Groq partners with Equinix to launch local AI inference capability

Groq is expanding its footprint with high-speed service for Australia and the region.

Groq has partnered with data centre provider Equinix on a 4.5MW high-speed AI inference infrastructure facility in Sydney. It represents Groq’s first footprint in Asia-Pacific, serving Australia and the region, and follows launches in the US and Europe.

There are more than two and a half million developers using Groq Cloud and nearly half of those originate from the Asia-Pacific region, highlighting there’s a real appetite for inference services in Australia and beyond.

“The world doesn’t have enough compute for everyone to build AI. That’s why Groq and Equinix are expanding access, starting in Australia,” said Jonathan Ross, CEO and founder, Groq.

Groq uses LPUs (language processing units), custom chips designed specifically to run large language model (LLM) inference. Groq said it’s offering lower cost compute power than traditional GPUs and hyperscaler clouds that’s up to five times faster.

With Equinix Fabric, a software-defined interconnection service, organisations can access secure, low-latency interconnectivity and access to GroqCloud for production AI workloads with data sovereignty assured.

“Organisations in Australia and the region can scale their AI workloads faster, more cost effectively and in compliance with local regulations, especially Australia's data sovereignty requirements,” said Guy Danskine, MD, Equinix Australia.

Groq is already working with customers across Australia, including Quantium and Canva, to deliver inference solutions tailored to their business needs, including improving customer experiences and employee productivity. More partner announcements are expected in the coming days.

“Australia is in a unique position to become a global hub for AI inference, including government support for AI, is the power of this new partnership for driving customer outcomes,” said Danskine.

Data science company Quantium, a Groq partner, helps organisations with the “last mile” which is harnessing the models and the infrastructure built by others to help them hit their business goals.

Speed and reliability are vital metrics, but data sovereignty is another critical consideration for business applications, said Victor Bajanov, AI transformation executive, Quantium.

“A lot of our customers are concerned about data sovereignty and it’s important that this keeps all of the data in Australia,” Bajanov said.