Pinakastra AI Cloud
Pinakastra is dedicated to making cloud computing infrastructure more accessible to enterprises, academic institutions, and research organizations. By leveraging the Pinakastra™ cloud platform, we assist these entities in IT infrastructure consolidation, ensuring they achieve optimal efficiency, scalability, and cost-effectiveness.
| Technical Specifications | Cirrus (Model -PIN-HCI-16-256-12) | StratoCumulus (Model -PIN-HCI-32-768-48) | GPU (Model -PIN-HCI-48-1024-48) |
|---|---|---|---|
| Chasis |
|
|
|
| GPU | "Upto to 2 AMD RX580 PCle GPU cards" | GeForce RTX 3080Ti GPU (12GB GDDR6X) | "Up to 10 x NVIDIA V100 PCle GPGPU cards" |
| Processor | Single AMD Ryzen 3700x CPU | "2nd Gen Xeon Scalable Processor with (12) swDIMM slots & 12"" x 13"" EATX Form Factor" | "2nd Gen. Intel) Xeon®: Scalable and Intel® Xeon® Scalable Processors" |
| RAM | 4-Channel DDR4 DIMMS | 8 x 32GB Regd. ECC DDR4 3200MHz = 256GB RAM | 6-Channel RDIMM/LRDIMM DDR4, 24 x DIMMs |
| Ethernet Card-1 | 1 x GbE LAN Ports | 1 x GbE LAN Ports | 2 x 10Gb/s BASE-T LAN ports (Intel® X550-AT2) |
| Ethernet Card-2 | 2 x 10GbE LAN ports | 2 x 10GbE LAN ports | 2 x GbE LAN ports (Intel® 1350-AM2) |
| OS Disk | 2xM.2 NVMe, 2 x SATA 3.5 | 1 x 500GB NVMe M.2 Internal SSD / Client(for OS) | "5 x NVMe, 2 x SATA/SAS 2.5"" hot-swappable HDD/SSD bays" |
| Data Disk | 2 x SATA 3.5 (Hot Swappable) | 1 x 8TB SATA HDD /tf Enterprise (for DATA) | 12 x 3.5" SATA/SAS hot-swappable HDD/SSD bays |
| PCle Slots | 2 x low profile PCle Gen3 expansion slots | 2 x low profile PCle Gen3 expansion slots | 2 x low profile PCle Gen3 expansion slots |
| PSU | 1 x 80 Plus Platinum 1500W PSU | 1 x 1200W PSU | 5 x 80 PLUS Platinum 2200W redundant PSU |
| GPU + Cloud | "multiple workloads GPU PCI Passthrough virtual machines for" | GPU PCI Passtrough virtual machines for multiple workloads | "GPU PCI Passtrough virtual machines for multiple workloads" |
| What you get? | "Inclusions: GPUs included with this price Pinakastra Cloud Platform Ceph Unified Storage (Block, Object & NFS) Exclusions: Transportation & Deployment Charges" |
"Inclusions: GPUs included with this price Pinakastra Cloud Platform Ceph Unified Storage (Block, Object & NFS) Exclusions: Transportation & Deployment Charges" |
"Inclusions: Pinakastra Cloud Platform Ceph Unified Storage (Block, Object & NFS) Exclusions: Transportation & Deployment Charges" |
| Average VM Size | 2 vCPU, 4GB RAM & 20 GB HDD (GPU PCI Passthrough) | ||
| Total Virtual Server | 48 | 96 | 256 |
Pinakastra Cloud is set to leverage NVIDIA technology for accelerating both compute-intensive and graphics workloads, achieved either through GPU PCIe Passthrough or via Nvidia vGPU drivers.
Technology Stack
Pinakastra’s AI Cloud stack combines open-source technologies with hyperconverged infrastructure (HCI) and NVIDIA GPU acceleration, delivering tailored private cloud solutions for compute and storage-intensive workloads
.
Customer Perception
Public cloud costs for AI/ML workloads are rising, making Pinakastra’s scalable private cloud solutions attractive to enterprises, academia, and research organizations anticipating growing AI demands.
AI & Deep Learning
Pinakastra’s OpenStack-based platform provides a scalable, automated environment for AI/ML, enabling flexible deployment and fostering innovation through open-source technology.
