AMD unveiled its complete end-to-end built-in AI platform imaginative and prescient and launched its open, scalable rack-scale AI infrastructure constructed on {industry} requirements at its annual Advancing AI occasion.
The Santa Clara, California-based chip maker introduced its new AMD Intuition MI350 Collection accelerators, that are 4 occasions quicker on AI compute and 35 occasions quicker on inferencing than prior chips.
AMD and its companions showcased AMD Intuition-based merchandise and the continued progress of the AMD ROCm ecosystem. It additionally confirmed its highly effective, new, open rack-scale designs and roadmap that convey management Rack Scale AI efficiency past 2027.
“We will now say we’re on the inference inflection level, and will probably be the driving force,” mentioned Lisa Su, CEO of AMD, in a keynote on the Advancing AI occasion.
In closing, in a jab at Nvidia, she mentioned, “The way forward for AI won’t be constructed by anybody firm or inside a closed system. It will likely be formed by open collaboration throughout the {industry} with everybody bringing their finest concepts.”
AMD unveiled the Intuition MI350 Collection GPUs, setting a brand new benchmark for efficiency, effectivity and scalability in generative AI and high-performance computing. The MI350 Collection, consisting of each Intuition MI350X and MI355X GPUs and platforms, delivers a 4 occasions generation-on-generation AI compute improve and a 35 occasions generational leap in inferencing, paving the best way for transformative AI options throughout industries.
“We’re tremendously excited concerning the work you might be doing at AMD,” mentioned Sam Altman, CEO of Open AI, on stage with Lisa Su.
He mentioned he couldn’t imagine it when he heard concerning the specs for MI350 from AMD, and he was grateful that AMD took his firm’s suggestions.
AMD demonstrated end-to-end, open-standards rack-scale AI infrastructure—already rolling out with AMD Intuition MI350 Collection accelerators, fifth Gen AMD Epyc processors and AMD Pensando Pollara community interface playing cards (NICs) in hyperscaler deployments comparable to Oracle Cloud Infrastructure (OCI) and set for broad availability in 2H 2025. AMD additionally previewed its subsequent era AI rack known as Helios.
It will likely be constructed on the next-generation AMD Intuition MI400 Collection GPUs, the Zen 6-based AMD Epyc Venice CPUs and AMD Pensando Vulcano NICs.
“I feel they’re focusing on a unique sort of buyer than Nvidia,” mentioned Ben Bajarin, analyst at Artistic Methods, in a message to GamesBeat. “Particularly I feel they see the neocloud alternative and an entire host of tier two and tier three clouds and the on-premise enterprise deployments.”
Bajarin added, “We’re bullish on the shift to full rack deployment methods and that’s the place Helios suits by which will align with Rubin timing. However because the market shifts to inference, which we’re simply at first with, AMD is properly positioned to compete to seize share. I additionally suppose, there are many prospects on the market who will worth AMD’s TCO the place proper now Nvidia could also be overkill for his or her workloads. In order that is space to observe, which once more will get again to who the precise buyer is for AMD and it is perhaps a really completely different buyer profile than the shopper for Nvidia.”
The most recent model of the AMD open-source AI software program stack, ROCm 7, is engineered to fulfill the rising calls for of generative AI and high-performance computing workloads— whereas dramatically enhancing developer expertise throughout the board. (Radeon Open Compute is an open-source software program platform that permits for GPU-accelerated computing on AMD GPUs, notably for high-performance computing and AI workloads). ROCm 7 options improved help for industry-standard frameworks, expanded {hardware} compatibility, and new improvement instruments, drivers, APIs and libraries to speed up AI improvement and deployment.
In her keynote, Su mentioned, “Opennesss must be greater than only a buzz phrase.”
The Intuition MI350 Collection exceeded AMD’s five-year aim to enhance the vitality effectivity of AI coaching and high-performance computing nodes by 30 occasions, in the end delivering a 38 occasions enchancment. AMD additionally unveiled a brand new 2030 aim to ship a 20 occasions improve in rack-scale vitality effectivity from a 2024 base 12 months, enabling a typical AI mannequin that at present requires greater than 275 racks to be skilled in fewer than one totally utilized rack by 2030, utilizing 95% much less electrical energy.
AMD additionally introduced the broad availability of the AMD Developer Cloud for the worldwide developer and open-source communities. Objective-built for speedy, high-performance AI improvement, customers can have entry to a totally managed cloud setting with the instruments and suppleness to get began with AI tasks – and develop with out limits. With ROCm 7 and the AMD Developer Cloud, AMD is reducing boundaries and increasing entry to next-gen compute. Strategic collaborations with leaders like Hugging Face, OpenAI and Grok are proving the ability of co-developed, open options. The announcement obtained some cheers from people within the viewers, as the corporate mentioned it might give attendees developer credit.
Broad Companion Ecosystem Showcases AI Progress Powered by AMD
AMD prospects mentioned how they’re utilizing AMD AI options to coach at present’s main AI fashions, energy inference at scale and speed up AI exploration and improvement.
Meta detailed the way it has leveraged a number of generations of AMD Intuition and Epyc options throughout its information middle infrastructure, with Intuition MI300X broadly deployed for Llama 3 and Llama 4 inference. Meta continues to collaborate intently with AMD on AI roadmaps, together with plans to leverage MI350 and MI400 Collection GPUs and platforms.
Oracle Cloud Infrastructure is among the many first {industry} leaders to undertake the AMD open rack-scale AI infrastructure with AMD Intuition MI355X GPUs. OCI leverages AMD CPUs and GPUs to ship balanced, scalable efficiency for AI clusters, and introduced it’s going to provide zettascale AI clusters accelerated by the most recent AMD Intuition processors with as much as 131,072 MI355X GPUs to allow prospects to construct, practice, and inference AI at scale.
Microsoft introduced Intuition MI300X is now powering each proprietary and open-source fashions in manufacturing on Azure.
HUMAIN mentioned its landmark settlement with AMD to construct open, scalable, resilient and cost-efficient AI infrastructure leveraging the total spectrum of computing platforms solely AMD can present.Cohere shared that its high-performance, scalable Command fashions are deployed on Intuition MI300X, powering enterprise-grade LLM inference with excessive throughput, effectivity and information privateness.
Within the keynote, Purple Hat described how its expanded collaboration with AMD allows production-ready AI environments, with AMD Intuition GPUs on Purple Hat OpenShift AI delivering highly effective, environment friendly AI processing throughout hybrid cloud environments.
“They’ll get probably the most out of the {hardware} they’re utilizing,” mentioned the Purple Hat exec on stage.
Astera Labs highlighted how the open UALink ecosystem accelerates innovation and delivers larger worth to prospects and shared plans to supply a complete portfolio of UALink merchandise to help next-generation AI infrastructure.Marvell joined AMD to share the UALink change roadmap, the primary actually open interconnect, bringing the last word flexibility for AI infrastructure.
