choose your region

Architecture

  • Architecture

  • HBMP technology
    makes BOARD superfast

  • The new Hybrid-in-Memory technology, called HBMP, revolutionizes BOARD performance and overcomes the limitations of traditional in-memory solutions.

  • Download the
    HBMP Whitepaper

  • HBMP: the new frontier of in-memory BI and CPM

    BOARD's approach to in-memory technology started with a precise goal: to exploit the full potential of IMC (in-memory computing) in terms of better performance without sacrificing the flexibility, scalability, total cost of ownership and ability to effectively support the simulation and planning processes, that have fueled international success for BOARD. From this research, a unique technology arose which BOARD has called HBMP that has no equal either in the Business Intelligence or in the Corporate Performance Management space.

  • What is HBMP (Hybrid Bitwise Memory Pattern)?

    Bitwise Memory Pattern

    Thanks to the adoption of a new algorithm, physical data is stored and mapped through unique bits, or rather “bitmapped”. In other words, BOARD maps the compressed multidimensional data structure (Pattern) at bit level (Bitwise) into the RAM (Memory).

    The new algorithm, designed to exploit all the potential of in-memory computing, brings enormous benefits, not only in terms of performance, but also in terms of data compression, parallelism and high-user concurrency.

    Furthermore, unlike what happens with the majority of in-memory solutions, the algorithm is designed to ensure full bi-directionality (read-write) in the use of information. Consequently, this allows full support of write-back and “on the fly” changes of the data structure, thus keeping all the functional capabilities of BOARD intact.


    Hybrid

    Another fundamental characteristic of HBMP technology is its Hybrid nature, or rather the ability to manage data in three different ways:

    • Full in-memory
      to do all processing against data held in-memory
    • On-demand in-memory
      to decide which subset of data to use in-memory as more appropriate to the workload
    • Hybrid in-memory
      to store in-memory data indices and mapping and to leave data on the disk

  • These three different ways of configuring the BOARD in-memory environment allows great flexibility in choosing the model that best suits your needs, both in terms of scalability and price-performance.

    Not only is any “out of memory” problem eliminated, but at the same time, BOARD offers all the options needed to optimize the technological cost of ownership of any BI and CPM implementation.

  • HBMP Benefits

      HBMP Stand-Alone In-Memory
    Write-back Supported
    Calculated value    		 Not supported (or supported just on a single non-calculated value)
    Data Spreading Supported Not supported (or poorly supported)
    Data model Flexible

    (Changes “on the fly”)    		 Rigid
    Data documentation Notes

    File attachment

    Images    		 Not supported
    Extensibility Drill-through

    Cross application data model    		 Drill-through not supported

    Sealed data model
    Scalability Virtually unlimited

    High incidence of concurrent users on RAM resources    		 Limited to existing RAM

    Low incidence of concurrent users on RAM resources
    Workflow Integrated Not supported
    Office Integration Complete and Advanced Limited
    Development
    environment    		 Programming-free, Visual

    Directly on-line

    Logical or physical view of data    		 Code based

    Need to re-upload the data to go live

    Physical view of data
WebsiteFeedback