AWS Graviton5 CPUs Now Available: Purpose-Built For AI With 25% Performance Uplift, 192 Cores, DDR5-8800 & PCIe Gen6 Support

Jun 11, 2026 at 11:50am EDT
AWS Graviton5 CPUs Now Available: Purpose-Built For AI With 25% Performance Uplift, 192 Cores, DDR5-8800 & PCIe Gen6 Support

Amazon says that its AWS Graviton5 is the fastest and most efficient CPU it has ever built and is now generally available for AI and other use cases.

AWS Graviton5 Starts Rolling Out To Customers, Offering Better Performance With Latest Technologies Including PCIe Gen6

Today marks the official "General" availability of AWS's next big chip, the Graviton5 CPU. The chip is said to be much faster than the existing generation of AWS CPUs, and also expand their capabilities with the latest IO features.

Related Story ASUS and MSI Finally Bring AMD’s EXPO Ultra Low Latency To 600-Series Boards, Closing The Gap With AM5’s 800 Lineup

Starting off with the performance, Amazon AWS claims that Graviton5 CPUs will offer up to 25% faster performance versus the prior generation, while applications on the same chip will run 35% faster, AI/ML Inferencing will be 35% faster, and databases will be 30% faster. Each CPU packs 192 cores built on a 3nm process technology & also offers 33% lower inter-core latency.

Some of those IO upgrades that we mentioned above include faster DDR5 memory support with up to 8800 speeds, the fastest DDR5 in the cloud, and support for PCIe Gen6 protocol. Each chip also offers a cache that is five times larger than the prior generation, with each core having access to 2.6x more cache, adding to the chip's expanded AI inference and ML capabilities.

The chip includes a 5x larger L3 cache—a high-speed memory buffer that keeps frequently accessed data close to the processor. Each Graviton5 core has access to 2.6x more L3 cache than Graviton4, which translates to fewer delays waiting for data and faster application response times. Memory performance has also improved, with Graviton5 providing faster memory speeds, enabling you to process larger datasets and run memory-intensive applications more efficiently.

Network and storage bandwidth have also increased, with up to 15% higher network bandwidth and 20% higher Amazon Elastic Block Store (EBS) bandwidth on average across instance sizes, and up to twice the network bandwidth for the largest instances—resulting in faster data transfers, quicker backups, and improved performance for distributed applications.

Amazon AWS

Currently, the Graviton5 CPUs are available in Amazon EC2 M9g and M9gd instances. Meta has already committed to deploying tens of millions of Graviton cores for its Agentic AI use, while Uber & Snowflake are also some of the many customers that are leveraging these chips. Amazon says that it has over 120K customers building on Graviton, and the number continues to rise.

As for an instance-to-instance comparison, the M9g instance offers 25% better compute performance vs the prior generation, while the M9gd instances are designed for firms that require high-speed local SSD storage with up to 11.4 TB of capacity and 30% higher IOPS versus the previous gen.

AWS Graviton5 marks a major leap forward in cloud computing, delivering the company’s fastest and most efficient CPU yet. With 192 cores, up to 25% higher performance, significantly larger caches, PCIe Gen6 support, and faster memory, it brings substantial gains for AI inference, databases, high-performance workloads, and general applications—all while continuing Amazon’s tradition of superior price-performance and energy efficiency.

Now generally available in M9g and M9gd instances, Graviton5 is already being adopted by major customers like Meta, Uber, and Snowflake. As the fifth generation in a highly successful custom silicon journey, it further strengthens AWS’s leadership in delivering powerful, cost-effective, and sustainable cloud infrastructure.

About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.

Follow Wccftech on Google to get more of our news coverage in your feeds.

Deal of the Day