IBM and AMD partner with Zyphra to boost frontier AI training on IBM Cloud

IBM and AMD have announced a collaboration with Zyphra, an open-source AI research and product company based in San Francisco, to deliver advanced AI infrastructure on IBM Cloud. The agreement will provide Zyphra with one of the largest clusters of AMD Instinct MI300X GPUs deployed to date, dedicated to training multimodal foundation models.

Under a multi-year partnership, IBM Cloud will deliver the cluster to support Zyphra’s mission of building a leading open-source superintelligence lab. Zyphra recently closed a Series A round at a USD 1B valuation to advance innovations in neural network architectures, long-term memory, and continual learning.

This initiative marks the first large-scale training deployment on IBM Cloud powered by AMD Instinct MI300X GPUs alongside AMD Pensando Pollara 400 AI NICs and AMD Pensando Ortano DPUs. The initial cluster was made available in September 2025, with further expansion planned in 2026.

Driving Zyphra’s multimodal AI research

Zyphra is training frontier models across language, vision, and audio modalities to support the development of Maia, its general-purpose superagent aimed at delivering productivity gains for enterprise knowledge workers. IBM and AMD’s combined expertise in cloud infrastructure and AI acceleration is expected to play a key role in scaling these efforts.

Krithik Puthalath, CEO & Chairman, Zyphra, said, “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.”

Alan Peacock, GM, IBM Cloud, added, “Scaling AI workloads faster and more efficiently is a key differentiator in achieving ROI for enterprises and startups alike. We are delighted to support Zyphra’s roadmap with AMD, delivering scalable, economical AI infrastructure that accelerates model training.”

Philip Guido, EVP & Chief Commercial Officer, AMD, commented, “The IBM and AMD collaboration delivers innovation at the speed and scale clients demand, representing a new standard in AI infrastructure. By combining IBM’s enterprise cloud expertise with AMD’s leadership in high-performance computing and AI acceleration, we are supporting Zyphra’s pioneering multimodal research and enabling organisations everywhere to unlock AI solutions that deliver real-world outcomes.”

Partnership momentum in AI infrastructure

This collaboration builds on IBM and AMD’s 2024 initiative to make AMD Instinct MI300X accelerators available as a service on IBM Cloud for generative AI and high-performance computing workloads. Together, the companies are advancing performance, efficiency, and scalability for both enterprises and emerging AI players.

Looking ahead, IBM and AMD also plan to explore quantum-centric supercomputing architectures, combining IBM’s quantum computing research with AMD’s leadership in HPC and AI accelerators.

The Zyphra agreement underlines how partnerships between cloud providers, semiconductor leaders, and AI innovators are shaping the infrastructure backbone of next-generation AI.

Read More:

The Gemini Trifecta: what three AI flaws reveal about the future of security

Partner Pulse: One Cube Solutions | System Integrator and Channel Partner (India)

Pure Storage on partner growth and sustainable data models in India

Freshworks and Sonata IT: partner-led SaaS growth and AI-first expansion in India, APAC