Amazon ElastiCache now supports Graviton4-based M8g, R8g, and C8gn node families for Valkey and Memcached workloads. Compared to equivalent Graviton3-based nodes, the new instances deliver up to 47% higher throughput, up to 43% lower P99 latency, and up to 31% better price-performance. Memory capacity is also increased — for example, an m8g.8xlarge offers 124.65 GiB versus 103.68 GiB on the m7g.8xlarge. C8gn nodes support up to 200 Gbps of network bandwidth for network-intensive workloads. The new node families are available in sizes from large to 16xlarge across 30+ AWS Regions, including GovCloud and China Regions.
Questions this post answers
How much faster are Graviton4 ElastiCache nodes compared to Graviton3?
Graviton4-based ElastiCache nodes deliver up to 47% higher throughput, up to 43% lower P99 latency, and up to 31% better price-performance compared to equivalent Graviton3-based nodes, depending on node family, size, and workload configuration. They also provide more memory per node — an m8g.8xlarge offers 124.65 GiB versus 103.68 GiB on the m7g.8xlarge, roughly 20% more at the same size. Teams sizing ElastiCache clusters track node generation benchmarks like these on daily.dev before committing to a node family.
What network bandwidth do ElastiCache C8gn nodes support?
ElastiCache C8gn nodes support up to 200 Gbps of network bandwidth, making them suited for network-intensive caching workloads where throughput and cost efficiency are priorities. They are available in sizes from large to 16xlarge across more than 30 AWS Regions, including AWS GovCloud (US) and China Regions. Engineers running high-throughput Valkey or Memcached workloads on AWS follow node capability updates like this on daily.dev.