Thunder Compute (YC S24) | C++ Systems, Infrastructure, BizOps | San Francisco (Onsite) | Full-time | https://www.thundercompute.com
We're building the VMware for GPUs. Most GPUs in the cloud sit idle a lot of the time because they're bolted to one machine over PCIe. We decouple them.
Our virtualization layer is a userspace shim, loaded through LD_PRELOAD, that intercepts CUDA calls and ships them over the network to a host with a physical GPU somewhere else in the data center. Workloads run unchanged. When a process goes idle, the GPU detaches and gets reassigned, and reattaches in tens of milliseconds when it's needed again. Think Ceph for GPUs: the GPU becomes a network resource a cluster-wide scheduler can pool and pack.
The hard part is making that fast. We're within ~10% of native on most AI workloads, and it lets us serve ~1.8x more users on the same fleet. We run it in production as our own GPU cloud and sell it to enterprises with underutilized fleets.
Team is from Citadel Securities, Aquatic, Old Mission, AWS, and Bain. Series A, $17.5M raised from Matrix and Y Combinator.
Thunder Compute is a GPU cloud for developers built on a proprietary software stack of C++ and GoLang. You will solve hard systems problems in production.
Quant dev / HFT SWE experience preferred, team from CitSec, Aquatic, Old Mission, AWS.
Hiring 4 full-time Software Engineers for low-level C++ systems work (HFT skillset preferred), reporting to the CTO, team from Citadel Securities, Aquatic, Old Mission, Bain
We're a cloud provider using GPU virtualization software to arbitrage GPU prices.
We research technology to use GPUs more efficiently than competitors. Our cloud network attaches GPUs over TCP to eliminate downtime. This lets us buy GPU compute, make it more efficient, and resell it at 60%+ profit margins.
We recently closed ~$4.5M of funding to recruit the world's best low-level talent.
If you're cracked at building C++ systems software and looking to do something more meaningful than creating liquidity in the market, reach out.
To apply: email met at [redacted] and mention HN. I read every email.
Hiring 4 full-time Software Engineers for low-level C++ systems work (HFT skillset), reporting to the CTO, team ex-Citadel, ex-Aquatic, ex-Bain
We're a GPU cloud provider making more efficient infrastructure.
To our customers, we are a direct substitute for Runpod, Lambda, Coreweave, etc. with a simple UX and low prices.
We recently closed ~$4.5M of funding to triple our team size with a focus on recruiting the world's best low-level talent.
If you're cracked at building C++ systems software, on noncompete and bored, or looking to do something more meaningful than creating liquidity (you can tell most of our team is from HFT), reach out.
To apply: email met at [redacted] and mention HN. I read every email.
We're a GPU cloud building software to use GPUs more efficiently.
To our customers, we are a direct substitute for Runpod, Lambda, Coreweave, etc. with a simpler UX and lower prices.
Our advantage is software - we have spent 3 years bringing research technology to market which lets us use each GPU more efficiently than competitors. At the same time, our software focus lets us provide a high-quality, industry-leading UX.
In a space where customers shop around and try several providers, having a better product at lower prices has let us quickly expand.
We recently closed ~$4.5M of funding to triple our team size and continue building best-in-class software.
If you're cracked at building C++ systems software (most of our team came from HFT) or kubernetes infrastructure, reach out.
To apply: email met at [redacted] and mention HN to guarantee I see your application.
We're building the VMware for GPUs. Most GPUs in the cloud sit idle a lot of the time because they're bolted to one machine over PCIe. We decouple them.
Our virtualization layer is a userspace shim, loaded through LD_PRELOAD, that intercepts CUDA calls and ships them over the network to a host with a physical GPU somewhere else in the data center. Workloads run unchanged. When a process goes idle, the GPU detaches and gets reassigned, and reattaches in tens of milliseconds when it's needed again. Think Ceph for GPUs: the GPU becomes a network resource a cluster-wide scheduler can pool and pack.
The hard part is making that fast. We're within ~10% of native on most AI workloads, and it lets us serve ~1.8x more users on the same fleet. We run it in production as our own GPU cloud and sell it to enterprises with underutilized fleets.
Team is from Citadel Securities, Aquatic, Old Mission, AWS, and Bain. Series A, $17.5M raised from Matrix and Y Combinator.
Open roles:
- Software Engineer, C++ Systems ($215K-$325K + meaningful equity). Low-level, latency-sensitive work on the virtualization layer. Quant dev / HFT background preferred: https://jobs.ashbyhq.com/thundercompute/2efae53b-817c-43e7-9...
- Software Engineer, Infrastructure ($150K-$250K + meaningful equity). Go, Kubernetes, reliable systems at scale: https://jobs.ashbyhq.com/thundercompute/927a922b-c607-4b0e-a...
- Business Operations Lead ($150K-$225K + meaningful equity): https://jobs.ashbyhq.com/thundercompute/4f28e418-b77b-446e-a...
If you're cracked at building systems software and want to do something more meaningful than creating liquidity, reach out.
Email me at [redacted] and mention HN. I read every email.