As a dedicated simulation solutions provider, Ansys leads the industry in physics-based and multiphysics simulation, supporting transformational product development for customers across industries.

“As engineers and designers face increasingly complex problems, cloud computing helps lower the barrier of access to high-performance computing, allowing users to solve problems faster. Ansys has also been focusing on green computing initiatives with the goal of improving energy efficiency and reducing costs to customers. We benchmarked 35% better performance on AWS Graviton3-based C7g instances compared to the previous generation instances. With the support of LS-DYNA on the AWS Graviton3 processor, Ansys customers will get the best of both worlds – access to a world-class multiphysics solver without comprising on speed, and lower energy and costs.”

Prith Banerjee, Chief Technology Officer - Ansys

Beamr is a leading provider of image and video optimization solutions that enable professional photographers to improve their workflows, photo sharing services to improve UX and reduce churn, and video service providers to reduce storage and delivery costs.

"Beamr's JPEGmini software, written in C/C++, optimizes JPEG image files by reducing their file size without compromising quality. The application is a compute-intensive and includes functions for image decoding, image encoding, and a quality measure algorithm that analyzes various image attributes. Since the mobile version of this software runs on Arm processors, we decided to test its performance on AWS Graviton3-based Amazon EC2 C7g instances. Rebuilding our software to run on C7g instances took us one working day, and the results were promising. When running on C7g instances, we saw 30% improved performance over comparable x86-based instances. Based on these results, we plan to recommend to our customers to run the Beamr JPEGmini software on Graviton3-based instances once those become GA, and we plan to benchmark Beamr’s H.264 and HEVC video encoders on Graviton instances as well."

Dan Julius, VP R&D - Beamr

Honeycomb.io develops an observability platform that allows engineering teams to visualize, analyze, and improve cloud application quality and performance.

“We're excited to have tested our high-throughput telemetry ingestion workload against early preview instances of Graviton3 and have seen a +35% performance increase for our workload over Graviton2. We were able to run 30% fewer instances of C7g than C6g serving the same workload, and with 30% reduced latency. Once this instance type reaches production, we believe we'll see over 50% price-performance improvement over Intel instances overall.”

Liz Fong-Jones, Principal Developer Advocate - honeycomb.io

Modern Electron is a technology company focused on the decarbonization of heat and power. Our power products generate electricity from heat in furnaces and boilers; and our gas decarbonization products generate clean hydrogen for fuel blending in commercial and industrial heating.

"We use the simulation software WarpX, part of the Exascale Computing Project, to simulate our novel heat-to-electricity devices. In 2020, we began using Graviton2-based C6g instances to realize a 50% cost improvement over C5 instances. Recently we tested C7g instances with simulations we are currently running on C6g, and have seen computation speed improvements of another 50%. This gives us a significantly lower time-to-solution that greatly improves our rate of learning."

Peter Scherpelz, Senior Computational Physicist - Modern Electron

The National Supercomputing Centre (NSCC) Singapore was established in 2015 and manages the nation’s petascale facilities. NSCC’s high performance (HPC) resources support the research needs of Singapore’s public and private sectors, including research institutes, institutes of higher learning, government agencies and companies.

“As part of a trial we ran Gromacs benchmarks on AWS Graviton3-based Amazon EC2 C7g instances for molecular dynamics simulations and found that it outperforms comparable x86-based instances by 15-20% and the previous generation of C6g powered by AWS Graviton2 instances by 40%-50%. The benchmarking took just two days and was achieved without any specific performance tuning efforts using solely GCC 11.2.0 and ARM performance libraries. We are impressed with the performance AWS Graviton3 offers for HPC workloads.”

Jernej Zidar, Senior HPC Application Analyst - NSCC

ProSiebenSat.1 Media SE is a German digital and media company focused on entertainment, dating, commerce, and other ventures. Its wholly owned subsidiary, Seven.One Entertainment Group, offers viewers entertainment content across a variety of media technologies.

"The Seven.One Entertainment Group runs video transcoding, based on the open source tool ffmpeg, for on-demand content in Amazon EC2. In 2021, we switched to AWS Graviton2-based C6g instances and realized significant cost benefits over comparable x86-based instances. With C6g instances, we could also run more parallel transcoding tasks per EC2 instance, resulting in fewer instances and more cost savings. We have run transcoding tests with ffmpeg version 4.4.1 on the new AWS Graviton3-based C7g instances and benchmarked up to 40% faster performance compared to C6g instances. With a new build of ffmpeg version 5 using Arm64 optimizations, there are even greater performance gains, especially in near-live use cases. This enables to provide video on-demand content faster to our users."

Wolfgang Dworzak, Head of Video Engineering - Seven.One Entertainment Group GmbH

Tuya Smart offers a cloud platform that connects a range of devices via the IoT to make lives smarter.

“Our application is network intensive with a large number of persistent connections. We were amazed to see a 40% performance improvement for this application on AWS Graviton2-based Amazon EC2 C6g instances compared to x86-based instances. We have been running most of our production workloads on Graviton-based instances for over 1.5 years now. After the launch of AWS Graviton3-based Amazon EC2 instances, we benchmarked 15% performance improvement on general tasks and 50% on encrypt/decrypt intensive tasks compared to the C6g instances. We are looking forward to adopting the new C7g instances in production soon.”

Simon Shu, Senior Platform Technical Director - Tuya Smart

Viant is a leading advertising software company enabling ad buyers to plan, create, execute and measure their digital advertising investments through a cloud-based platform.

“We are extremely excited about AWS Graviton3 performance. As our application heavily relies on CPU vectorization, AWS Graviton3 SVE comes with substantial performance improvement over the AWS Graviton2 Neon extension. On production testing, we’ve seen ~45% performance improvement for heavy CPU workloads on AWS Graviton3-based Amazon EC2 C7g instances compared to AWS Graviton2-based C6g instances. AWS Graviton3 is going to be pivotal component for growing our platform and reducing operational cost.”

Adrian Witas, SVP Chief Architect - Viant