The Gen AI Deployment Benefits of the IBM/AMD Partnership
Gen AI has emerged as the driving force behind AI advancements in 2024, revolutionising industries and reshaping business operations across the globe.
In response to this transformative trend, IBM has taken a bold step to provide its customers with a competitive edge in the rapidly evolving AI landscape.
IBM has partnered with AMD to empower enterprises with cutting-edge AI capabilities.
This collaboration aims to make AMD Instinct MI300X accelerators available as a service on IBM Cloud, offering businesses unprecedented access to high-performance computing resources for Gen AI workloads
A new era of AI acceleration
As companies seek to harness the power of large language models and complex AI applications, the demand for robust, efficient, and scalable computing solutions has never been greater.
IBM and AMD's partnership addresses this need head-on, promising to deliver a potent combination of IBM's cloud expertise and AMD's advanced hardware capabilities.
IBM Cloud customers are set to benefit significantly from this partnership, gaining access to AMD Instinct MI300X accelerators as a service from 2025.
"As enterprises continue adopting larger AI models and datasets, it is critical that the accelerators within the system can process compute-intensive workloads with high performance and flexibility to scale,” says Philip Guido, AMD's Executive Vice President and Chief Commercial Officer.
These accelerators are specifically designed for Gen AI inferencing workloads, a crucial aspect of AI deployment that involves using trained models to generate outputs based on new inputs.
The AMD Instinct MI300X is a powerhouse of AI computation. Boasting 19,456 stream processors and 1,216 matrix cores across 304 compute units, this accelerator is capable of reaching a peak engine clock of 2100 MHz.
Its performance metrics are equally impressive, with a peak eight-bit precision (FP8) performance of 2.61 PFLOPs, which can be doubled to 5.22 PFLOPs with structured sparsity.
Empowering AI development
The integration of AMD Instinct MI300X accelerators with IBM's watsonx AI and data platform will provide additional AI infrastructure resources for scaling AI workloads across hybrid cloud environments.
This integration is crucial for enterprises looking to upskill their Gen AI inference workloads and advance their AI development efforts.
One of the key advantages of the AMD Instinct MI300X is its substantial memory capacity. With 192GB of high-bandwidth memory (HBM3) and a 8192-bit memory interface, it offers a peak memory bandwidth of 5.3 TB/s.
This expansive memory allows for larger models to run on fewer GPUs, potentially reducing costs for businesses deploying AI at scale.
"Leveraging AMD's accelerators on IBM Cloud will give our enterprise clients another option to scale to meet their enterprise AI needs, while also aiming to help them optimise cost and performance,” says Alan Peacock, IBM Cloud general manager, highlighted the significance of this partnership.
Enhancing AI infrastructure
The proposed architecture offers access to AMD Instinct MI300X accelerators through IBM Cloud Virtual Servers for VPC and container support with IBM Cloud Kubernetes Service and IBM Red Hat OpenShift on IBM Cloud.
This flexibility in deployment options is designed to help enterprises optimise performance and security when running AI applications.
Furthermore, the partnership extends to enabling Red Hat Enterprise Linux AI and Red Hat OpenShift AI platforms to run Granite family large language models (LLMs) with alignment tooling using InstructLab on MI300X accelerators.
The partners intend to use IBM Cloud’s security and compliance capabilities to support highly regulated industries’ use of AI.
This integration of software and hardware solutions demonstrates a comprehensive approach to AI infrastructure development.
This collaboration between IBM and AMD is not occurring in isolation. It's part of a broader trend in the tech industry towards more specialised and efficient AI computing solutions.
IBM's "AI First" strategy, as outlined by CEO Arvind Krishna, emphasises the importance of integrating AI into every business function.
Arvind's framework of "building blocks" for AI—data, models, governance, assistants, and agents—aligns closely with the capabilities offered by this new partnership.
The AMD Instinct MI300X accelerators provide the computational power necessary to handle the complex data processing and model training required for advanced AI applications.
Looking ahead
As businesses continue to explore the potential of Gen AI, partnerships like this one between IBM and AMD position IBM as an attractive provider of AI solutions.
By combining IBM's cloud expertise with AMD's hardware innovations, this collaboration promises to deliver a powerful platform for AI development and deployment.
As Gen AI continues to transform business operations across industries, the availability of powerful, efficient, and scalable computing solutions will be crucial for companies looking to stay competitive in an increasingly AI-driven world.
Explore the latest edition of AI Magazine and be part of the conversation at our global conference series, Tech & AI LIVE.
Discover all our upcoming events and secure your tickets today.
AI Magazine is a BizClik brand
