Company description
Thunder Compute is building VMWare for GPUs. We are a 3-person team, backed by Y-Combinator, and seed funded.
Our GPU virtualization software eliminates idle GPU cycles. This occurs through a mix of sharing and oversubscription.
The core software network-attaches GPUs over IP through a form of API remoting. This lets us allocate compute wherever it is needed most. This is a hard technical challenge: we are building a complicated, low-level system with extremely strict performance requirements.
Job description
While the problem is tough, the opportunity is equally large: we estimate well over $100B. In this systems engineering role as our fourth team member, your job is to help us build a virtualization layer with near-native performance.
If you are a top 0.1% systems engineer with a strong belief in your technical ability, here is a rare chance to directly unlock massive opportunity through technical execution.
Core skills
- Experience optimizing performance of NICs and/or GPUs
- Top 0.1% C++ ability
- Strong understanding of low-level networking, compilers, or similar
- Experience working in latency-sensitive environments
Soft skills
- High potential to grow into a senior leadership position as the company expands
- Communication
- Self-motivation (we expect you to drive your own work and contribute from day 1)
Preferred pre-requisites
- Ph.D. or Masters degree in Computer Science from a top ~50 U.S. university, 3.5+ GPA (Bachelors required)
- 1+ year of industry experience at a hedge fund, hardware company (e.g., NVIDIA, Samba Nova), or similar preferred
Compensation
- This roll is full-time, in-office, salaried with stock options (~1-1.5% ownership).
- We are willing to move our headquarters to find the right person for this role.
- You will report directly to Co-founder and CTO Brian (ex-citadel quant dev).
- Thunder GPU, Inc. is an equal opportunity employer, and adheres to all applicable employment laws.
