Amazon Web Services Inc. today announced the general availability of Amazon S3 Express One Zone, a new tier of its S3 object storage service that is up to 10 times faster than the standard edition.

AWS already offers multiple specialized versions of S3. Most of them, such as the Glacier storage tiers, trade off speed for lower pricing. They enable companies to more cost-efficiently run applications that don’t require the full performance of a standard S3 storage bucket, or data repository.

Express One Zone, in contrast, is geared toward demanding workloads that must retrieve data at a faster pace than a typical S3 environment can manage. One target use case for the new storage tier is artificial intelligence training. According to AWS, Express One Zone can be used to build large AI models that require the ability to access their training datasets millions of times per minute.

The offering also lends itself to a range of other applications. AWS envisions companies using Express One Zone for tasks as varied as powering analytics projects and delivering personalized ads.

S3 users store their data in repositories called buckets. According to AWS, a single Express One Zone bucket can process hundreds of thousands of data requests per second. Moreover, it’s capable of completing those requests with single-digit millisecond latency.

“This new storage class can handle objects of any size, but is especially awesome for smaller objects,” AWS Chief Evangelist Jeff Barr wrote in a blog post. “This is because for smaller objects the time to first byte is very close to the time for last byte. In all storage systems, larger objects take longer to stream because there is more data to download during the transfer, and therefore the storage latency has less impact on the total time to read the object.”

AWS says Express One Zone is not only faster than standard S3 storage but can also be more cost-efficient. According to the cloud giant, the offering allows companies to reduce the cost of data requests by up to 50%. It can also lower the associated compute expenses by as much as 60%.

The compute-related savings stem from the fact that on-demand AWS instances are charged based on usage. The more time passes between the moment an instance is created and the moment it’s shut down, the more the instance costs. The opposite is also true, which creates opportunities to reduce costs. 

Express One Zone’s ability to rapidly complete data requests allows applications to process information faster than would otherwise be possible. The sooner an application completes its assigned data processing task, the sooner the instances hosting it can be shut down. Deprovisioning instances earlier lowers the associated compute costs.

Under the hood, Express One Zone is powered by what AWS describes as a purpose-built hardware and software stack. Customers can deploy storage buckets on hardware hosted in the same Availability Zone as their applications. Keeping records in close proximity to the workloads that use them reduces the amount of time it takes data requests to travel over the network, which lowers latency.

Express One Zone is currently available in AWS’ Northern Virginia, Oregon, Stockholm and Tokyo cloud regions. The Amazon.com Inc. unit will roll out the storage tier to additional locations in the future.

Image: AWS