DBT Bureau
Pune, 2 Oct 2025
IBM and AMD announced a collaboration to deliver advanced AI infrastructure to Zyphra, an open-source AI research and product company based in San Francisco, California. Under a multi-year agreement between IBM and Zyphra, IBM is positioned to deliver a large cluster of AMD Instinct™ MI300X GPUs on IBM Cloud for Zyphra to use for training frontier multimodal foundation models. This collaboration is expected to deliver among the largest advanced generative AI training capabilities to date powered by an AMD stack running on IBM Cloud.
Zyphra recently closed a Series A financing round at a $1B valuation to build a leading open-source/open-science superintelligence lab focused on advancing fundamental innovations in novel neural network architectures, long-term memory, and continual learning.
Zyphra partnered with IBM and AMD for their cutting-edge product roadmaps and ability to deliver GPU accelerators at the rapid pace required to drive Zyphra’s innovation forward. This agreement is the first large-scale, dedicated training cluster on IBM Cloud leveraging AMD Instinct MI300X GPUs and AMD Pensando™ Pollara 400 AI NICs and AMD Pensando Ortano DPUs. The initial deployment was made available to Zyphra in early September with planned expansion in 2026.
Zyphra will use the advanced training cluster to develop multimodal foundation models across language, vision and audio modalities to power Maia, a general purpose superagent designed to deliver productivity benefits for knowledge workers across enterprise. IBM and AMD are uniquely positioned to continue scaling computational resources as Zyphra’s AI model training needs expand.
“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 honored to lead the way in developing frontier models with AMD silicon on IBM Cloud,” said Krithik Puthalath, CEO and Chairman of Zyphra. “We’re excited to partner with IBM and AMD to power the next era of open-source, enterprise superintelligence.”
IBM and AMD Unlocking a New Era in AI Training Infrastructure
Last year, IBM and AMD announced a collaboration to deploy AMD Instinct MI300X accelerators as a service on IBM Cloud. Known for its security, reliability and scalability, IBM Cloud’s robust infrastructure complements the capabilities of the AMD Instinct MI300X. This offering was designed to enhance performance and power efficiency for Gen AI models and high-performance computing (HPC) applications.
“Scaling AI workloads faster and more efficiently is a key differentiator in achieving ROI for established enterprises and emerging companies alike,” said Alan Peacock, GM of IBM Cloud. “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.”
“The IBM and AMD collaboration delivers innovation at the speed and scale clients demand, representing a new standard in AI infrastructure,” said Philip Guido, EVP and Chief Commercial Officer, AMD. “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 organizations everywhere to build smarter businesses and unlock AI solutions that drive real-world outcomes.”
IBM and AMD are forging new ground in AI infrastructure, helping redefine performance, efficiency, and scale for enterprise and startup customers. Their hybrid infrastructure environment offers a foundation for scaling AI, with options such as hybrid multi-cloud that can help boost cloud ROI and value for clients’ generative AI deployments.
IBM and AMD also recently announced plans to develop next-generation computing architectures, known as quantum-centric supercomputing, leveraging IBM’s leadership in developing the world’s most performant quantum computers and software, and AMD’s leadership in high-performance computing and AI accelerators.
