Posted inCloudInfrastructure

AWS re:Invent: 4 new innovations to power high performance computing

“At AWS we never compromise on security and we’re laser-focused on lowering costs,” says Peter DeSantis, SVP, Utility Computing

Peter DeSantis, Senior Vice President of AWS Utility Computing

Keeping up with tradition, Amazon Web Services (AWS) held its ‘Monday Night Live’ keynote with Utility Computing senior vice president Peter DeSantis at AWS re:Invent.

The session gave a glimpse of what’s underneath the hood and offered a picture of the ‘how’ behind AWS’s cloud innovations.

“The ‘how’ is the way we deliver some of the most important attributes of the cloud including –security, elasticity, performance, cost, availability and sustainability,” said DeSantis.

He added, “These are fundamental to how we build our services and they accrue from long-term investments.”

Throughout his keynote, DeSantis focused on how AWS is ensuring high performance without compromising security and increasing costs.

“There are typically easy ways to get good performance if you’re willing to compromise on security or to blow up your cost structure. But at AWS, we never compromise on security and we’re laser-focused on lowering costs,” he said.

In line with this promise, DeSantis announced several new updates and launches aimed at accelerating AWS’s solutions and services:

AWS Nitro v5

AWS announced the latest version of its Nitro system the underlying platform for its EC2 instances.

Nitro v5 is the latest iteration in the firm’s proprietary data processing unit (DPU). DeSantis noted that this new version built by its Annapurna Labs team has about twice the transistors than the previous generation nitrogen, a PCIe adapter that provides twice the bandwidth and 50 percent higher memory speeds. This means that Nitro v5 delivers 60 percent higher performance with lower latency.

Graviton3E

Annapurna Labs Amazon Web Services Photo Copyright Noah Berger / 2022

Designed for high-performance workloads, AWS introduced the new Graviton3E CPU. It is the latest addition to the company’s Graviton line. According to the AWS utility VP, the chip offers significant performance improvements, including a 35 percent increase in workload performance that’s heavily reliant on vector instructions.

The new CPU sets new benchmark performances for healthcare and financial modelling workloads.

New EC2 instances

Powered by Nitro v5 and Graviton3E, DeSantis also revealed three new EC2 instances – C7gn, R7iz, and Hpc7g.

  • C7gn – powered by the AWS Graviton3E, this instance supports 200 Gbps of network bandwidth. It is designed for the most demanding network-intensive workloads such as virtual network appliances and data analytics. It will be available in multiple sizes with up to 64 vCPUs and 128 GiB of memory.
  • Hpc7g – also powered by Graviton3E, this new instance is will be available in multiple sizes with up to 64 vCPUs and 128 GiB of memory. It is designed to give companies the best price/performance for tightly coupled compute-intensive HPC and distributed computing workloads.
  • R7iz – this instance will be available in multiple sizes with up to 128 vCPUs and 1 TiB of memory. It is powered by the latest 4th generation Intel Xeon Scalable Processors (code-named Sapphire Rapids) and runs at a sustained all-core turbo frequency of 3.9 GHz. With high performance and DDR5 memory.

AWS Lambda SnapStart

AWS announced a small but significant update to Lamba, its serverless, event-driven compute service. Lamba enables users to run code for virtually any type of application or backend service without provisioning or managing servers.

In his keynote, DeSantis introduced Lambda SnapStart, a new performance optimisation that can significantly improve the startup time for applications. It can deliver improved performance of up to 10x faster startup times for latency-sensitive Java applications with minimal or no code changes required.