Hey everyone! I’m building a small startup called GPU Signals — think of it as an “Airbnb for GPUs.” It tracks rental platforms like RunPod and Vast and helps users spot and secure high-demand GPUs like A100s and 4090s before they disappear.
The problem we’re tackling is simple but frustrating: finding the right GPU at the right time without sitting around refreshing pages or losing out to someone faster.
Here’s what we’re building:
• Real-time alerts for specific GPU types (via Discord, email, or webhook)
• Custom filters for price, location, reliability, uptime, etc.
• Historical availability trends so you can plan ahead
• Predicted network congestion — see when demand spikes and plan around it
• Auto-claim (GPU sniping) — optionally auto-book instances the moment they become available, based on your filters
• Host reliability tracking to avoid unstable or flaky nodes
We’re early-stage, but building fast — and we’re putting together a list of early testers to help shape the roadmap.
If you’ve ever struggled with GPU availability or want to automate your rental workflow, we’d love to hear from you.
Check it out here: https://gpusignals.com
Thanks!