Groq
Private
AI Inference Chips
Founded 2016Mountain View, CA~300 employees
Valuation
$2.8B (2024)
Total Funding
$640M
Profit Status
Pre-profit
Employees
~300
About Groq
Groq is a chip startup that has developed the Language Processing Unit (LPU), custom silicon designed specifically for LLM inference. Their hardware delivers inference speeds 10-25x faster than GPU-based alternatives.
Company History
Founded in 2016 by Jonathan Ross, a former Google TPU architect. Burst into public awareness in early 2024 with API demonstrating 500+ tokens/second for Llama and Mixtral.
AI Models & Benchmarks
LPU (Language Processing Unit)
HardwareCustom ASIC for deterministic, low-latency LLM inference. 750GB/s memory bandwidth.
2023
Benchmarks
Llama 3.1 70B speed250 tok/s (vs ~40 tok/s GPU)
Mixtral 8x7B speed575 tok/s
Latency<0.5s first token
Products & Services
LPU Inference EngineGroqCloudGroq API
Company Details
Focus Area
AI Inference Chips
Stage
Private
Founded
2016
Headquarters
Mountain View, CA
Employees
~300
Financial Overview
Total Funding
$640M
Valuation
$2.8B (2024)
Profit Status
Pre-profit
Funding Rounds
- Series C$300M2021Tiger Global, D1 Capital
- Series D$640MAug 2024BlackRock, Neuberger Berman
Key Investors
BlackRockTiger GlobalBatCapitalNeuberger Berman
Key People
- Jonathan Ross (CEO/Founder)
- Douglas Wightman (CFO)
Competitive Landscape
NVIDIACerebrasSambaNovaTenstorrent
Tags
Hardware
Inference
Chips
Infrastructure
Fast