AMD Unveils Vision For Open AI Ecosystem
Dr. Lisa Su, chairman and CEO of AMD, emphasized the company's role in accelerating AI innovation. 'We are entering the next phase of AI, driven by open standards, shared innovation and AMD's expanding leadership across a broad ecosystem of hardware and software partners who are collaborating to define the future of AI,' Su said.
AMD announced a broad portfolio of hardware, software and solutions to power the full spectrum of AI: AMD unveiled the Instinct MI350 Series GPUs , setting a new benchmark for performance, efficiency and scalability in generative AI and high-performance computing. The MI350 Series, consisting of both Instinct MI350X and MI355X GPUs and platforms, delivers a 4x, generation-on-generation AI compute increase and a 35x generational leap in inferencing, paving the way for transformative AI solutions across industries. MI355X also delivers significant price-performance gains, generating up to 40% more tokens-per-dollar compared to competing solutions.
, setting a new benchmark for performance, efficiency and scalability in generative AI and high-performance computing. The MI350 Series, consisting of both Instinct MI350X and MI355X GPUs and platforms, delivers a 4x, generation-on-generation AI compute increase and a 35x generational leap in inferencing, paving the way for transformative AI solutions across industries. MI355X also delivers significant price-performance gains, generating up to 40% more tokens-per-dollar compared to competing solutions. AMD demonstrated end-to-end, open-standards rack-scale AI infrastructure —already rolling out with AMD Instinct MI350 Series accelerators, 5th Gen AMD EPYC processors and AMD Pensando Pollara NICs in hyperscaler deployments such as Oracle Cloud Infrastructure (OCI) and set for broad availability in 2H 2025.
—already rolling out with AMD Instinct MI350 Series accelerators, 5th Gen AMD EPYC processors and AMD Pensando Pollara NICs in hyperscaler deployments such as Oracle Cloud Infrastructure (OCI) and set for broad availability in 2H 2025. AMD also previewed its next generation AI rack called 'Helios .' It will be built on the next-generation AMD Instinct MI400 Series GPUs – which compared to the previous generation are expected to deliver up to 10x more performance running inference on Mixture of Experts models, the 'Zen 6'-based AMD EPYC 'Venice' CPUs and AMD Pensando 'Vulcano' NICs.
.' It will be built on the next-generation AMD Instinct MI400 Series GPUs – which compared to the previous generation are expected to deliver up to 10x more performance running inference on Mixture of Experts models, the 'Zen 6'-based AMD EPYC 'Venice' CPUs and AMD Pensando 'Vulcano' NICs. The latest version of the AMD open-source AI software stack, ROCm 7 , is engineered to meet the growing demands of generative AI and high-performance computing workloads—while dramatically improving developer experience across the board. ROCm 7 features improved support for industry-standard frameworks, expanded hardware compatibility and new development tools, drivers, APIs and libraries to accelerate AI development and deployment.
, is engineered to meet the growing demands of generative AI and high-performance computing workloads—while dramatically improving developer experience across the board. ROCm 7 features improved support for industry-standard frameworks, expanded hardware compatibility and new development tools, drivers, APIs and libraries to accelerate AI development and deployment. The Instinct MI350 Series exceeded AMD's five-year goal to improve the energy efficiency of AI training and high-performance computing nodes by 30x, ultimately delivering a 38x improvement. AMD also unveiled a new 2030 goal to deliver a 20x increase in rack-scale energy efficiency from a 2024 base year, enabling a typical AI model that today requires more than 275 racks to be trained in fewer than one fully utilized rack by 2030, using 95% less electricity.
of AI training and high-performance computing nodes by 30x, ultimately delivering a 38x improvement. AMD also unveiled a new 2030 goal to deliver a 20x increase in rack-scale energy efficiency from a 2024 base year, enabling a typical AI model that today requires more than 275 racks to be trained in fewer than one fully utilized rack by 2030, using 95% less electricity. AMD also announced the broad availability of the AMD Developer Cloud for the global developer and open-source communities. Purpose-built for rapid, high-performance AI development, users will have access to a fully managed cloud environment with the tools and flexibility to get started with AI projects – and grow without limits. With ROCm 7 and the AMD Developer Cloud, AMD is lowering barriers and expanding access to next-gen compute. Strategic collaborations with leaders like Hugging Face, OpenAI and Grok are proving the power of co-developed, open solutions.
