ActivePort Group Limited (ASX:ATV) has released ActivePortal Compute, the company’s private cloud orchestration software solution for high-performance, GPU-centric computing.
Graphics Processing Units (GPU’s) are at the heart of high-performance computers used to power advanced technologies including artificial intelligence, machine learning, cloud gaming, virtual reality and more.
ActivePortal Compute is a software system used to create and manage private clouds of GPUcentric servers that can be seamlessly integrated with virtual GPU servers at cloud providers like AWS and Azure.
Using ActivePort’s newest release, customers can build private clouds anywhere in the world and connect them via ActivePort’s Global Edge SD-WAN network.
Adding private cloud orchestration for GPU-centric computing opens up new revenue generating opportunities for ActivePort in markets such as artificial intelligence, machine learning, cloud gaming, virtual reality and more.
The company is currently conducting performance trials with several technology solution providers globally and is already tapping into the cloud gaming market which is expected to grow at a CAGR of 48.2 per cent between 2021 and 2027, to reach a total global market size of US$7.2 billion.
Mark Middleton, ActivePort’s founder and CTO, said that in response to accelerating market demand for GPU-centric computing solutions, the company has added a new feature-set to its orchestration software that automates creation of private cloud systems, compatible with public cloud and integrated with global SD-WAN services.
ActivePortal Compute helps users bring their GPU-centric private cloud solutions to market faster and manage them globally from a single, central console.
Public cloud compatibility means ActivePortal Compute customers can easily move to a hybrid GPU cloud without having to re-engineer their automation and orchestration scripts, saving them time and money as they transition to this new platform.
Mr Middleton said that what’s more, users can take advantage of the network orchestration features in ActivePort’s software to integrate their hybrid cloud using software-defined wide-area networking (SD-WAN) for global connectivity.
GPU’s were originally used in game consoles and gaming computers but also in dedicated computer clusters for high-intensity computational work-loads.
Public cloud providers including AWS (Amazon) and Azure (Microsoft) have, in recent years, partnered with the two largest global producers of GPU’s, AMD and NVIDIA to provide GPU-centric compute in the Cloud.
As use of these services grew, a new market for private cloud instances that can be faster and cheaper opened up, where larger consumers can purchase dedicated clusters of GPU servers on a pay-per-use basis.
The release of this new product is the culmination of almost two years of R&D. GPU’s capable of running multiple users on a single card and streaming their output across a network are relatively new and they paved the way for GPU clouds to be possible,” Mr Middleton said.
“We have extended our software to make it simple to deploy and manage GPU servers, integrated by our SD-WAN, all from one screen.
“This new ActivePortal Compute orchestration software solution sets the Company up for significant new near-term revenue opportunities in fast-growing markets such as artificial intelligence, machine learning, cloud gaming, virtual reality and more.”