Shadeform (YC S23)

With Shadeform, you can launch GPU instances from any cloud in just a few clicks. Want an A100 instance from Lambda? You got it. Want to schedule a Hyperstack H100 instance for next week? Go for it. Want 8 of them in a cluster from Denvr Dataworks? Not a problem. Any cloud, any instance, and at no extra cost. Visit our website to learn more.

Having a strong and aggressive on call process is critical to ensuring availability and customer experience. It raises awareness to customer pain points and puts pressure on the team to address them. Even at startups, on call should be embraced and prioritized. On call allows start ups to move fast and break things - without breaking the customer experience. At Shadeform (YC S23), we heavily leverage PagerDuty to bring surface issues and reach out to customers both proactively and reactively. Alternatively, here’s the only two contacts that ever call me ??: #oncall #startups

Shadeform is now the first and only GPU provider to offer NVIDIA, Intel, and AMD instances on demand ?? Not sure which one to choose for your workload? Here’s the breakdown: AMD: Introducing the AMD MI300X, courtesy of Hot Aisle Inc., a powerhouse designed for next-gen AI workloads. With its unmatched memory bandwidth and advanced architecture, it’s built for large-scale model training and inference tasks, providing a robust alternative to traditional GPU offerings. Intel: Our Intel Gaudi 2 GPUs are offered through Denvr Dataworks and deliver cost-efficient performance for AI training tasks. Known for excelling in deep learning frameworks like TensorFlow and PyTorch, they provide a unique edge in power efficiency and scalability for training larger models at lower costs. NVIDIA: Perfect for diverse AI workloads, from cutting-edge large language models to real-time inferencing. With on-demand access to NVIDIA H200s, H100s, A100s and more, you’ll benefit from unparalleled performance in deep learning, massive parallel processing, and HPC applications. Ideal for developers prioritizing efficiency and versatility in AI training and deployment. Don’t know which GPU fits your workload? Shadeform makes it easy to compare and manage your GPU resources across cloud providers—all from one platform. Sign up for free and try it out for yourself today! #NVIDIA #Intel #AMD #A100 #H100 #Gaudi2 #MI300X #GPUCloud #GPU #AI #ML #PyTorch #TensorFlow

Who else is tired of the brokerage process for cloud GPU commitments? It has been a pain for providers and customers this year. As a provider, you get a request from a broker with half baked information. After a few back and forths, there’s nothing there and the broker goes silent. As a customer, you deal with multiple brokers and a completely disorganized process. The whole thing is chaos right now with over 5 brokers working on the same deals and complete obfuscation of deal and client information. That’s why we’ve built the Angie’s List for GPUs. Fill out ONE form with the details of your request and get a quote from 15+ providers for your cluster or your reserved instances in 24 hours. No more back and forth with brokers, no more dragged out deal processes. Get connected directly to clouds, close, and get back to building. This space is growing in a rapid and fragmented way. We're here to clean it up for reservations as we have done with on-demand GPUs. If you’re a provider and want to automatically get requests for quotes in your inbox, DM me, and we’ll add you to the list! If you’re a customer, go to shadeform (link in comments) to start getting quotes directly from our providers as soon as today!

Let’s be real, getting quotes for cloud GPU reservations sucks. Nobody enjoys manually reaching out to providers one by one for quotes, and hiring a broker to do it for you is an added cost that more often than not drags the process out even longer. This issue hits home for a lot of our users, so we wanted to make the process easier for everyone involved. We’re excited to share that we’ve built the “Angie’s List” for GPU reservation. Fill out one simple form with your requirements, and get quotes from over 15 leading providers in just 24 hours. This is a cloud procurement intervention. Break up with your broker and get the quotes you want tomorrow.

Love it when a plan comes together! ?? ?? Thank you Shadeform (YC S23) for the shout out below! ??

Having trouble deciding which cloud to use for your AI / ML workloads? We help customers like this one test out GPU configurations across leading cloud providers to see what works for them. The best part? It doesn’t cost you a dime extra. Looks like Denvr Dataworks is turning out to be a solid option! ?? Sign up for free to start comparing instances and clouds today!

Shadeform now supports teams!??? Invite users to a team with the brand new team management tab in the navigation menu. Your team members will be able to: ? Access and manage your shared GPU instances. ? Share a fee-free centralized cloud billing account. Invite your teammates today!??????

Which Neocloud should you grow your workloads in? On top of price, geo, and specs, reliable gear, full support, and platform longevity are required for comfortably growing your cloud workloads. Here are some recommendations: Lambda Top Neocloud with a supplemental hardware business. Lambda supports some of the fastest growing startups in the space today along with Sony, Samsung, Pika and more. Their reliable IaaS has always been top of the line with the widest footprint on the market. Denvr Dataworks Denvr based out of Calgary has a fast growing fleet of A100s, H100s and more. Their proprietary stack includes advanced networking features like VPCs, programmatically available interconnected clusters, and features you’ll be hard pressed to find across the market. Their proprietary immersion modular DCs give them a unique edge by expanding DC space, reducing power consumption, and more. Crusoe Crusoe is one of the largest emerging platforms who is now building out some of the largest data centers in the world for OpenAI. They have one of the most elegant cloud console UX with high uptime and reliability for machines supporting some of the most critical workloads out there. DigitalOcean DO is tried and true. Post the acquisition of Paperspace by Digital Ocean, Digital Ocean has started launching GPU inventory under the DO brand. With DO, you of course get GPUs, but you also get all the added features that come with DO making it lead the pack with managed services in this emergent space. Nebius One of the leading AI cloud providers in Europe. You might have noticed them advertising across SF recently and Nasdaq listing. They have a solid interface and advanced capabilities like programmatic clusters, k8s, and more. Hyperstack One of the most price performant and reliable options on the market. Hyperstack has solid gear and a massive inventory for users ranging from A6000s to H100s and more coming soon. Massed Compute Based out of the midwest, Massed has emerged as one of the top Neoclouds in the past 6 months. Their direct offering focuses on virtual desktops, and they have rock solid inventory, prices, unmatched support, and some of the most reliable gear out there. Reach out to them for great commit prices! DataCrunch Datacrunch with servers in Northern Europe and Iceland is green and mean. Their dynamic pricing, which changes daily, means you can sometimes catch UNREAL deals. Their team and support can ensure a high uptime for your servers with a solid cloud console with H200s and B-Series GPUs coming online soon. Sesterce Our partners out of France have deployed 20+ clusters for clients to date, and their new on-demand platform is providing some of the largest variety of compute out there. Hot Aisle Inc. Hot Aisle provides the Ferrari of compute. Their MI300X servers are top of the line Dell boxes in one of the most reliable facilities and setups in the country. Their team delivers an exceptional metal rental service with more to come!

Welcome to the team Dylan Condensa!