IBM (NYSE: IBM) and AMD (NASDAQ: AMD) have announced a multi-year collaboration to provide advanced AI infrastructure to Zyphra, an open-source AI research and product company based in San Francisco. As part of the agreement, IBM Cloud will deliver a large-scale cluster of AMD Instinct™ MI300X GPUs for Zyphra to train frontier multimodal foundation models—one of the largest generative AI training deployments to date.
Zyphra, recently valued at $1 billion following its Series A round, is building an open-science superintelligence lab focused on novel neural architectures, long-term memory, and continual learning. Its flagship project, Maia, is designed as a general-purpose AI superagent to boost productivity for knowledge workers across industries.
The partnership combines IBM Cloud’s scalability and security with AMD’s compute-to-networking stack—including MI300X GPUs, Pensando™ Pollara 400 AI NICs, and Ortano DPUs—to deliver cutting-edge AI performance. Initial deployment began in September, with expansion planned through 2026.
“This marks the first time AMD’s full-stack training platform has been integrated and scaled on IBM Cloud,” said Krithik Puthalath, CEO of Zyphra. “Together with IBM and AMD, we are powering the next era of open-source enterprise superintelligence.”
Alan Peacock, GM of IBM Cloud, emphasized the importance of efficient scaling: “We are delighted to support Zyphra’s roadmap with economical, enterprise-grade AI infrastructure.” Philip Guido, EVP & CCO of AMD, added: “This collaboration sets a new standard in AI acceleration, combining IBM’s enterprise cloud expertise with AMD’s leadership in HPC and AI.”
The move builds on IBM and AMD’s broader initiatives, including quantum-centric supercomputing—merging IBM’s quantum research with AMD’s high-performance accelerators—to deliver the future of AI and hybrid multi-cloud computing.
See What’s Next in Tech With the Fast Forward Newsletter
Tweets From @varindiamag
Nothing to see here - yet
When they Tweet, their Tweets will show up here.
