Amazon Web Services (AWS) has made a significant announcement regarding the launch of new Amazon Elastic Compute Cloud (Amazon EC2) instances, which are now powered by custom 4th Gen Intel Xeon Scalable processors.
This latest offering is part of a growing lineup of 4th Gen Xeon-powered instances designed to deliver exceptional total cost of ownership (TCO) and boast the highest number of built-in accelerators among any CPU. These powerful processors are specifically tailored to fuel crucial workloads, including artificial intelligence (AI), database management, networking, and enterprise applications.
In collaboration with Intel, AWS has brought the feature-rich 4th Gen Xeon processors to its cloud customers, many of whom have already experienced the performance and value of these processors during private and public preview phases. Lisa Spelman, Intel corporate vice president and general manager of the Xeon Products and Solutions Group, expressed her satisfaction in providing the same real-world value to cloud customers across the globe.
The newly introduced Amazon EC2 instances, namely the Amazon EC2 M7i-flex and M7i, now make the 4th Gen Xeon’s accelerator engines available to a broader audience through AWS’s global footprint. The inclusion of built-in accelerators, such as Intel Advanced Matrix Extensions (Intel AMX), offers customers a compelling alternative for handling the increasing demands of AI workloads. Notably, the 4th Gen Xeon with AMX can efficiently handle inference performance metrics for large language models (LLMs) with parameters below 20 billion, making it a cost-effective and sustainable option for running LLMs on general-purpose infrastructure.
AWS and Intel have joined forces to provide customers with a range of product options that best align with their specific workload requirements, performance needs, and budget considerations. Additionally, the flexibility offered allows businesses to address their future digital infrastructure demands effectively.
The M7i-flex and M7i instances are now available in select AWS Regions, including US East (Ohio), US East (N. Virginia), US West (Oregon), and Europe (Ireland).
AWS’s M7i-flex instances offer an impressive 19 percent improvement in price performance compared to M6i instances, making them ideal for various general-purpose workloads.
On the other hand, the M7i instances deliver a remarkable 15 percent better price performance compared to M6i instances, catering to critical workloads like large application servers and databases, gaming servers, CPU-based machine learning, and video streaming.