Graviton-4 Powered X8g Instances: Efficient and High-Performance Cloud Solutions
AWS has introduced its new Graviton-4-powered, memory-optimized X8g instances, which are now available in ten virtual sizes and two bare metal sizes. These instances offer up to 3 TiB of DDR5 memory and up to 192 vCPUs, making them the most energy-efficient and highest-performing EC2 Graviton instances to date. The X8g instances feature a 16 to 1 ratio of memory to vCPU, making them ideal for applications such as Electronic Design Automation (EDA), in-memory databases and caches, relational databases, real-time analytics, and memory-constrained microservices. Additionally, these instances fully encrypt all high-speed physical hardware interfaces and incorporate enhanced security features from the AWS Nitro System and Graviton4 processors.
Growing Adoption of Graviton-Powered Instances
Over 50,000 AWS customers are already leveraging the existing suite of more than 150 Graviton-powered instances. These customers run a diverse array of applications, including Valkey, Redis, Apache Spark, Apache Hadoop, PostgreSQL, MariaDB, MySQL, and SAP HANA Cloud. With twelve available sizes, the new X8g instances offer even greater flexibility for scaling up (using a bigger instance) or scaling out (using more instances), providing additional options for memory-bound workloads currently running on separate instances.
Enhanced Performance and Specifications
Compared to the previous generation X2gd instances, the X8g instances deliver significant improvements:
- 3x more memory
- 3x more vCPUs
- Over twice the EBS bandwidth (40 Gbps vs. 19 Gbps)
- Twice the network bandwidth (50 Gbps vs. 25 Gbps)
The Graviton4 processors in X8g instances feature twice as much L2 cache per core as Graviton2 processors (2 MiB vs. 1 MiB), 160% higher memory bandwidth, and up to 60% better compute performance.
Built using the 5th generation of the AWS Nitro System and Graviton4 processors, these instances also include security features such as Branch Target Identification (BTI), which protects against low-level attacks aimed at disrupting control flow at the instruction level. For more information about these security enhancements, you can read about how Amazon’s new CPU fights cybersecurity threats or watch the re:Invent 2023 AWS Graviton session.
Instance Specifications
Here are the detailed specifications for the X8g instances:
Instance Name vCPUs Memory (DDR5) EBS Bandwidth Network Bandwidth x8g.medium 1 16 GiB Up to 10 Gbps Up to 12.5 Gbps x8g.large 2 32 GiB Up to 10 Gbps Up to 12.5 Gbps x8g.xlarge 4 64 GiB Up to 10 Gbps Up to 12.5 Gbps x8g.2xlarge 8 128 GiB Up to 10 Gbps Up to 15 Gbps x8g.4xlarge 16 256 GiB Up to 10 Gbps Up to 15 Gbps x8g.8xlarge 32 512 GiB 10 Gbps 15 Gbps x8g.12xlarge 48 768 GiB 15 Gbps 22.5 Gbps x8g.16xlarge 64 1,024 GiB 20 Gbps 30 Gbps x8g.24xlarge 96 1,536 GiB 30 Gbps 40 Gbps x8g.48xlarge 192 3,072 GiB 40 Gbps 50 Gbps x8g.metal-24xl 96 1,536 GiB 30 Gbps 40 Gbps x8g.metal-48xl 192 3,072 GiB 40 Gbps 50 Gbps
### Networking and Storage Support
The X8g instances support Elastic Network Adapter (ENA), ENA Express, and Elastic Fabric Adapter (EFA) Enhanced Networking, providing generous amounts of EBS bandwidth and supporting all EBS volume types, including io2 Block Express, EBS General Purpose SSD, and EBS Provisioned IOPS SSD.
### X8g Instances in Action
#### Databases
X8g instances enable SAP HANA and SAP Data Analytics Cloud to handle larger and more complex workloads than before. Running on Graviton4-powered instances, SAP has observed up to 25% better performance for analytical workloads and up to 40% better performance for transactional workloads compared to the same workloads on Graviton3 instances. This allows SAP to expand its Graviton-based usage to even larger memory-bound solutions.
#### Electronic Design Automation
EDA workloads are critical for designing, testing, verifying, and taping out new generations of chips, including Graviton, Trainium, Inferentia, and the building blocks for the Nitro System. AWS and many other chip manufacturers use the AWS Cloud for these workloads, leveraging the scale and elasticity to provide the necessary compute power for each phase of the design process. This enables engineers to innovate faster by not having to wait for results. Here’s a long-term snapshot from one of the clusters used to support the development of Graviton4 in late 2022 and early 2023:
You can observe bursts of daily and weekly activity, followed by a spike in overall usage during the tape-out phase. The instances in this cluster are on the larger end of the spectrum, representing several hundred thousand cores running concurrently. This capability to scale up compute resources as needed without a dedicated investment in hardware provides unprecedented flexibility.
The new X8g instances will enable AWS and its EDA customers to run even more workloads on Graviton processors, reducing costs and energy consumption while accelerating the time to market for new products.
### Availability
X8g instances are now available in the US East (N. Virginia), US West (Oregon), and Europe (Frankfurt) AWS Regions. They can be utilized in various purchasing options, including On-Demand, Spot, Reserved Instances, Savings Plan, Dedicated Instances, and Dedicated Hosts. To learn more, visit the [X8g page](https://aws.amazon.com/ec2/instance-types/x8g).
### Conclusion
The introduction of Graviton-4-powered X8g instances marks a significant advancement in cloud computing, offering remarkable improvements in memory, compute performance, and energy efficiency. These instances are poised to benefit a wide range of applications, from databases to electronic design automation, making them a versatile and powerful addition to the AWS portfolio. As cloud technology continues to evolve, the X8g instances represent a critical step forward in delivering high-performance, scalable, and secure cloud solutions.
For more details and to start using X8g instances, visit the [AWS EC2 X8g page](https://aws.amazon.com/ec2/instance-types/x8g).
