Hitachi Vantara, the data storage, infrastructure, and hybrid cloud management subsidiary of Hitachi, Ltd, has officially announced the launch of Hitachi iQ with the NVIDIA HGX™ platform, something which translates to an end-to-end solution stack of fast, scalable, and resilient infrastructure for modern AI systems.
According to certain reports, the stated platform brings forth cost-optimized solutions, tailored to a variety of industries and use cases, including large language models, inferencing, model training, analytics, digital twins, and image processing.
To understand the significance of such a development, we must take into account one survey conducted by ESG Research. The stated survey reveals that, despite 97% of organizations viewing GenAI as a top-five priority, no more than one-third (37%) believe their infrastructure and data ecosystem is well-prepared for implementing GenAI solutions. O the other hand, 77% agreed that data quality issues needed to be addressed before they could adopt GenAI processes.
Talk about the benefits of this collaboration on a slightly deeper level, we begin from how Hitachi iQ- NVIDIA HGX duo combines storage, networking, servers with NVIDIA H100 and H200 Tensor Core GPUs options, as well as the NVIDIA AI Enterprise end-to-end, cloud-native software platform to facilitate the development and deployment of generative AI applications.
Next up, there is the promise to provide enhanced data processing of distributed file systems which, in turn, will allow for more effective pattern identification, while access to diverse datasets simultaneously helps AI models learn from a wider range of scenarios.
“Our customers are looking for advanced solutions to meet the ever-evolving demands of AI, and the Hitachi iQ offering with NVIDIA HGX is the latest example of how we are at the forefront of AI innovation,” said Jason Hardy, chief technology officer for artificial intelligence at Hitachi Vantara. “As one of the only vendors to offer a complete, end-to-end AI infrastructure solution, we marry world-class technology with deep industry expertise for a powerful, one-stop-shop experience.”
Furthermore, thanks to the support from NVIDIA HGX, Hitachi iQ improves the economics and sustainability of AI through a zero-copy architecture which, on its part, eliminates wasteful data copying and transfer times between storage silos for different phases of AI.
The updated Hitachi Content Software for File platform, with generation 5 PCIe, features the latest AMD EPYC processor technology and high-performance networking with NVIDIA ConnectX®-7 400Gb/s InfiniBand or Ethernet NICs. Not just that, the presence of PCIe Generation 5 E3.S NVMe drive technology installs extreme performance capabilities for Hitachi iQ and other data solutions where performance is mission-critical. Apart from that, it provides significant performance upgrades for high-performance computing (HPC) and artificial intelligence (AI) workloads.
Rounding up highlights would be high-density object storage integrated with Hitachi Content Software for File which supports the scaling of compute and storage independently with the flexibility to adapt to diverse and fluctuating data sizes, data types, and workloads.
“Pairing NVIDIA’s accelerated computing platform with the deep industrial expertise of Hitachi Vantara helps address the growing demand for enterprise-grade AI infrastructure and applications, delivering high-performance computing capabilities for AI workloads to enable faster training, fine-tuning, and inferencing,” said Shar Narasimhan, director of data center GPUs and AI at NVIDIA.
Available since July 2024, Hitachi iQ portfolio of AI-ready infrastructure, solutions, and services was markedly designed to help customers identify the most valuable AI use cases, assess their data readiness, determine ROI, and create a strategic roadmap for successful AI implementation. Despite such a short time, though, it has managed to reach industries like rail, energy, and manufacturing. The portfolio has also emerged as a leader for integrating IT infrastructure and business applications.