Accelerate Your Workflows With
Bare Metal Instances
Unlock the full potential of dedicated GPUs with Shakti Cloud's bare metal instances. Gain complete control, low-level user access, and lightning-fast storage and networking, delivering unparalleled scalability.
Harness the power of multiple GPUs with incredibly fast connections and a software stack that is fully optimised for acceleration to tackle the processing of massive datasets, accommodate the growth of model sizes, and handle the complexity of simulations with ease.
Solutions Spotlight
BM Large
Shakti Cloud BM Large offers NVIDIA HGX with eight H100 cards. Designed specifically for Artificial Intelligence (AI) and HPC workloads. Start with single bare metal and scale more than 1000 nodes seamlessly.
BM Large Specification
- 224vCPU
- 2048 GB RAM
- 640 GB GPU Memory with NVIDIA NVLink®
- 50TB Usable Internal Storage
- NVIDIA BlueField-3, NVIDIA Quantum-2 InfiniBand Network
- Unlimited ingress and egress with 500Mbps per node Internet Bandwidth
BM Universal
Shakti Cloud BM Universal offers four NVIDIA L40S servers packed to perform to boost your LLM inference and simulation workloads.
BM Universal Specification
- 128 vCPU
- 1024 GB RAM
- 640 GB GPU Memory with NVIDIA NVLink®
- 960 GB OS Drive, 7TB Usable Internal Storage
- NVIDIA BlueField-3, NVIDIA InfiniBand Network
- Unlimited ingress and egress with 100Mbps per node Internet Bandwidth
BM Max
Boost your most demanding high-performance computing and AI tasks with powerful NVIDIA A100 cards. Shakti Cloud BM Max offers 4 NVIDIA A100 GPUs.
BM Max Specification
- 160 vCPU
- 768GB RAM
- 1TB Disk
- Upto 10/25G network
- Unlimited ingress and egress with 200Mbps per node Internet Bandwidth.