IBM and AMD are partnering Zyphra to deliver advanced AI infrastructure for frontier multi-modal model development on IBM Cloud.
Under the multi-year agreement, Zyphra will have access to a large cluster of AMD Instinct MI300X GPUs (top), AMD Pensando AI NICs and DPUs.
“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,” said Krithik Puthalath, CEO and Chairman of Zyphra.
Zyphra is an open-source AI research and product company valued at US$1 billion. It intends to leverage the new cluster to train and deploy large multi-modal foundation models across language, vision and audio on its Maia superagent platform.
“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, General Manager of IBM Cloud.
“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, Executive Vice President and Chief Commercial Officer of AMD.
GenAI competition
The move heightens competition in the generative AI landscape, particularly following the recent NVIDIA-Intel alliance. That landmark partnership, blending NVIDIA’s GPU leadership with Intel’s high-speed networking and data centre solutions, is expected to accelerate AI development cycles and influence enterprise adoption strategies worldwide.
The IBM-AMD axis may help counterbalance the competitive edge NVIDIA-Intel seek to establish, especially in open-source and inference-efficient AI workloads.
Entelechy Asia 