News
AWS on Compute Power Trip, Adding 5 New EC2 Instance Types
Amazon Web Services (AWS) added five new instance types to its Elastic Compute Cloud (EC2) this week, along with other enhancements intended to increase the flexibility of its compute offerings.
The announcements came Tuesday as part of the company's 2020 re:Invent conference, being held as a three-week online event this year due to COVID-19 constraints.
The addition of five new EC2 instance types -- each promising various compute, storage, performance and cost benefits -- improves on an already comprehensive selection of compute options from AWS.
"AWS already has more compute instance types than any other cloud provider, with instances based on the fastest processors from Intel, cost-optimized instances with AMD processors, the most powerful GPU instances from NVIDIA, instances that feature up to 400 Gbps networking performance, and the only Arm-based instances in the cloud offering customers 40% better price performance with AWS-designed AWS Graviton2 processors," the company said in its announcement.
The new EC2 instances are as follows:
- C6gn, powered by Arm-based AWS Graviton2 processor: Designed for network-intensive workloads such as data modeling, high-performance computing and video transcoding, but with better packet processing chops than other instances. C6gn lets users "consolidate their workloads onto fewer instances or smaller instance sizes, and reduce infrastructure costs," according to AWS.
- G4adm, powered by AMD GPUs: Designed for graphics-intensive applications, these instances can render high-quality images for virtual workstations, game streaming, video transcoding and other jobs, but at a "45% better price performance" over NVIDIA-powered instances (namely, G4dn).
- M5zn, powered by Intel Xeon Scalable processors: Designed to provide fast compute performance and high networking throughput, but with less memory. Users can turn to this instance to avoid buying more compute resources than they need for tasks like "performing complex calculations and real-time analysis for their financial, analytics, and gaming workloads."
- D3/D3en, powered by Intel Xeon Scalable processors: For storage-heavy workloads that require high network speeds and CPU performance. These instances deliver good value for storage, as well as better processing and network performance over the earlier D2 instance.
- R5b, powered by Intel Xeon Scalable processors: Designed for large relational database workloads that leverage Elastic Block Store (EBS) storage, R5b promises a threefold "instance-to-EBS" performance improvement over the existing R5 instance. "This drives significantly increased performance for large database workloads that process large data sets in memory," AWS said.
In other AWS compute news, the company announced that it is developing much smaller versions of the AWS Outposts product for release sometime in 2021. AWS Outposts lets organizations run the full AWS cloud on-premises using server racks that are delivered and installed by AWS in the location of their choice.
Currently, AWS Outposts comes as a 42-rack unit that's 80 inches tall, 24 inches wide and 48 inches deep. For customers in tight spaces, AWS is developing a 1.75-inch one-rack unit server and a 3.5-inch two-rack unit server. Both new form factors consume considerably less power and network bandwidth than the current AWS Outposts product.
Finally, AWS is expanding its footprint of "Local Zones." Local Zones are not quite full-blown AWS regions, though they extend AWS region capabilities to population-dense areas (such as Los Angeles). They're designed for latency- and throughput-sensitive applications that require closer proximity to end users than what's already provided by existing AWS regions.
At re:Invent this week, AWS announced the launch of Local Zones in Houston, Boston and Miami, with a dozen more opening next year in Atlanta, Chicago, Dallas, Denver, Kansas City, Las Vegas, Minneapolis, New York, Philadelphia, Phoenix, Portland and Seattle.