NVIDIA DGX POD
Fully Integrated Applied Data Systems NVIDIA DGX POD
While verified in the laboratory, as with all complex HPC systems, NVIDIA DGX Reference Architectures are complex and difficult to get up and running in order to be productive. Applied Data Systems (ADS) solves all this. Our custom built NVIDIA DGX POD includes any number of NVIDIA DGX H100 Servers along with high-speed 400Gb NVIDIA Mellanox Infiniband networking, high-speed NVMe storage using IBM Storage Scale systems. Red Hat or NVIDIA Ubuntu OS, Bright Cluster Manager software, SLURM and all NVIDIA NGC software packages as part of the stack – all designed to work together, optimally and seamlessly. The ADS NVIDIA POD is factory integrated and tested as one before shipment ensuring you get an optimized DGX POD, saving you weeks of on-site installation hassle.
Delivered Ready for Immediate Use
The ADS NVIDIA DGX POD is setup on-site by our white glove AgilityCare professional service technicians. So when a DGX POD from Applied Data Systems arrives at your door, it can be rolled into your data center to transform your data into impactful insight immediately. TS-SCI cleared personnel for government or sensitive installations are available.
Experts in Parallel File Systems
ADS offers a variety of storage solutions based upon customer requirements. Available with IBM ESS Storage Scale systems, NetApp FAS or E series storage running IBM Spectrum Scale Data Management or Data Access edition, ExtremeStor ECE using IBM Scale Storage erasure code edition.
Our ExtremeStor ECE solution (using IBM Storage Scale erasure code edition) offers both a hybrid and all NVMe flash storage system for all AI and Machine Learning workloads. Our solution is designed to ensure that no NVIDIA DGX H100 GPU server is waiting on data. ExtremeStor with IBM Storage Scale ECE file system software delivers extreme scalability, flash-accelerated performance, and automatic policy-based storage tiering from flash through disk to cloud or even tape—helping reduce storage costs up to 90 percent while providing better security and management efficiency in cloud, big data, and analytics environments. ExtremeStor ECE delivers high IOPS and high bandwidth to support data-hungry GPUs with tens of GB/s of bandwidth per DGX H100, simultaneously across as many DGX servers as needed.
Cut Time to Production from Months to Hours
Applied Data Systems unique pre-integration and delivery process dramatically reduces the time the customer needs to spend getting their system into production – from months to hours. Applied Data Systems customize offerings to fit our customer’s requirements in our state of the art Customer Integration and Manufacturing Center (CIMC). We work hard to eliminate integration issues by doing a full burn in at our CIMC. This work includes full integration of all components such as CPU’s, accelerators, network adapters, NVMe drives, etc, into our rack mount and blade servers, with complete end to end storage customization and setup including loading and optimization of parallel file-systems and all middleware components. We strive for 100% customer satisfaction and offer a ‘white glove’ installation service ensuring 100% success.
Applied Data Systems can assemble and stage the fully-functioning DGX POD system and enable the customer to access it remotely to test and validate the solution before shipment. A customer can perform remote benchmark tests on their new AI infrastructure to verify quality and interoperability prior to the NVIDIA DGX POD being delivered.
The Result
This state-of-the-art ADS DGX POD uniquely enables customers to rapidly deploy new infrastructure for AI workloads immediately upon delivery. Let us discuss your project so we can help determine what is the best setup for your requirements. If there is a requirement for discussions to be done in a classified setting, we can help with that too. No project or discussion is too large or too small. We’re here to help out. Get in touch with us now.