Groq closed a $350 million Series A led by Disruptive, valuing the restructured inference cloud business at $3.5 billion and completing roughly $1 billion raised in two months after its earlier pivot and talent shift.
Groq’s $350 million Series A led by Disruptive with planned participation from NVIDIA, values the company at $3.5 billion. Combined with the $650 million raised in June 2026, this brings recent funding to $1 billion in roughly two months. The round is described as a Series A under the company’s restructured form and remains subject to customary closing conditions.
Groq, founded in 2016 by former Google engineer Jonathan Ross (a key figure in TPU development), originally focused on custom AI inference hardware, its Language Processing Units (LPUs), positioned as a high speed, low cost alternative to NVIDIA GPUs for running trained models. It reached a peak valuation of $6.9 billion in a $750 million Series E in September 2025, led by Disruptive with participation from BlackRock, Neuberger Berman, Deutsche Telekom Capital Partners, Samsung, Cisco, and others.
In December 2025, Groq entered a non exclusive licensing agreement with NVIDIA covering its inference technology. As part of that arrangement, founder/CEO Jonathan Ross, President Sunny Madra, and other key personnel joined NVIDIA. Reports have described the broader transaction as involving substantial consideration (around $20 billion in some accounts for related assets/technology access). NVIDIA later incorporated elements of Groq’s technology into its LPX platform.
Post deal, Groq repositioned from a primary chip developer to an AI inference “neocloud” / data center operator. It focuses on deploying and operating high performance inference infrastructure (including NVIDIA accelerated computing alongside its own technology) to meet surging demand for running AI models at scale. The June 2026 $650 million growth round (led by Disruptive and Infinitum, with existing investors reinvesting) funded this transition, leadership rebuild, and capacity expansion; the valuation was reset but not publicly disclosed at the time.
The August 2026 $350 million round formalizes the $3.5 billion valuation for the evolved entity and injects further capital for scale.
How will Groq use the funds?
Proceeds support customers seeking medium and larger sized clusters of NVIDIA accelerated computing for both training and inference. Groq plans to expand capacity from 54 megawatts to 200+ megawatts in 2027 (earlier guidance had targeted ~200 MW by end of 2027). It currently operates 13 data centers across North America, Europe, the Middle East, and Asia Pacific.
The company reports serving more than 6 million developers, Fortune 500 enterprises, and thousands of AI native companies, generating trillions of tokens weekly. It is an NVIDIA Cloud Partner, certified to design, deploy, and operate NVIDIA accelerated computing to NVIDIA’s reference architecture and standards. The capital accelerates fit-out of existing footprint with the latest inference technology, including NVIDIA’s LPX systems that incorporate licensed Groq technology.
Valuation Trajectory and Investor Dynamics
- September 2025: $6.9 billion post money.
- Post NVIDIA deal and restructuring: Valuation reset (undisclosed in the June 2026 round).
- August 2026: $3.5 billion (approximately half the prior peak).
This constitutes a down round relative to the 2025 peak, reflecting the talent and technology licensing shift rather than pure operational distress. Existing backer Disruptive (Dallas-based growth firm; founder Alex Davis serves as Groq Executive Chairman) led both recent rounds and has been a major long term supporter. NVIDIA’s planned participation signals continued commercial alignment after the licensing deal.
Recommended: Malachyte Raises $10 Million In Seed Funding Round
Earlier funding included a $640 million Series D in 2024 at $2.8 billion (led by BlackRock) and significant prior capital that pushed cumulative equity raises into the multi billion dollar range. A separate $1.5 billion Saudi Arabia commitment in early 2025 supported infrastructure expansion in the Kingdom (customer/capacity oriented rather than pure equity).
Inference is increasingly viewed as the larger, more persistent, and more critical layer of AI infrastructure compared with training, as models move into production use across applications. Groq emphasizes speed, cost efficiency, and reliability at scale for this workload. By combining proprietary inference expertise/IP (now licensed) with NVIDIA hardware and its own global data center footprint, the company aims to operate as a specialized inference cloud rather than a pure play chip rival.
Leadership has been refreshed with operational, data center, enterprise software, and platform expertise (additions noted around the June round included figures with experience at Meta, xAI Colossus, and other large scale infrastructure efforts). Executive Chairman Alex Davis has framed the opportunity around inference becoming the dominant AI infrastructure layer and Groq’s focus on supporting major model builders.
The dual $1 billion capital raise in mid 2026 equips the restructured Groq to aggressively expand capacity and serve growing demand from developers and enterprises, while the NVIDIA relationship provides both technology integration and potential co-investment alignment. The lower valuation relative to 2025 reflects the fundamental shift in the company’s asset base and team composition following the licensing and talent moves, even as operational metrics (developer base, token volume, geographic footprint) continue to grow.
Please email us your feedback and news tips at hello(at)techcompanynews.com
