AWS recently introduced the EC2 I4i instance type for data-intensive storage and IO-heavy workloads requiring fast access to medium-sized datasets. The new instances can benefit high-performance real-time relational databases, distributed file systems, data warehouses and key-value stores.
The EC2 I4i instances run 3rd generation Intel Xeon Scalable (Ice Lake) processors and offer NVMe-based Nitro SSD local storage, providing high I/O performance, low latency and minimal latency variability. Jeff Barr, vice president and chief evangelist at AWS, suggests possible use cases:
Transactional databases such as MySQL, Oracle DB, and Microsoft SQL Server, as well as NoSQL databases: MongoDB, Couchbase, Aerospike, Redis, and the like. They are also an ideal fit for workloads that can benefit from very high compute performance per TB of storage such as data analytics and search engines.
In comparison to the previous generation I3, the new instances introduce a bigger instance size (i4i.32xlarge), with 1TB of memory and 30TB of local NVMe storage. According to AWS, they provide up to 60% lower storage I/O latency and up to 75% lower storage I/O latency variability. Barr adds:
The i4i.16xlarge and i4i.32xlarge instances give you control over C-states, and the i4i.32xlarge instances support non-uniform memory access (NUMA). All of the instances support AVX-512, and use Intel Total Memory Encryption (TME) to deliver always-on memory encryption.
Performing a benchmark against the previous generation, Michał Chojnowski, software engineer at ScyllaDB, writes:
We observed up to 2.7x higher throughput per vCPU on the new I4i series compared to I3 instances for reads. With an even mix of reads and writes, we observed 2.2x higher throughput per vCPU on the new I4i series, with a 40% reduction in average latency than I3 instances (...) Just three I4i.16xlarge nodes support well over a million requests per second – with a realistic workload.
Source: https://www.scylladb.com/2022/05/09/scylladb-on-the-new-aws-ec2-i4i-instances-twice-the-throughput-lower-latency/
The I4i is not the only write storage-optimized EC2 option released by AWS: at the latest re:Invent, the cloud provider released the Im4gn/Is4gen instances, as covered separately on InfoQ.
The I4i instances support only AMIs running the latest ENA drivers and NVMe 1.4. They are currently available in a small subset of regions in the US and Europe: Northern Virginia, Ohio, Oregon, and Ireland. The new class starts at $0.172/hr on demand for i4i.large, a 10% increase on the pricing of the i3.large ($0.156/hr).