AWS launched R8i and R8i-flex instances powered by custom Intel Xeon 6 processors with DDR5 7200 MT/s memory. The instances scale to 96xlarge, providing up to 384 vCPUs and 3 TB of memory via dual 96-core Xeon 6 chips with two vCPUs per physical core. AWS reports the highest performance and fastest memory bandwidth among comparable Intel cloud processors, though specific CPU customizations are not disclosed. R8i offers up to 20% more compute, 2.5x memory throughput, and 15% better price-performance versus R7i, with larger gains for PostgreSQL, NGINX, and AI workloads. Target use cases include SAP HANA, SQL/NoSQL databases, in-memory caches, and big data analytics, and customers can tune network versus EBS bandwidth to optimize performance.
Amazon Web Services (AWS) has teamed up with Intel to announce the eighth generation of memory-optimized EC2 instances: the R8i and R8i-flex. These new instance types run on specially developed Intel Xeon 6 processors with DDR5 7200 MT/s memory. According to AWS, they deliver the highest performance and fastest memory bandwidth of any comparable Intel processors in the cloud. For Intel, which has been under considerable pressure in recent weeks, this is a welcome boost: the world's largest cloud provider has explicitly chosen customized Xeons.
The fact that AWS is highlighting this collaboration is an important signal that Intel can still compete in the hyperscale cloud market, according to The Register, where competitors such as AMD and Arm have made significant gains in recent years. The fact that AWS is once again opting for customized Xeons shows that Intel still knows how to offer unique value for specific workloads.
The new generation offers up to 20 percent more computing power, 2.5 times higher memory throughput, and 15 percent better price-performance than the R7i instances. Specific workloads benefit even more: PostgreSQL databases run up to 30 percent faster, NGINX web applications up to 60 percent faster, and AI deep learning models up to 40 percent faster. Customers can also adjust the bandwidth distribution between the network and EBS, further improving database and logging performance.
Collection
[
|
...
]