Groq, a startup that develops chips to run generative AI models faster than traditional hardware, has its eye on enterprises – and the general public sector.
Today Groq announced that it’s launching a brand new division – Groq Systems – focused on significantly expanding its customer and developer ecosystem. Groq Systems' purview covers organizations, including government agencies, that wish to add Groq's chips to existing data centers or construct recent data centers with Groq processors.
With the creation of the brand new entity, Groq has acquired Definitive Intelligence, a Palo Alto-based company that gives a variety of business-focused AI solutions, including chatbots, data evaluation tools and documentation creators. Sunny Madra, CEO of Definitive Intelligence, now leads GroqCloud, Groq's cloud platform, which provides Groq hardware documentation, code samples, and self-service API access to the corporate's cloud-hosted accelerators.
“At Groq, we’re committed to creating an AI economy that’s accessible and inexpensive to anyone with a superb idea,” Groq co-founder and CEO Jonathan Ross said in a press release. “We are pleased to welcome Sunny and his team at Definitive Intelligence to assist us achieve this mission…The Definitive team has expertise in AI solutions and go-to-market strategies, in addition to a proven commitment to sharing knowledge with the community. ”
Madra founded Definitive Intelligence in 2022 with Gavin Sherry, former technical director of EMC. Prior to launching Definitive Intelligence, Madra and Sherry co-launched Autonomic, a cloud-based platform for connecting mobility systems that Ford acquired in 2018.
Definitive Intelligence offers several business-focused GenAI products, including OpenAssistants (a group of open source libraries for developing AI chatbots) and Advisor (a visualization generator that connects to each enterprise and public databases). One of Definitive's leading tools is Pioneer, an “autonomous data science agent” designed to handle various data evaluation tasks, including predictive modeling.
Prior to the acquisition, Definitive Intelligence had raised $25.5 million in enterprise capital.
“The world is just now realizing how essential high-speed inference is to generative AI,” Madra said in an emailed statement. “At Groq, we offer developers with the speed, low latency and efficiency they should deliver on the promise of generative AI. I’ve been a giant fan of Groq since I first met Jonathan in 2016, and I’m excited to support him and the Groq team of their quest to bring the fastest inference engine to the world.”
Groq, which emerged from obscurity in 2016, is developing a so-called LPU (Language Processing Unit) inference engine. The company claims its LPU can run existing large language models with an analogous architecture to OpenAI's ChatGPT and GPT-4 at 10x speed.
Ross's claim to fame helps invent the Tensor Processing Unit (TPU), Google's custom AI accelerator chip used to coach and run models.
Definitive Intelligence is Groq's second acquisition after high-performance computing and AI infrastructure solutions company Maxeler Technologies in 2022. It will not be the last. The marketplace for custom AI chips is very competitive, and – to the extent that the ultimate purchase makes Groq's plans clear – Groq is clearly wanting to gain a foothold before its competitors have a probability.
