If you want the multi-cloud freedom for your hyper-converged infrastructure without any public-cloud lock-in, then look no further. Start in minutes through just a web browser, with self-registration. Immediately and securely access 24 tensor petaflops (a petaflop is the ability of a computer to do one quadrillion floating-point operations per second). Stay tuned; we are expanding to 200+ tensor PFLOPS later in 2020. Option to choose capacities at different data centers based on geographical, technical and compliance requirements. Yes, you can think of us as the Airbnb of supercomputing.
A simple user interface belies the power of a supercomputer via web browser. Zeblok is a one stop shop. Whether you are looking to start a multi-user data science project using various open-source AI frameworks, or you are looking for a high-performance computing (HPC) environment to distribute your jobs across a large cluster of computational power, we have you covered.
Powerful Machine Learning / Artificial Intelligence notebook
Model Training, with ML/DL pipelines
Out-of-the-box CUDA optimization on GPU for AI- Frameworks
Supports popular data science language bindings such as R, Scala, Docker and Python
We embrace open-source frameworks and tools
Ask your data scientists. They'll tell you they build open pre-trained models and derive inspiration from existing models. Bring your data scientists and AI-engineers onto our platform to launch your next model development in minutes. We've curated ground-breaking algorithms, some of which have patents pending, to enable your team to hit the ground running. Our ecosystem, comprised of university affiliations, start-up companies and top algorithm researchers, will inspire your team to build world-class AI applications. Our large community of data scientists is available to answer your questions on AI-development. We also customize algorithms to fit into your business processes and to correspond exactly to your customers' needs.
Build Intelligence Into Your Processes NOW
While building our own AI models, we were surprised that there was no turnkey solution available offering a straightforward service to access accelerated computing power, or Supercomputing-as-a-Service. AI should not be limited to the realms of academics or tech giants. So, we decided to address the white space in the market.
As we developed alliances with universities, we also discovered an AI-model commercialization chasm. Such grassroots research produces high quality algorithms. Businesses want to take the AI journey, but there is no easy way for enterprises get access to these algorithms. We know you want access to the best algorithms in a way that they can be integrated into your company's workflows. We know you want to reduce wastage by building upon curated algorithms.
Whether you are a start-up, SMB or an enterprise, and working on big-data acceleration, machine learning, deep learning, quantum simulation, computational biology, business/marketing insights, protein simulation, smart city IoT, these are computationally heavy problems. Zeblok's ecosystem provides a capital-efficient AI development environment. You will save both on infrastructure and algorithm development. Like GEICO, our gecko says - 15 minutes will save you 15% or more on your next AI project. And this is just a start. We have lowered the barrier to hyper-converged-AI like no one else.
Our proprietary orchestration layer software supports multi-class scheduling, allowing simple multi-user AI/ML workstations to co-exist with High-Performance Computing (HPC) workloads needed for substantial AI training. It permits on-demand usage of GPU capacities for AI algorithm development. It provides a rapid prototyping environment to promote AI models as API for rapid consumption into enterprise business processes. The highest quality security protocols ensure the integrity of data. Users have access to curated AI/ML algorithms through our AlgoStore to build products that access an algorithm, paying only a user royalty, while securing the intellectual property of their end product.
PhD Candidate, Laufer Center
From 1st Data Center @ CEWIT Stony Brook
“Very simple to install my protein simulation software. Saw 6x improvement in performance compared to Blue Waters Cray GPU based system using 1 GPU”
GET IN TOUCH
1500 Stony Brook Road
Stony Brook, NY 11784
+1 (201) 637-3603