ASBIS Delivers End-to-End AI Infrastructure with Pure Storage and NVIDIA DGX SuperPOD
Pure Storage has taken an important step for enterprise AI: its FlashBlade//S-based data platform has been certified by NVIDIA for DGX SuperPOD deployments built on the new DGX GB200 and DGX GB300 systems. This combination of Blackwell-generation DGX compute and scale-out all-flash storage is designed for large-scale, production AI, where data infrastructure has to scale linearly, stream data at extreme throughput, and keep thousands of GPUs continuously fed.
For organisations building AI factories, this certification means the storage layer is no longer a bottleneck or an experiment. FlashBlade//S is tested as part of NVIDIA DGX SuperPOD reference architectures to deliver the throughput, latency, and resiliency required for training and serving the largest models, while staying simple to operate and expand.
ASBIS is ready to turn this milestone into real projects across its markets.
ASBIS is an NVIDIA Elite Partner and NVIDIA Solution Provider, working closely with NVIDIA on advanced data centre, AI, and accelerated computing solutions. In addition to NVIDIA GPUs and DGX systems, ASBIS also delivers NVIDIA networking technologies for high-performance data centre fabrics, including InfiniBand and Ethernet-based solutions designed for AI and HPC environments.
At the same time, ASBIS is an authorised Pure Storage distributor and long-term value-added partner, delivering FlashArray and FlashBlade solutions to customers through a strong regional network of local offices and partners. This combination – deep NVIDIA expertise and a mature Pure Storage practice – allows ASBIS to design and supply fully validated AI infrastructures based on NVIDIA DGX SuperPOD with Pure Storage as the certified storage tier.
Together with NVIDIA and Pure Storage, ASBIS can provide a complete stack for AI-ready data centres, including:
- NVIDIA DGX systems based on the Blackwell architecture (DGX GB200 / GB300 and related platforms) for large-scale training and inference
- Pure Storage all-flash platforms – FlashBlade//S as the certified storage for DGX SuperPOD, plus complementary FlashArray systems for surrounding workloads
- NVIDIA high-speed networking to build the AI fabric connecting compute and storage
- Architecture design, sizing, and validation based on NVIDIA and Pure reference designs, adapted to local facilities, regulations, and workloads
- Implementation, integration, proof-of-concept support, and ongoing services through ASBIS expert teams
For more details, please do not hesitate to contact our experts.