An exclusive Tech Tribune Q&A with Jonathan Ross (founder and CEO) of Groq, which was honored in our:
Tell us the origin story of Groq – what problem were you trying to solve and why?
Democratizing compute because we saw the performance to cost projections of current generation processors hurting the potential of future innovation in AI and ML.
What was the biggest hurdle you encountered in your journey?
Talent density is at the core of our business. This is a rapidly growing market category and everyone’s competing for the best talent. On top of that, choosing to create a radically new architecture rather than continuing to iterate on the challenges of legacy architectures has meant more development.
What does the future hold for Groq?
Moving into the era of deterministic compute, Groq should help break barriers that complement all levels of compute. With workloads that utilize CPUs and GPUs for AI, ML, and HPC as their end-to-end system, Groq has already begun demonstrating how we can help with key areas of complex workloads. Just as one example, look at what we’ve done with Argonne National Laboratory regarding drug discovery. Our future right now will be a next level solution that accelerates former accelerators, like GPUs. Anywhere a CPU and GPU are at the heart of an AI, ML, and HPC workload, a GroqChip™ can be an optimal element that frees up other processors to do more of what they are good at, while utilizing Groq for what we’re best at – deterministic compute.
What are your thoughts on the local tech startup scene in Mountain View?
Hardware is the new software in Silicon Valley. Software companies have been soaking up talent and doing a lot of the innovating for nearly a decade, with a smaller percentage going to the hardware side of tech. Now, we are seeing more software developers get involved with hardware teams at all parts of the stack. For us, this is key because we are focused on developer velocity and taking the pain points out of the developer experience.
What’s your best advice for aspiring entrepreneurs?
The quality of people you involve in your project is always going to be the most important decision you make. Really skilled people aren’t enough. You need to think about the culture of your company, and what you all provide one another as you grow. Also, avoid following the path of the incumbent. They’ve already done everything on that journey, which means they already have a really good idea of everything you will do. Learn about it, but then find the shortest path to your own goal. Accept that it will be hard, but that what you’ll learn will be greater than what a path of convenience will teach you.