AMD unveils MI300X, steps up play into AI chip segment

Chipmaker AMD has unveiled MI300X, one of its most-advanced GPUs (graphics processing units) for artificial intelligence, which would be available from the second half of this year.

With this, the Santa Clara-based semiconductor company has challenged Nvidia, which currently dominates the AI chips market.

GPUs are specialised chipsets, which are used by firms like OpenAI to build advanced AI programmes.

While unveiling MI300X at its Data Center and AI Technology Premiere here, AMD Chair and CEO Lisa Su said AI is the defining technology shaping the next-generation of computing and the largest strategic growth opportunity for the company.

"We are laser-focused on accelerating the deployment of AMD AI platforms at scale in the data centre, led by the launch of our Instinct MI300 accelerators planned for later this year and the growing ecosystem of enterprise-ready AI software optimised for our hardware," she said.

AMD on Tuesday revealed new details of the AMD Instinct MI300 Series accelerator family, including the introduction of the AMD Instinct MI300X accelerator, being billed as the world’s most advanced accelerator for generative AI.

"The MI300X is based on the next-gen AMD CDNA 3 accelerator architecture and supports up to 192 GB of HBM3 memory to provide the compute and memory efficiency needed for large language model training and inference for generative AI workloads," the company said.

AMD also introduced the AMD Instinct Platform, which brings together eight MI300X accelerators into an industry-standard design for the ultimate solution for AI inference and training.

"The MI300X is sampling to key customers starting in Q3," it said.

Besides, AMD showcased the ROCm software ecosystem for data center accelerators, highlighting collaborations with industry leaders to bring together an open AI software ecosystem.

It also unveiled Ryzen AI, touted as the world’s first integrated AI engine on an x86 processor, which is deployed in their consumer and commercial notebooks.

This AI engine revolutionises business laptops by enabling premium AI experiences.

AMD VP GPU Platforms Brad McCredie said MI300X creates a huge opportunity in this space and is unique for leadership in the industry.

“One of the things we expect to see, there will be continued AI experience that will happen. Some of these would be cloud-driven and some locally driven. It will be depending on the price point of the system, what kind of capability the particular device has,” he said.

Read | Microsoft is helping finance AMD’s expansion into AI chips

Besides, during the 'Data Center and AI Technology Premiere', AMD also announced a continuation of its relationship with the world's largest cloud computing provider Amazon Web Services (AWS).

AMD also announced software ecosystem collaborations with Hugging Face and PyTorch, which is a machine learning framework, originally developed by Meta AI and now part of the Linux Foundation umbrella.

Hugging Face is an organization and a popular platform that specialises in natural language processing (NLP) and is known for its open-source libraries and pre-trained models. More than 15,000 companies uses its software.

"This new technology partnership between Hugging Face and AMD is very important for the open-source AI community, establishing a new hardware platform for training and inference workloads that will support the most popular architectures and frameworks," said Hugging Face CEO Clement Delangue.