• Groq has deployed its LPU inference in Sydney for faster, cheaper AI compute.
• The move supports demand and adds access through GroqCloud and Equinix Fabric.

Groq has extended its AI infrastructure into the Asia-Pacific region, bringing its LPU Inference Engine to Equinix’s data centre in Sydney. This follows earlier rollouts in the US and Europe and gives organisations, developers, and public-sector users across Malaysia and the region closer access to faster and lower-cost AI compute.

The plan includes building one of Australia’s largest high-speed AI inference sites, based on a 4.5MW Groq facility in Sydney. The company says the setup can deliver higher speed and lower cost than traditional GPUs and major cloud services. Through Equinix Fabric, a network service that links organisations to cloud and digital partners, users across Asia-Pacific can connect to GroqCloud with low delay while keeping control over their data.

AI use in the region continues to rise—IDC estimates that AI and generative AI spending in Asia-Pacific could reach US$110 billion by 2028, growing at a compound annual rate of 24 per cent from 2023.

The partnership brings together Groq’s approach to AI inference with Equinix’s global reach. The goal is to help organisations scale AI workloads in a predictable and cost-controlled way.

Canva Co-Founder and COO Cliff Obrecht said the shift marks a moment where new tools can “massively accelerate human creativity.” He pointed to Australia’s growing strength in AI and compute infrastructure and said the company looks forward to helping more people bring ideas to life.

Groq CEO and Founder Jonathan Ross stressed the need for more accessible compute, saying, “The world doesn’t have enough compute for everyone to build AI,” which is why the company is expanding access through its work with Equinix.

Scott Albin, Groq’s General Manager for APAC, said the region already accounts for more than half of the company’s global developer base. He added that the Sydney deployment marks the first step in bringing high-performance and cost-efficient inference closer to local users, with a focus on secure, low-latency access and support for data sovereignty.

Equinix’s Asia-Pacific President, Cyrus Adaggra, said Groq’s rapid global expansion through Equinix highlights how the company’s ecosystems and global footprint act as a gateway for enterprise AI workflows at scale.

Groq’s LPU Inference Engine sits at the centre of this effort. The chip is supported by a custom compiler and is built for high throughput and low latency when running next-generation open-source language models. Unlike GPUs, which were created for graphics and later used for AI, the LPU was designed from the start for inference and language tasks, which Groq says helps with speed, cost control, and energy use.

Equinix Fabric plays a key role as well, where it gives organisations a direct, on-demand way to connect to cloud and digital partners across more than 270 data centres in 37 countries. Through this setup, Groq’s compute can be reached from many markets without adding new physical infrastructure.

GroqCloud remains the main access point for the company’s technology, and enterprises can choose between cloud access or on-premise deployment, depending on their needs. The companies say this flexibility will make it easier for teams across Asia-Pacific to adopt and scale AI systems without building a full global footprint on their own.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.