GPU Accelerated Database

Speed of thought analytics on data at scale and in the cloud

Extreme data crunching on GPU (Graphics Processor Units) provides up to 1000x performance improvement while reducing hardware cost and power consumption.


Today's organisations all use data processing to inform decision making. However, the amount of data processing an organisation can undertake is limited; to process more data, use more sophisticated algorithms or get answers faster requires greater data processing power – exponentially greater. Indeed, according to Forrester Research the single biggest challenge, shared across organisations’ database management, is delivering improved performance with 80% stating this is creating serious challenges. This is a rapidly-growing problem. 90% of the world’s data were created in the last two years alone. Cisco forecast 50 billion connected devices by 2020, each of which will be generating a data stream. The Solution brytlyt’s solution is to massively improve data processing power without a correlated massive investment in technology. brytlyt’s core product MakuluXtrm fundamentally changes the speed of data processing and the cost of data processing. Organisations acquiring MakuluXtrm acquire a transformational capability to process data into more readily available, more accurate, richer and more targeted information.  An organisation’s data can be exploited to much greater impact on both internal and external information consumers - both for existing use cases and to identify new strategic options for value creation, which would not otherwise be observable. MakuluXtrm combines proprietary software and hardware that delivers extreme performance up to 1,000x faster than existing technology. The cornerstone of the proprietary hardware is the use of Graphics Processing Units (GPUs) to accelerate data crunching. General Processing on GPU is the only technology that can provide anything close to the performance that real-time analytics at scale demands. The cornerstones of the proprietary software are patent-pending data processing algorithms that allow the most efficient use of GPUs. 32% of respondents in a recent Forrester Research survey indicated they planned expenditure on In-Memory database products - the closest existing alternative to MakuluXtrm. brytlyt's second product, Makulu, is a version of MakuluXtrm that runs on standard hardware. Despite the obvious commercial drivers and market size, brytlyt does not face any meaningful competition in the GPU space. Other players do not appear to have overcome fundamental hardware constraints that brytlyt has addressed (specifically the PCIe and network bottlenecks) and do not offer a feature-rich and flexible platform.  Existing large and established players in the wider database space don’t appear to be able to even begin to compete with brytlyt on any given metrics.  Quite aside from raw performance, MakuluXtrm hardware is up to 50% cheaper, 75% more energy efficient with a physical form factor that is far smaller.

Key Benefits

Massively improve data processing power without a correlated massive investment in technology. Delivers extreme performance up to 1,000x faster than existing technology Up to 50% cheaper 75% more energy efficient

Register to view this innovation

  • Access all of the innovation information
  • Contact the innovator
  • Follow this innovation's progress
  • View innovation updates