Broad Partner Ecosystem Showcases AI Progress Powered by AMD
Today, seven of the 10 largest model builders and Al companies are running production workloads on Instinct accelerators. Among those companies are Meta, OpenAI, Microsoft and xAI, who joined AMD and other partners at Advancing AI, to discuss how they are working with AMD for AI solutions to train today's leading AI models, power inference at scale and accelerate AI exploration and development: Meta detailed how Instinct MI300X is broadly deployed for Llama 3 and Llama 4 inference. Meta shared excitement for MI350 and its compute power, performance-per-TCO and next-generation memory. Meta continues to collaborate closely with AMD on AI roadmaps, including plans for the Instinct MI400 Series platform.
detailed how Instinct MI300X is broadly deployed for Llama 3 and Llama 4 inference. Meta shared excitement for MI350 and its compute power, performance-per-TCO and next-generation memory. Meta continues to collaborate closely with AMD on AI roadmaps, including plans for the Instinct MI400 Series platform. OpenAI CEO Sam Altman discussed the importance of holistically optimized hardware, software and algorithms and OpenAI's close partnership with AMD on AI infrastructure, with research and GPT models on Azure in production on MI300X, as well as deep design engagements on MI400 Series platforms.
CEO Sam Altman discussed the importance of holistically optimized hardware, software and algorithms and OpenAI's close partnership with AMD on AI infrastructure, with research and GPT models on Azure in production on MI300X, as well as deep design engagements on MI400 Series platforms. Oracle Cloud Infrastructure (OCI) is among the first industry leaders to adopt the AMD open rack-scale AI infrastructure with AMD Instinct MI355X GPUs. OCI leverages AMD CPUs and GPUs to deliver balanced, scalable performance for AI clusters, and announced it will offer zettascale AI clusters accelerated by the latest AMD Instinct processors with up to 131,072 MI355X GPUs to enable customers to build, train and inference AI at scale.
(OCI) is among the first industry leaders to adopt the AMD open rack-scale AI infrastructure with AMD Instinct MI355X GPUs. OCI leverages AMD CPUs and GPUs to deliver balanced, scalable performance for AI clusters, and announced it will offer zettascale AI clusters accelerated by the latest AMD Instinct processors with up to 131,072 MI355X GPUs to enable customers to build, train and inference AI at scale. HUMAIN discussed its landmark agreement with AMD to build open, scalable, resilient and cost-efficient AI infrastructure leveraging the full spectrum of computing platforms only AMD can provide.
discussed its landmark agreement with AMD to build open, scalable, resilient and cost-efficient AI infrastructure leveraging the full spectrum of computing platforms only AMD can provide. Microsoft announced Instinct MI300X is now powering both proprietary and open-source models in production on Azure.
announced Instinct MI300X is now powering both proprietary and open-source models in production on Azure. Cohere shared that its high-performance, scalable Command models are deployed on Instinct MI300X, powering enterprise-grade LLM inference with high throughput, efficiency and data privacy.
