IBM & AMD power Zyphra’s USD $1B open-source AI push
IBM and AMD have announced a collaboration to provide advanced artificial intelligence infrastructure to Zyphra, an open-source AI research and product company based in San Francisco, leveraging integrated capabilities for AMD training clusters on IBM Cloud.
Under a multi-year agreement, IBM will deliver a substantial cluster of AMD Instinct MI300X GPUs on IBM Cloud for Zyphra, enabling the company to train frontier multimodal foundation models. The infrastructure, also featuring AMD Pensando Pollara 400 AI NICs and AMD Pensando Ortano DPUs, represents one of the largest generative AI training platforms to date relying on an AMD stack deployed through IBM Cloud.
Zyphra intends to use this cluster to advance its development of multimodal foundation models, including those for language, vision and audio processing. These models will support Maia, Zyphra's general-purpose superagent designed to improve productivity for knowledge workers across enterprise sectors.
Zyphra, which recently closed a Series A financing round at a USD $1 billion valuation, is working towards establishing an open-source superintelligence laboratory. Its research focuses on novel neural network architectures, long-term memory, and continual learning.
The collaboration is the first large-scale deployment of AMD Instinct MI300X GPUs and AMD networking hardware on IBM Cloud tailored specifically for dedicated AI training by Zyphra. The initial deployment went live in early September with plans for further expansion in 2026.
Partnership remarks
"This collaboration marks the first time AMD's full-stack training platform-spanning compute through networking-has been successfully integrated and scaled on IBM Cloud, and Zyphra is honoured to lead the way in developing frontier models with AMD silicon on IBM Cloud. We're excited to partner with IBM and AMD to power the next era of open-source, enterprise superintelligence," said Krithik Puthalath, CEO and Chairman of Zyphra.
IBM and AMD have positioned themselves to scale resources further as Zyphra's requirements grow, offering a flexible approach to enterprise and research clients needing powerful AI capabilities. The agreement highlights the combined strengths of IBM's cloud infrastructure-regarded for security and scalability-and AMD's hardware acceleration for AI and high-performance computing workloads.
Infrastructure scaling
"Scaling AI workloads faster and more efficiently is a key differentiator in achieving ROI for established enterprises and emerging companies alike. We are delighted to support Zyphra's strategic roadmap as we collaborate with AMD to deliver scalable, economical AI infrastructure that can accelerate Zyphra's model training," said Alan Peacock, GM of IBM Cloud.
"The IBM and AMD collaboration delivers innovation at the speed and scale clients demand, representing a new standard in AI infrastructure. By combining IBM enterprise cloud expertise with AMD leadership in high-performance computing and AI acceleration, we are supporting Zyphra's pioneering work in multimodal and inference-efficient AI, enabling organisations everywhere to build smarter businesses and unlock AI solutions that drive real-world outcomes," said Philip Guido, EVP and Chief Commercial Officer, AMD.
This partnership forms part of wider efforts by IBM and AMD to redefine data centre performance and efficiency for both startups and large enterprise customers. Their solution provides a foundation for hybrid and multi-cloud deployments, aiming to optimise cloud investments and support extended generative AI use cases.
Alongside the AI infrastructure project, IBM and AMD have also revealed plans to develop next-generation computing architectures, known as quantum-centric supercomputing. This research integrates IBM's expertise in quantum computing and software with AMD's advancements in high-performance and adaptive computing technologies. The collaboration seeks to establish future capabilities that combine quantum and classical computation for complex workloads.
The relationship between IBM, AMD, and Zyphra reflects ongoing trends of combining cloud-based AI acceleration with open-source research and development, as enterprises and research labs increase their demand for scalable, flexible, and high-performance AI training environments.
