Portable Ai-MicroCloud

Changing the AI Paradigm 



 Portable Multi-Cloud.   HPC Orchestration.    Algorithms.    APIs.  Quantum-Safe.  


Trusted By
NVIDIA logo.webp
IBM logo.png
Laufer Center for news1.jpg
TW-full-color 600 dpi.jpg
Alef edge logo.JPG
AI institute logo.JPG
Milestone logo.png
QNU for news.jpg




Key Features


Accelerated AI-Pipelines

High performance Data Lake for handling large datasets – 10-15x query acceleration


Open Source Support

Supports all open source frameworks


Micro Cloud  GPU Supercomputing

Ready to use 20+ Tensor Petaflops in a GPU MicroCloud


Multicloud from Core to Edge

Runs anywhere: AWS, Azure, On-Premises or at the Edge


Cloud Native 

Container-based – runs and scales anywhere




Curated algorithms – closed loop validation – IP exclusives

Composable Foundational Components


* Coming soon

Our Solutions

  • We deploy our Ai-MicroCloud™ to your data center or cloud

  • Single enterprise-wide platform

  • Unify entire AI life-cycle 

  • Conforms to your security & compliance parameters 

  • Open source framework support

  • Scale up to HPC infrastructure in one click

  • Access curated AI algorithms

Cloud Service Providers
  • Onsell Ai-MicroCloud™ to​

  • Compete with public clouds

  • Offer low-latency AI solutions

  • Promote higher margin services

  • Increase customer retention

  • API integration for metering & IaaS orchestration apps

Edge Operators
  • Integrate Ai-MicroCloud™ into the Edge-Telco-Cloud ecosystem​

  • Reap benefits of 5G & low-latency 

  • AI inferencing at the Edge for quick decisions

  • Deliver real-time, production-ready AI/ML solutions 

Case Studies

  • Accelerate drug discovery for the SARS-CoV-2 virus, using Computational Physics (MELD) modeling. 

  • HPC notebook with MPI using – 128 NVIDIA RTX 6000 GPUs, 500 GB of Parallel File System

  • Multiple Containers to support multi-GPU, multi-CPU compute engines 

Laufer Center for news1.jpg
In silico.JPG
  • In Silico drug discovery simulations, computer-aided drug design (CADD), molecular modeling and quantitative structure-activity relationship (QSAR) for pharmaceutical companies

  • HPC notebook usage – 8 NVIDIA RTX 6000 GPUs

  • Zeblok’s platform provides 6x improvement in performance compared to NSF’s Blue Waters Cray GPU-based system

Laufer Center for news1.jpg
  • Dance Classification

  • Image classification by computer vision using convolution blocks and VGG family of convnets to identify dance genre

  • Open source Jupyter notebook

  • Multiple containers to support multi-CPU, multi-CPU compute engines using – 1 NVIDIA RTX 6000 GPU

CNN lung.JPG
  • Covid-19 CT Scan Segmentation

  • Unlabeled largest data set of infected lung CT scans called MOSMEDDATA

  • 800 CT volumes of COVID infected patients which results in around 25k CT slices.

  • 2 RTX 6000 GPUs, 500 GB of Parallel File system

AI institute logo.JPG
  • COVID-19 Epidemiology 

  • Assist health care administration, insurance, and government agencies in emergency preparedness.

  • Open source Kaggle repository dataset - UNCOVER COVID-19 Challenge provided by Roche Data Science Coalition 

  • Zeblok AI-Rover™ WorkStation, 1 RTX 6000 GPU, 50 GB of Block Storage, 100 GB Object Storage

Senti ana.JPG
  • Predictive Analytics

  • TRMI data from Refinitiv, combined with pricing data in a unique project that aimed to understand the impact of market sentiment on returns.

  • Zeblok AI-Rover™ WorkStation, 7 vCPUs, 16GB RAM, 50 GB of Block Storage

Study des.JPG
  • Rare Disease Study Design

  • C-Path institute dataset from Friedreich's Ataxia Clinical Outcome Measure Study (FA-COMS) natural history data to identify novel endpoints, biomarkers, and baseline characteristics that may optimize clinical study

  • Zeblok AI-Rover™ WorkStation, 1 RTX 6000 GPU, 50 GB of Block Storage

SCD image.JPG
  • Skin Cancer Classification

  • ISIC 2020 Challenge Dataset containing 33,126 dermoscopic training images of unique benign and malignant skin lesions from over 2000 patients. 

  • Zeblok AI-Rover™ WorkStation, 1 RTX 6000 GPU, 1 vCPU, 16GB RAM, 50 GB of Block Storage, 100GB Object Storage

Just 3 Steps 

Select a Notebook
Pick Infrastructure
Start modeling
  • Choice of Open Source framework Notebooks that are CUDA optimized

  • Choice of curated Algorithm Notebooks

AI Showroom

Zeblok Ai-MicroCloud™ 21-day trial

A complete user experience on the Zeblok MicroCloud for Data Engineers, Data Analysts and Data Scientists 

Step #1 
Select a Notebook

Choice of

  • CUDA optimized open source framework notebook


  • Notebook, with algorithm exclusively available on the Zeblok platform

Step #2
Select Infrastructure

Choice of

  • CPU environment on our Micro Cloud


  • Single or hundreds of GPUs in our MicroCloud (up to 20+ petaFLOPS)

Step #3
Start Modeling
  • Ingest data into high performance Data Lake

  • Train models and integrate runtime AI solutions into your enterprise business processes from Core to Edge

Partner Programs


Cloud Service Providers (CSPs)

Managed Service Providers (MSPs)

Specialized Hardware Manufacturers

  • Empower your customers to unlock growth in their business with innovative and customized AI/ML predictive services

  • Accompany your customers every step of the way on their AI journey

  • Create unique solutions integrating your value-added services


Data Companies

  • Activate your datasets for integration with AI/ML models

  • Easily monetize datasets as part of Zeblok’s Global Partner Ecosystem

  • Expand revenues and cross sell/up-sell with unmatched curated algorithms in Zeblok’s Intelligence Marketplace    


Algorithm Developers Within Academia

AI Startups

Other Algorithm Developers

  • Distribution through our Intelligence Marketplace

  • Security that safeguards your intellectual property

  • Generate additional revenues customizing your algorithm to meet specific enterprise requirements


System Integrators (SI)

AI Solution Providers

  • Get access to Ai-MicroCloud™ including development licenses for test and development purposes

  • Leverage Ai-MicroCloud™ for more time and cost efficient client deliverables

  • Access technical library, algorithms, API documentation and sample data

Announcements and News

Tangent Works and Zeblok Partnership ann


“Very simple to install my protein simulation software. Using only 1 GPU on Zeblok's Ai-MicroCloud™ , we saw a 6x improvement in performance compared to Blue Waters Cray GPU-based system.”

- Laufer Center   

compare graph.tif