Hello Product Hunt! I’m Jonathan from TensorDock. After 7 months in beta, we’re finally launching Core Cloud, our platform to deploy GPU virtual machines in as little as 45 seconds! https://www.tensordock.com/produ... 🤔 Why? Training machine learning workloads at large clouds can be extremely expensive. This left us wondering, “how did cloud ever become more expensive than on-prem?” I’ve seen too many ML startups buy their own hardware. Cheaper dedicated servers with NVIDIA GPUs are not too hard to find, but they lack the functionality and scalability of the big clouds. We thought to ourselves, what if we built a platform that combines the functionality of the large clouds but made it priced somewhere between a dedicated server and the large clouds? That’s exactly what we’ve done. 👩‍💻 Built to make engineers more productive - 3 machine learning images so you can start training ML models in 2 minutes, not 2 hours - We provide a REST API, so you can integrate directly your code with ours - There’s also a CLI you can use to manage your servers directly via command line 💪 The feature set only large clouds can match - Storage-only billing when the VM is stopped (for only $0.073/GB/month) - Edit virtual machines after they’re created to downsize costs. If you provision a NVIDIA A6000 and find out you’re only using 50% of it, stop the VM, modify it to a NVIDIA A5000, and you’ll be billed the lower hourly rate without needing to recreate your server and migrate data over! - 3x-replicated NVMe-based network storage. Seriously fast - 10 Gbps networking - 3 locations (New York, Chicago, Las Vegas), and more coming soon! - Custom alerts - Billing pro-rated to the minute 🤑 How cheap are we? - CPU-only servers from $0.027/hour - NVIDIA Quadro RTX 4000s from $0.29/hour - NVIDIA Tesla V100s from $0.52/hour … and 8 other GPU types that let you truly right-size workloads so that you’re never paying for more than you actually need ✏️ Example use case of our platform: 1) Spin up a CPU-only server for $0.027/hour to upload all of your data 2) Convert it to a GPU server to train your ML workload for $0.282/hour 3) Convert it back to a CPU-only server to analyze the results for $0.027/hour 4) Stop the server but keep the data for only $0.073/GB/month in storage costs 5) Start the server back up again, and train your next workload! 🎉 Celebrate our launch with us with $1 in free credit! Yes, we sound cheap… but $1 is all you need to get started with us! That’s more than 3 hours of compute time on our cheapest configuration! Use code PRODUCTHUNT_1 on the billing page to redeem this free credit. We’re around to answer your questions, so post them below! Or, email us at hello@tensordock.com :) TensorDock: https://www.tensordock.com/ Product page: https://www.tensordock.com/produ... API: https://documenter.getpostman.co... Community CLI: https://github.com/caguiclajmg/t... Deploy a GPU: https://console.tensordock.com/d...

TensorDock GPU Cloud

Cloud GPUs from $0.29/hour. Cost-effective. API. Scalable.

Cloud GPUs from $0.29/hour. Cost-effective. API. Scalable.