Red Hat d escribed how its expanded collaboration with AMD enables production-ready AI environments, with AMD Instinct GPUs on Red Hat OpenShift AI delivering powerful, efficient AI processing across hybrid cloud environments.
escribed how its expanded collaboration with AMD enables production-ready AI environments, with AMD Instinct GPUs on Red Hat OpenShift AI delivering powerful, efficient AI processing across hybrid cloud environments. Astera Labs highlighted how the open UALink ecosystem accelerates innovation and delivers greater value to customers and shared plans to offer a comprehensive portfolio of UALink products to support next-generation AI infrastructure.
highlighted how the open UALink ecosystem accelerates innovation and delivers greater value to customers and shared plans to offer a comprehensive portfolio of UALink products to support next-generation AI infrastructure. Marvell joined AMD to highlight its collaboration as part of the UALink Consortium developing an open interconnect, bringing the ultimate flexibility for AI infrastructure.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Arabian Post
2 days ago
- Arabian Post
Binance Identifies North Korea's Growing Cyber Threat
Binance's Chief Security Officer, Jimmy Su, has identified North Korean hacking groups as one of the most significant threats facing the cryptocurrency sector in 2025. As the world's largest cryptocurrency exchange, Binance has been at the forefront of monitoring and responding to cyber threats. Su's recent statements have highlighted the urgency of addressing the escalating risk posed by North Korean cyber actors targeting the crypto industry. The North Korean government has long been linked to a series of cyberattacks aimed at funding its regime through illicit means, with cryptocurrency exchanges being prime targets. These attacks have become more sophisticated over time, with state-backed hackers employing advanced techniques to infiltrate systems and steal millions of dollars' worth of digital assets. Experts argue that this is part of a broader strategy to bypass international sanctions imposed on North Korea. Su's comments, made during an industry event in early 2025, point to a clear rise in the volume and sophistication of these attacks. According to Binance's internal threat intelligence, these cybercriminals are using increasingly advanced malware, social engineering tactics, and even direct network intrusions to steal sensitive data and crypto assets. ADVERTISEMENT An ongoing investigation by Binance's security team revealed that these hackers have been exploiting vulnerabilities in the infrastructure of crypto exchanges and DeFi platforms. In some cases, they have used phishing emails and fake job offers to target employees of key financial institutions, hoping to gain access to privileged systems. Such tactics are a clear indication of a well-coordinated operation with global ambitions. One of the most concerning trends identified is the use of sophisticated ransomware attacks, which have forced some exchanges to pay hefty ransoms to prevent the loss of critical data. The high-profile hacks have raised questions about the resilience of the global financial ecosystem to such breaches, especially as more governments and institutions begin to integrate blockchain technology into their operations. North Korean actors, according to cybersecurity experts, have also shown an increasing willingness to attack smaller and lesser-known exchanges. This shift in focus comes as these platforms often lack the robust security protocols that larger exchanges like Binance and Coinbase have implemented. As a result, they are seen as softer targets, vulnerable to exploitation. The financial implications of these attacks are staggering. Since 2019, estimates suggest that North Korean hackers have stolen over $2 billion in cryptocurrency. This is in addition to the $250 million reportedly taken from a major South Korean exchange in 2023. The stolen funds are typically laundered through complex networks of wallets and mixers, making it challenging for authorities to track and recover the stolen assets. Su's remarks have sparked a broader discussion about the need for improved cybersecurity measures within the cryptocurrency sector. Many industry leaders have called for a more collaborative approach, involving private companies, governments, and international law enforcement agencies. A unified strategy could help create a more secure environment for the rapidly expanding digital asset market. ADVERTISEMENT The North Korean government has long denied any involvement in cyberattacks. However, international cybersecurity firms have linked numerous incidents to Pyongyang's operatives, underscoring the persistent threat posed by the country's cyber espionage efforts. Furthermore, it is widely believed that these hacking groups are funded by the regime to help finance its nuclear weapons programme and other illicit activities. Efforts to combat these threats have led to increased investment in security technology, including more advanced encryption methods, real-time monitoring tools, and fraud detection systems. In addition to these technological solutions, some exchanges are considering implementing stricter Know-Your-Customer regulations to combat money laundering and ensure compliance with international financial rules. As the stakes continue to rise, both industry players and regulators face a significant challenge. The rapid growth of blockchain technologies, along with the increasing use of decentralised finance applications, has created a complex and evolving landscape where traditional security measures are often inadequate. As a result, companies must remain vigilant and proactive in defending against cyber threats from state-sponsored actors like North Korea. While North Korea's cyber campaign remains a major concern, it is not the only threat facing the cryptocurrency sector. Other nation-states, including Iran and Russia, have been linked to similar activities, further complicating efforts to secure the digital assets market. However, the ongoing attention on North Korean actors serves as a reminder of the global reach and scale of the problem.
