AWS has added fast compute instances called I4i using Intel Xeon IceLake servers and 30TB of NVMe Nitro SSD storage.
These Nitro SSDs and the Xeons deliver up to 60 per cent lower I/O latency, 75 per cent lower latency variability compared to the prior I3 instances, and up to 30 per cent better compute price/performance. The processors have an all-core turbo frequency of 3.5GHz. There is a I41.32xlarge instance size with 128 vCPUs and 1,024GiB (1.1 TB) of storage using 8x 3,750MiB (3.9GB) Nitro SSDs – a total of 31.2TB of SSD capacity.
Yiftach Shoolman, co-founder and CTO of Redis, offered a testimonial: “Our testing shows I4i instances delivering an astonishing 2.9x higher query throughput than the previous generation I3 instances. We have also tested with various read and write mixes, and observed consistent and linearly scaling performance.”
Avi Kivity, co-founder and CTO of ScyllaDB, had similar to say: “When we tested I4i instances, we observed up to 2.7x increase in throughput per vCPU compared to I3 instances for reads. With an even mix of reads and writes, we observed 2.2x higher throughput per vCPU, with a 40 per cent reduction in average latency than I3 instances.”
The I4i.16xlarge and I4i.32xlarge instances give users 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. I4i instances give customers up to 75Gbit/sec of networking speed and 40Gbit/sec of bandwidth to Amazon’s EBS (Elastic Block Store). Overall, applications go faster for less money.
Intel says AWS’s Nitro System is a combination of dedicated hardware and lightweight hypervisor. The I4i instances deliver practically all compute and memory resources of the host hardware to instances.
AWS says a Nitro SSD has firmware inside responsible for implementing many lower-level functions, with designed-in operational telemetry and diagnostics. It has code to manage the instance-level storage to improve reliability and deliver consistent performance.
AWS says these SSDs are engineered to deliver maximum performance under a sustained, continuous load, with the firmware responsible for garbage collection and wear-leveling to even out Program-Erase cycles per cell over time. The firmware can also be updated, AWS says, at at cloud scale and at cloud speed – more often than commodity SSDs, at any rate.
The new instances, AWS says, “are designed to minimize latency and maximize transactions per second (TPS) on workloads that need very fast access to medium-sized datasets on local storage. This includes 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.”
AWS customers can launch I4i instances today in the AWS US East (N Virginia), US East (Ohio), US West (Oregon), and Europe (Ireland) Regions (with more to come) in On-Demand and Spot form. Savings Plans and Reserved Instances are available, as are Dedicated Instances and Dedicated Hosts.
In order to take advantage of the performance benefits, customers must use recent AMIs that include current ENA drivers and support for NVMe 1.4. The I4i instance home page has more information.