Signalbase · Jul 2026
Signalbase reported that Thunder Compute secured $15.7M in Series A funding, describing the company as a systems research lab with a commercial focus on GPU virtualization.
Company research
Public company, workplace, funding, and market signals
Overview
Thunder Compute is a GPU cloud and systems-research company that says it is building the virtualization layer for GPUs, turning cutting-edge GPU virtualization research into a developer-friendly cloud offering for AI/ML workloads. Public sources describe it as a low-cost, on-demand GPU platform with enterprise and partner-fleet use cases; the company’s public profile pages conflict on some administrative details such as founding year and HQ, but the official site and YC present it as a San Francisco-based company originating in 2022.
Primary product
On-demand cloud GPU instances with GPU virtualization (GPU-over-TCP) for AI/ML workloads
Founded
2022
Headquarters
San Francisco, California, United States
Team size
1-10 employees
Industry
Software Development
Sub-industry
GPU cloud infrastructure / cloud-based GPU services
Offices
San Francisco, CA (HQ)
Wilmington, DE
Business model
On-demand cloud GPU instances
Pay-as-you-go usage-based billing
Enterprise GPU virtualization software for partner fleets
Funding
Stage
Series A
Total raised
$20.7M
Latest round
Series A · Jun 2026
Latest amount
$15.7M
People & investors
Leadership
Co-Founder & CEO
Co-Founder & CTO
Founding Systems Software Engineer
Work & culture
Research-heavy, systems-lab culture focused on hard infrastructure problems; the team is described as fast-growing and composed of systems researchers and infrastructure engineers from Citadel Securities, Aquatic, AWS, and other startups. Current hiring signals emphasize technical depth, speed, and in-person collaboration in San Francisco.
Compensation
Public job posts advertise full-time roles with generous salary + equity, but no public salary bands or formal compensation grid were found.
Benefits
Health insurance
Dental insurance
Vision insurance
401(k)
Generous holiday schedule
Free lunches
Product & market
Pricing
Usage-based, per-minute billing for on-demand GPU instances; public pricing is quoted per GPU per hour on the pricing page.
Differentiators
- —GPU virtualization / GPU-over-TCP architecture
- —Low-cost on-demand GPUs
- —No data egress fees
- —One-click developer workflow with VS Code/Cursor integrations
- —Persistent storage via snapshots
- —Hardware swapping without redeploying
- —Fast spin-up and restart
- —Security controls including GPU memory wipe/reset between jobs
- —Enterprise-grade infrastructure
- —Partner-fleet capacity optimization
Technology
GPU-over-TCP
CUDA
VS Code
Cursor
PyCharm
Kubernetes
Linux
Terraform
Ansible
Go
Python
PyTorch
Git
Bash
NVIDIA
Customers
Data scientists
Startups
Researchers
Developers
AI/ML teams
Enterprise customers
Partner GPU fleet users
Competitors
RunPod
CoreWeave
Cudo Compute
Crusoe
io.net
Modal
TensorWave
Lambda
Latitude.sh
Genesis Cloud
Scale & contact
Estimated revenue
~$900K estimated ARR (GetLatka estimate, 2025-04-10)
Estimated monthly visits
94.8K
Traffic estimate as of Jul 2026
Latest news
Thunder Compute blog · Jul 2026
Company comparison post positioning Thunder Compute as a low-cost, developer-friendly option versus RunPod and CoreWeave, with explicit pricing comparisons and workflow claims.
TechBullion · Mar 2026
TechBullion reported Thunder Compute raised $4.5M in seed funding led by Matrix Partners, with participation from Y Combinator, Avesta Fund, Preston-Werner Ventures, CyberAgent Capital, Transpose Platform, and Amino Capital.
LinkedIn post by Carl Peterson · Jan 2026
Carl Peterson announced the $4.5M Matrix-led round and thanked Matrix, Y Combinator, Avesta Fund, Preston-Werner Ventures, CyberAgent Capital, Transpose Platform, Amino Capital, and angels.
Thunder Compute blog · Jul 2025
Company engineering post explaining the GPU-over-TCP architecture and how Thunder Compute detaches GPUs from physical servers to improve utilization.
Jobs and careers at Thunder Compute
0 jobs at Thunder Compute
No open positions
Check back later for new openings