Among the many announcements that were made at Hitachi Vantara’s NEXT 2018 event last month in San Diego, was the announcement of an enhanced converged and hyperconverged portfolio to help customers modernize their data centers as part of their digital transformation strategies. Hitachi Vantara is committed to providing customers of our IT infrastructure products with the latest technology innovations to address their evolving business and data requirements. There was a lot that was covered in this announcement, so t would like to go into more detail on four of the key innovations that I feel are more significant. These Innovations include: DS7000 Scalable Servers, NVIDIA Tesla GPUs, All NVMe, and 3D XPoint storage memory.
DS7000 Scalable Servers
Hitachi Advanced Server DS7000 Series of Scalable Servers are built with a unique modular architecture which can be configured and scaled to meet the needs of a wide variety of application workloads. Each module supports a pair of Intel Xeon Scalable processors, with up to 56 cores and 3TB of memory per module. An 8 socket DS7000 server (DS7080) can support up to 224 cores and 12TB of memory. Each model can be smoothly upgraded to the next, preserving your investment in hardware and software as you grow, and compute modules can be individually configured to support a wide variety of compute and storage options. The DS7000 server family is available in Converged and Rack Scale configurations and is the ideal platform to deliver the high availability and scalability needed for Hitachi’s solutions for business-critical applications. Hitachi Solution for the SAP HANA Platform makes full use of the huge memory capacity of the DS7000 to deliver real-world business benefits from data analytics. Hitachi Solution for Databases relies on the powerful performance and massive I/Ocapacity of DS7000 to meet demanding service level agreements (SLAs) and rapidly changing workloads, allowing businesses to thrive and grow
Hitachi Advanced Server DS225 with NVIDIA Tesla GPUs
GPUs which were first developed for the gaming industry can now play an important role in the modern data center. Hitachi Advanced Server DS225 with NVIDIA Tesla GPUs is architected to accelerate advanced analytics, Machine Learning, CAD/CAE, and specialized graphics.
GPUs and CPUs process tasks in different ways: A typical CPU is optimized for sequential serial processing and is designed to maximize the performance of a single task within a job. GPUs, on the other hand, use a massively parallel architecture aimed at handling multiple functions at the same time. As a result, GPUs are 50 to 100 times faster than CPUs in tasks that require multiple parallel processes such as machine learning and big data analysis. The following Youtube Myth buster video is a fun way of showing the difference between CPU and GPU processing.
According to NVIDIA, data scientists and researchers can use their NVIDIA® Tesla® GPUs to parse petabytes of data, orders of magnitude faster than they could by using traditional CPUs, in applications ranging from energy exploration to deep learning, enabling them to tackle challenges that were once thought impossible. Tesla accelerators also deliver the horsepower needed to run bigger simulations faster than ever before and deliver the highest performance and user density for virtual desktops and workstations.
The DS225 with NVIDIA can be combined with the DS7000 server in the Hitachi Unified Compute Platform CI (UCP CI) converged systems and Hitachi Unified Compute Platform RS (UCP RS) rack-scale systems to address different environments, use cases and applications.
All NVMe and 3D XPOINT
Hitachi Unified Compute Platform HC (UCP HC) hyperconverged infrastructure appliance series will include a new all-flash NVMe UCP HC V124N system providing high storage density and capacity in a low-latency 1U form factor housing up to 12 NVMe SSDs (up to four ultra-fast Intel 3D XPOINT-based Intel Optane SSDs for cache and eight NVMe SSDs for capacity) delivering a total raw capacity of up to 80TB. It is designed to address the needs of customers requiring higher IOPS and ultra-low latency for applications such as relational databases, real-time analytics, stock trading, fraud detection and transaction processing (OLTP). The all-NVMe hyperconverged system will offer customers dramatically improved performance and efficiencies with a higher ROI by delivering hundreds of thousands of IOPS at sub-millisecond response time, for a 3X increase in IOPS performance and 4X lower latency compared to the prior all-flash UCP HC system. Hitachi and Intel engineering teams collaborated to run joint tests over many months to determine and optimize the performance results for real-world scenarios. For more information about the all NVMe hyperconverged solution and the test results, you can read my colleague Dinesh Singh’s blog here.
NVMe (non-volatile memory express) is a host controller interface and storage protocol created to accelerate the transfer of data between computers and solid-state drives (SSDs) over a computer's high-speed Peripheral Component Interconnect Express (PCIe) bus. (You can read my previous blog post on NVMe and Enterprise storage arrays via this link)
While NVMe NAND SSDs have been available on UCP hyperconverged systems for some time, this latest enhancement introduces a new solid-state device which is a major improvement over NAND SSD technology. Intel’s 3D XPOINT technology is a new type of persistent solid-state memory which sits between the performance of NAND SSD and DRAM. 3D XPoint has about one thousandth the latency of NAND flash (or about ten times the latency of DRAM), and ten times the density of DRAM. Since its cost is higher than NAND SSD but lower than DRAM it can be used as a cache between NAND SSD and DRAM to boost performance. Here is another YouTube Video that does a great job of describing the 3D XPoint technology.
We will be showcasing all-NVMe hyperconverged system Unified Compute Platform HC at VMworld Barcelona in a few weeks. Please visit Hitachi booth #P311for a live demo and see how to beat latency.
These new innovative technology options provide a flexible, low-risk path to modernizing data center infrastructure for a broader set of enterprise applications. New mix-and-match server options increase server flexibility and density to address different environments, use cases and applications.