IBM Cloud Pak for Data System updates provide a modern hyperconverged solution for transformation

By , Klaus Roder, and Alice Wang | 3 minute read | November 20, 2020

In this age of hybrid IT, companies that have barriers to cloud adoption need to modernize their infrastructure to be successful. This need is addressed by hyperconverged infrastructure (HCI), which injects modernization into any IT infrastructure by providing easy data and AI deployment and management functions.

IBM Cloud Pak® for Data System combines the power of hyperconverged infrastructure with cloud-native architecture and is designed to help customers align their data center strategy with their cloud strategy.

The latest version of IBM Cloud Pak for Data System (V1.0.7.5) includes updates from  IBM Cloud Pak for Data version 3.5, which is packed with capabilities to drive essential transformation. Combining a comprehensive set of IBM open source and third-party services within an integrated platform, the enhancements in version 3.5 build upon the existing benefits of the platform in two critical areas: (1) cost optimization and (2) innovation to drive digital transformation.

In addition to the new software benefits,  IBM Cloud Pak for Data System comes with a best-in-class support experience and is optimized for best performance.

Best-in-class customer support

What does best-in-class support entail? Consider the following:

  • IBM will install your complete solution.
    IBM does it all – setting up the hardware and installing the IBM Cloud Pak for Data software for you, even under COVID-19 restrictions. After the install team leaves you will have an up-and-running, fully functional system.
  • You can start out small and scale to grow on demand.
    IBM Cloud Pak for Data v3.5 is available as a starter bundle. If and when you need it, it is easy to grow by adding expansion units to your system.
  • Avoid additional hardware purchase and verification processes.
    Deploying on premises without IBM Cloud Pak for Data System means another vendor, another procurement and above all, delays, delays, delays. Avoid all of that and accelerate your time to value by letting IBM handle the entire IT management process, including:

    • Proactive maintenance of hardware to help avoid possible failures and critical services ensure on-site help where applicable.
    • Available IBM Technical Account Managers to help with offer guidance and expertise for successful deployments and planning/consultations.
    • Expert services to help your teams quickly get started on use cases. Clients have access to experience, knowledge, resources and skills from a worldwide team of IBM experts.

Can you do this on your own? It’s possible but you would need a plan for everything below:

Figure 1: “Do it yourself”

Performance matters

Deploying on a single hyperconverged system gives an advantage by combining all of these factors on hardware built for the all-important data and AI platform you are investing in. This hardware, beyond being the only hardware built specifically for IBM Cloud Pak for Data, carries several important performance characteristics.

First, in terms of storage, the groundbreaking NVMe (non-volatile memory express) drive in IBM Cloud Pak for System achieves 5x faster read and 2x faster write than traditional industry-standard SATA drives. This is because exponential rise in data can slow down even the highest performing SSDs (solid state drives) if connected to legacy storage protocols.

Also, the Active-Active 25Gb networking storage in IBM Cloud Pak for Data System not only prevents network downtime but also provides load balancing, resulting in throughput and response time improvement of 50x faster than public cloud.

Beyond storage, IBM Cloud Pak for Data System offers multiple improved resource efficiencies compared to a full rack Netezza® Mako, including:

  • 3x SQL performance with 1/5 the footprint
  • Increased data science efficiency, achieving the same training speed using 1/3 of the cores

In addition, IBM Cloud Pak for Data System offers additional performance optimization with field programmable gate arrays (FPGA). In performance testing, running IBM Cloud Pak for Data System achieved an average of 40% more jobs per hour than running on SoftLayer® infrastructure.

Overall, the IBM system offers seamless scalability for both hardware and software. Compared to industry-standard public cloud offerings—which often have storage, memory, and virtual compute limits—IBM Cloud Pak for Data System hardware supports in-place expansion on demand so customers can scale their environments without having to redeploy or reinvest in a different machine. IBM HCI also offers sophisticated automation where compute and storage resources can be expanded in place as needed from a common resource pool, ensuring maximum resource allocation.

Considering a superior customer experience and the highest-performing hardware, IBM Cloud Pak for Data System is the clear choice for customers looking to deploy IBM Cloud Pak for Data on premises. To learn more about how this hyperconverged system stands up against the competitors, watch the webinar with Cabot and IBM experts on demand here. Or learn more by checking out the product details.