Remote GPU system benchmarking for edge AI projects

14-03-2023 | Impulse Embedded | Test & Measurement

Impulse Embedded now allows customers to test their code on a GPU computing system before purchasing project hardware and deploying AI project at the edge.

The Remote Benchmarking service enables developers to test their model and application code and choose a hardware platform that scales with end users. It is adaptable to aim at environments that will reduce development costs rising during the hardware selection and testing process.

Users of this service will be provided credentials to remote login and deploy their code to any of the test systems to determine which platform will be most appropriate for their project. This will save time and money in the hardware selection process and enable faster time to market.

The company offers customers access to a broad range of development systems from the entry-level Jetson Nano to the latest AGX Orin and servers qualified for housing multiple GPU cards. The opportunity to test AI code on various platforms saves time and money and gives insight into what device works best for the application.

Using the knowledge gained from testing, the company can use its industry insight and technology know-how to supply the ideal solution considering I/O requirements, environmental extremes and any industry-specific certifications that may be required.

Protex AI uses AI-powered technology to give businesses more significant visibility of unsafe behaviours in their facilities. The privacy-preserving platform plugs into existing CCTV infrastructure and utilises its computer vision technologies to capture unsafe events autonomously in settings, including warehouses, manufacturing facilities and ports.

"At Protex AI, we require edge devices to deploy an on-premises, secure platform to our clients. Some of these clients require tailored solutions with a range of compute demands. We use Impulse Embedded to supply purpose-built hardware for our deployments. Impulse's remote testing platform has played a valuable part in helping us meet our hardware needs, they are quick to get us up and running on a remote system to benchmark the latest devices, thus enabling us to select the best device for our needs," Matt Moynihan, Protex AI.

By Seb Springall