Nvidia, and even the U.S. government. In a significant development, Meta Platforms has invested $14.3 billion to acquire a 49% stake in Scale AI. This acquisition marks a pivotal moment for the company, which now specialises in data curation and intelligent dataset management. As part of this deal, Mr. Alexandr is expected to assume a leadership role within Meta, leading the company's new 'superintelligence' research lab. This move is a much-needed talent refresh and reputation boost for Meta, which has been losing AI researchers and struggling with its Llama 4 launch. The company is bleeding AI talent as researchers are leaving for rival French startup Mistral. Meta's investment in AI is a significant move, and it aligns with the strategies employed by other tech giants. For instance, Microsoft has backed OpenAI, while Amazon supports Anthropic. These partnerships highlight the crucial role of specialised AI firms in the broader ecosystem. However, the Meta-Scale AI deal has sparked discussions about its implications. Some believe it represents an 'exit' for Scale AI, potentially impacting its ability to serve Meta's competitors like Google and OpenAI. Others question whether it's a strategic partnership to enhance Scale AI's capabilities while bolstering Meta's AI standing, or if it's a talent acquisition move by Meta. The deal's structure, with Mr. Alexandr reportedly retaining voting control over Scale despite his move to Meta, indicates a complex arrangement aimed at balancing these interests. Regardless, this move underscores Meta's commitment to leading the AI race. Mr. Alexandr, a vocal advocate for the strategic importance of AI for the U.S. as a matter of national security, was shaped by his upbringing in Los Alamos, New Mexico, the birthplace of the atomic bomb. In a TED Talk, he drew parallels between the transformative impact of the atomic bomb on past warfare and the current potential of AI to redefine global power. AI arms race Mr. Alexandr asserts that, as seen in the Second World War with the atomic bomb, the country that rapidly and effectively integrates new technology into warfare emerges victorious. He warns that the U.S. risks falling behind in the AI arms race, particularly against China, which he claims is ahead in accumulating the vast datasets necessary to develop powerful AI models. Mr. Alexandr has urged the U.S. government to establish a national data reserve, treating data as a strategic national resource akin to petroleum. Scale AI has actively pursued defence contracts, collaborating with the U.S. Department of Defense, Air Force, and the Army to test and evaluate large language models for military purposes. Mr. Alexandr views this work as a 'moral imperative', especially after witnessing China's advances in AI for surveillance and military applications. Mr. Alexandr predicts that AI will dominate warfare within the next decade. He emphasises that 'data will become a new kind of ammunition in the era of AI warfare'. This underscores his call for robust investment in data infrastructure, not only for military superiority but also as a deterrent, akin to nuclear capabilities. For Mr. Alexandr and Scale AI, the mission is clear: to provide the foundational data layer that accelerates AI development across all sectors, including national security. While Meta's significant investment marks a new chapter for Scale AI, potentially altering its client dynamics, Mr. Alexandr's core message about the paramount importance of data and the strategic implications of AI remains unchanged.
