Infrastructure for Artificial Intelligence Workloads

The GPU revolution has dramatically increased processing power of PC’s and servers, and the GPU’s massive parallel compute capabilities have made them an ideal foundation for all manner of artificial intelligence workloads – from simple machine learning to deep learning on simulated neural networks. XENON introduced GPU computing to Australia and built the first GPU cluster. Let us help you right-size your artificial intelligence infrastructure.

Proof-of-Concept projects can be run on smaller PC’s or workstations like the DEVCUBE or DGX Station, or in the public cloud if GPU’s are available. These small workstations form factors are ideal for individuals or small teams, running quiet enough to operate in an office environment. They are ideally suited for experimentation and model refinement individually or in small teams. Scaling up in size and processing power, dense, Mutli-GPU server solutions create petaflops of processing power and are useful as datasets and models grow. Integrated design within these larger DGX-1 and DGX-2 units allows for optimal data transfers and makes the most of the power of linked GPU’s in a single chassis. One you hit this scale, single racks of DGX units can replace rows of racks of servers, and decrease data set processing time by factors of 100 or more. In addition to this centralised artificial intelligence infrastructure, small form factor GPU’s like the NVIDIA Jetson are now available that transfer your AI models to the edge.

Containerisation of the artificial intelligence infrastructure stack allows for quick and easy set-up and scaling of your artificial intelligence process across any of these form factors. This greatly speeds experimental iteration time, and makes scaling up truly quick and easy.

Artificial Intelligence Infrastructure Consulting

XENON’s deep experience in artificial intelligence infrastructure allows us to provide you with unique insights into how to plan your AI infrastructure to best optimise your AI journey. The correct advice at the right time will save you money, and more importantly accelerate your return on investment, as your models evolve quicker into applied AI applications.

Learn How to Get Started in AI

If you are early in your AI journey, Download this eBook from NVIDIA and XENON – How to Get Started in AI. It explains how the software stack scales across the NVIDIA platforms, the learning models, and how to start your AI journey. A great resource that will help you make sense of the AI landscape.

Talk to a Solutions Architect