STAC Report: DDN’s AI400X2 and EXAScaler 6.0 under STAC-M3 (tick analytics)

Hardware/software generational performance gains with half the appliances.

10 November 2021

STAC recently performed STAC-M3 Benchmarks on the DDN’s newest all-flash storage appliance, the AI400X2. The solution used the latest release of DDN’s EXAScaler software to access the data on the appliance.   The results are available here.

STAC-M3 is the set of industry standard enterprise tick-analytics benchmarks for database software/ hardware stacks that manage large time series of market data ("tick data"). In this project, we ran both the baseline benchmark suite (code named Antuco) and the optional scaling suite (code named Kanaga).

The stack under test (SUT) was KX’s kdb+ 4.0 DBMS distributed across 15 Intel HNS2600BPB servers using DDN EXAScaler 6.0.0 software to access a single DDN AI400X2 All-Flash appliance through a parallel filesystem. The SUT used Compatibility Revision H of the kdb+ STAC Packs. DDN chose to highlight that this solution:

• Faster than a solution involving 2 of the previous generation of DDN’s storage appliance (AI400X) with kdb+ 3.6 and previous versions of the kdb+ STAC Packs and 15 database servers (SUT ID KDB200915)
  • in 10 of 17 mean response-time Antuco benchmarks, including:
    • 3.1x speedup in 10-user volume curve (STAC-M3.β1.10T.VOLCURV.TIME)
    • 2.9x speedup in 10-user market snapshot (STAC-M3.β1.10T.MKTSNAP.TIME)
    • 2.1x speedup in 10-user theoretical P&L (STAC-M3.β1.10T.THEOPL.TIME)
  • in 19 of 24 mean response-time Kanaga benchmarks, including:
    • 2.1x to 2.3x speedup in the four 10-user market snapshots (STAC-M3.β1.10T.YR{2,3,4,5}-MKTSNAP.TIME)
• Faster than a solution in a public cloud involving a parallel filesystem with 15 database servers, 40 storage servers, and previous versions of the kdb+ STAC Packs (SUT ID KDB210507)
  • in 12 of 17 mean response-time Antuco benchmarks, including:
    • 4.4x speedup in 10-user volume curve (STAC-M3.β1.10T.VOLCURV.TIME)
    • 4.0x speedup in 10-user market snapshot (STAC-M3.β1.10T.MKTSNAP.TIME)
    • 2.6x speedup in 10-user theoretical P&L (STAC-M3.β1.10T
  • in 12 of 24 mean response-time Kanaga benchmarks
    • 16.2x to 19.9x speedup in 10-user market snapshots (STAC-M3.β1.10T.YR{2,3,4,5}-MKTSNAP.TIME)
• Faster than a solution involving network-attached flash storage and 9 database servers (SUT ID KDB210929) in 11 of 17 mean response-time Antuco benchmarks, including:
  • 11.4x speedup in single-user intervalized statistics (STAC-M3.β1.1T.STATS-UI.TIME)
  • 6.6x speedup in 10-user market snapshot (STAC-M3.β1.10T.MKTSNAP.TIME)
  • 4.2x speedup in 10-user aggregate stats (STAC-M3.β1.10T.STATS-AGG.TIME)
  • 3.5x to 5.9x speedup in multi-user intervalized statistics (STAC-M3.β1.{10,50,100}T.STATS-UI.TIME)

For details, please see the report at the link above. Premium subscribers have access to the code used in this project as well as the micro-detailed configuration information for the solution. To learn about subscription options, please contact us.