Arabian Post
4 days ago
- Arabian Post
Trump's AI Chip Deal Could Open China Market to Downgraded Blackwell
President Donald Trump has authorised Nvidia and AMD to resume limited sales of artificial intelligence chips to China, marking a significant shift in U. S. trade policy. The agreement allows the companies to sell specific AI chips in exchange for a 15% revenue share paid to the U. S. government. This move deviates from previous strategies aimed at restricting China's access to advanced semiconductors due to national security concerns. The deal includes Nvidia's H20 chip and AMD's MI308, both intentionally downgraded for export. However, the Trump administration is considering extending this arrangement to Nvidia's more advanced Blackwell chips, albeit in a scaled-down form. Even with a 30%–50% performance reduction, these chips would still outperform China's domestic alternatives, potentially giving Nvidia and AMD a competitive edge in the Chinese market. Chinese authorities have responded cautiously to the deal. The Cyberspace Administration of China has cautioned major domestic tech firms, including Tencent, ByteDance, and Baidu, over their purchases of Nvidia's H20 AI chips, expressing concerns about national security and data privacy. The government has advised against using these chips in government-related projects and, in some cases, ordered a suspension of purchases. ADVERTISEMENT Despite these concerns, demand for the chips remains high due to their superior performance. Nvidia has emphasized that the H20 is not a military or government-focused product, noting that China does not rely on U. S. chips for official uses. The company is also preparing to launch a new artificial intelligence chip for the Chinese market at a significantly reduced price compared to its previously restricted H20 model. The upcoming GPU, part of Nvidia's latest Blackwell-architecture AI chips, is expected to cost between $6,500 and $8,000, making it substantially lower than the H20's price tag of $10,000 to $12,000. The Trump administration's decision to allow the sale of these downgraded chips represents a significant shift from the previous administration's approach to semiconductor exports. While the move has been welcomed by Nvidia and AMD, it has raised concerns among some lawmakers and analysts about the potential erosion of national security safeguards and the risks of enabling Chinese progress in AI.
Tahawul Tech
6 days ago
- Tahawul Tech
NVIDIA reach deal with US to restart sale of AI chips to China
NVIDIA and AMD have reached an agreement with the US government that will allow both companies to resume the sale of their AI chips to China. It has been reported that NVIDIA and AMD have been granted export licenses to resume trading in China, but they will pay a 15% levy on the revenue made to the US. In April, US President Donald Trump expressed his dismay that the US was losing some of its technological edge, and announced that companies would need to apply for export licenses. Washington and Beijing have been locked in an economic arm wrestle for years, and now the race for AI-dominance is at the heart of the Trump administration. The US has major national security concerns when it comes to Chinese companies, but the decision to allow NVIDIA and AMD to sell their H20 and MI308 chips will be seen as a surprise by many analysts. Since returning to The Oval Office, US President Donald Trump has launched a war on tariffs and has made no secret of his disdain at the country's tech giants having most of their manufacturing outside of the United States. Last week, Apple pledged another $100bn into manufacturing in the United States, which many believed was done to avoid any future tariffs that the Trump administration should impose. Trump said he wants to see iPhones that are sold in the US, manufactured domestically. Nvidia reconfigured its H100 chip for China after the US introduced export controls in October 2023. At the beginning of the year, there was a huge surge in demand for the H20 chip from Chinese tech players such as Tencent, Alibaba and ByteDance.