Time of India
4 hours ago
- Time of India
Taiwan Blacklists Chinese technology company that is designated as 'National Security Threat' by the US
Taiwan has blacklisted China's Huawei Technologies Co. and Semiconductor Manufacturing International Corp. According to a report in Bloomberg, Taiwan's International Trade Administration has included Huawei, SMIC and several of their subsidiaries in an update of its so-called strategic high-tech commodities entity list, according to the latest version that was made available on its website on Saturday, June 14. This is said to deal a serious blow to the two companies spearheading China's efforts to develop cutting-edge AI chip technologies. The change, which was not publicly announced, requires Taiwanese companies to obtain government approval before exporting to listed entities. When America designated Huawei as 'National Security Threat' Both the companies are also banned by the United States of America (USA). In June 2020 the America's Federal Communications Commission (FCC) designated two Chinese companies -- Huawei Technologies Company (Huawei) and ZTE Corporation (ZTE), as well as their parents, affiliates, and subsidiaries -- as covered companies for purposes of the agency's November 2019 ban on the use of universal service support to purchase equipment or services from companies posing a national security threat. America has also placed SMIC and some of its subsidiaries on the US entity list, which has significantly limited the companies' ability to acquire foreign technology. Citing Taiwan's regulations, the report noted that the new restrictions could significantly limit Huawei and SMIC's access to Taiwan's plant construction technologies, materials, and equipment critical for producing AI semiconductors, such as those manufactured by Taiwan Semiconductor Manufacturing Co. (TSMC) for companies like Nvidia. The updated list also includes Huawei's units in Japan, Russia, and Germany. In 2023, Bloomberg News reported that several Taiwanese firms were assisting Huawei in developing a covert network of chip plants in southern China. This latest move by Taipei may disrupt those efforts, further constraining the Chinese companies' semiconductor ambitions. The report highlighted that Taiwan had previously imposed blanket bans on exporting chipmaking equipment, such as lithography machines, to China but had not targeted specific Chinese tech firms on its entity list until now. TSMC, a key supplier for Apple and Nvidia, halted supplies to Huawei in 2020 due to U.S. export controls. Huawei and SMIC made headlines in 2023 by unveiling a 7-nanometer chip made in China, surprising U.S. policymakers. Despite global restrictions, the companies remain China's best hope for bridging the AI chip gap left by limited access to Nvidia's advanced semiconductors. The restrictions come amid escalating tensions, with Taiwan President Lai Ching-te earlier this year labeling China a 'foreign hostile force' and introducing measures to counter infiltration, Bloomberg reported. China, which claims Taiwan as its territory and has threatened unification by force, has not yet responded to the new measures. AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Time of India
6 hours ago
- Time of India
Nvidia CEO Jensen Huang says Anthropic CEO is very wrong, and on almost everything he said about AI: Don't do it in a dark room and tell me ...
Nvidia CEO Jensen Huang strongly refuted Anthropic CEO Dario Amodei's predictions regarding AI-driven job automation during a press briefing. Huang dismissed Amodei's claim that AI could eliminate 50% of entry-level white-collar jobs within five years, advocating for transparent AI development. While acknowledging job transformation, Huang emphasized AI's potential to create new opportunities and boost productivity. Nvidia CEO Jensen Huang expressed sharp disagreement with Anthropic CEO Dario Amodei 's predictions about AI-driven job automation during a press briefing at VivaTech in Paris. Huang specifically challenged Amodei's recent claim that AI could eliminate up to 50% of entry-level white-collar jobs within five years, dismissing the notion as overly alarmist. 'I pretty much disagree with almost everything [Amodei] says,' Huang stated, critiquing three key points he attributed to Amodei: that AI is so dangerous only a select few should develop it, that it's too costly for widespread development, and that its power will lead to massive job losses. 'If you want things to be done safely and responsibly, you do it in the open … Don't do it in a dark room and tell me it's safe,' Huang added, advocating for transparent and collaborative AI development. What Anthropic CEO said on AI Amodei, in a recent Axios interview, had warned that AI could disrupt half of entry-level office roles and urged policymakers to prepare for economic impacts. Huang countered that while AI will transform jobs, it will also create new opportunities, citing increased productivity as a driver for hiring. 'Some jobs will be obsolete, but many jobs are going to be created,' he said. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like 선납금·위약금 없이 벤츠를 원하는 기간만큼 이용하세요. 계약 만료시 반납, 인수 모두 가능합니다. 써치오토모빌 더 알아보기 Undo Anthropic on Nvidia CEO Jensen Huang's statement In response, an Anthropic spokesperson clarified to Fortune that Amodei has never claimed only Anthropic should build AI. 'Dario has advocated for a national transparency standard for AI developers … so the public and policymakers are aware of the models' capabilities and risks,' the statement read. The spokesperson reaffirmed Amodei's concerns about AI's economic impact and his commitment to addressing them. Anthropic, founded in 2021 by Amodei and other ex-OpenAI researchers, prioritizes safety in AI development, a focus rooted in the founders' reported concerns over OpenAI's direction. The public clash between Huang and Amodei highlights ongoing debates in the AI industry about its societal and economic implications. AI Masterclass for Students. Upskill Young Ones Today!– Join Now
