Groq is a systems company simplifying and reducing compute to its essential elements, everywhere. Committed to providing machine learning at push-button performance, Groq’s compiler-led design approach enables customers to easily deploy scalable, deterministic, highest performance per watt systems. As we transition to a machine learning driven environment, our vision is to provide the world’s compute.
Groq is a well-funded, rapidly growing AI systems company in stealth mode addressing the need for high performance, low latency inference to understand and categorize the torrent of data generated by individuals, machines and IoT devices. Groq’s software and hardware addresses a market that is expected to reach $66B by 2025. Groq systems will enable hyperscalers to deploy significantly more TFLOPS per datacenter and will help drive autonomous vehicles and expedite the growth of 5G networks.
Groq’s state of the art acceleration technology was created by engineers experienced in designing and building world class AI accelerators, including the first Google TPU. We are growing quickly and building an exceptional engineering team to meet our aggressive performance roadmap.
Our focus is on providing Groq customers with easy to use, high performance AI hardware and software with sub-millisecond latency for a variety of ML workloads, along with a few compelling features that will not be available on any other ML platform. Our customers avoid lock-in with our open source drivers and libraries and industry standard frameworks, including TensorFlow.
We are a collaborative group of senior engineers organized into functional areas solving hard problems.
We design acceleration hardware and software tools to tackle the most demanding ML/DL workloads. Building a compiler for our unique hardware architecture offers a new set of difficult problems to solve. We love it!
Optimize computation graphs for efficient execution our unique hardware architecture.
Build visualizations of compiled program executions and resource allocations.
Implement a cycle accurate software simulator used for debugging compiled programs.
Though we encourage face-to-face contact to build rapport with team members, we believe in work/life balance. If you can't be in the office because you have to meet the cable technician at home, or you have a parent/teacher conference, or you have to take Fido to the vet, just let the team know via Slack and feel free to work from home that day.
Groq provides up to $25 every day for lunch. We order from Forkable and lunch is delivered everyday. We all enjoy gathering at the table and eating our meals together.
Groq pays 100% of the premiums for employees and 80% for dependents. How cool is that?!?
As long as you're getting the work done and informing the team via Slack. We just ask that you don't abuse this privilege.
We have a 401K plan, but currently do not have a corporate match. As a startup, we've decided that our VC funding is best put towards the creation and development of our product.
Groq will provide up to 6 weeks of paid maternal/paternal leave.
We offer 20 days of paid vacation/year, unlimited paid sick days!
We occupy the whole top floor of a six-story building with gorgeous 360-degree views and access to the wrap-around balcony.
Interested in this company?
Skip straight to final-round interviews by applying through Triplebyte.