CHATSWORTH, Calif., Dec. 7, 2021 – HPC-AI data storage and management company DDN today announced what it called “dramatically improved” results in the STAC-M3 benchmark for measuring performance in high frequency trading utilizing DDN’s newly announced A3I AI400X2 appliance.
The STAC-M3, used by banks, brokerage houses, exchanges, hedge funds, and proprietary trading shops, is the financial services industry’s standard benchmark test suite for tick analytics, with results audited by STAC Research. The benchmark specifications are maintained by the STAC Benchmark Council, which has more than 450 financial institutions and vendor organizations, which share technical issues in financial services and develop benchmark standards.
The new benchmark results, audited by STAC Research, show the performance on a single DDN A3I AI400X2 all-flash NVMe appliance connected to 15 servers running the KX Systems kdb+4.0 database in distributed mode via a high-speed InfiniBand switch.1 After completing its previous STAC benchmark tests in September 2020,2 DDN has since released the new AI400X2 appliance and the underlying file system, EXAScaler 6, both of which were utilized in this year’s tests. With these improvements, DDN was able to reduce the number of appliances in the test, while still exceeding 10 of 17 baseline (Antuco) benchmark tests and 19 of 24 scaling (Kanaga) benchmarks, compared to the previous tests using two DDN AI400X appliances.
“While DDN A3I storage systems can scale without limit to increase capacity and throughput, these results demonstrate extreme performance in a cost-effective form factor,” the company said in its announcement. “And, the A3I AI400X2 appliance has a smaller footprint, less complexity and lower overall TCO than competitive storage solutions.”
“Simplifying the management of data-intensive workloads and delivering faster insights to our customers is the basis of our development here at DDN,” said Kurt Kuckein, vice president of marketing at DDN. “These STAC benchmark results reflect the progress we’ve made toward streamlining data management while significantly accelerating analytics and AI workloads – particularly for our financial services customers.”
Other notable DDN results from this STAC testing:
- DDN outperformed a solution using kdb+ 4.0, three NFS-based network-attached flash storage systems and nine database servers3 (DDN said it was faster in 11 of 17 mean response time Antuco benchmarks, including a 11.4x speedup in single-user intervalized statistics, STAC-M3.β1.1T.STAT-UI.TIME)
- DDN also outperformed a solution using kdb+ 4.0, a parallel file system, 15 database servers, and 40 storage servers on a public cloud4 (DDN was faster in 12 of 17 Antuco benchmarks, and 12 of 24 Kanaga benchmarks, including a 16.2x to 19.9x speedup in the four 10-user market snapshots, STAC-M3.β1.10T.YR{2,3,4,5}-MKTSNAP.TIME)
“Like many of DDN’s enterprise customers, financial services firms are looking to modernize their business models with analytics and AI to generate new opportunities from their data,” the company said. “Existing approaches to data management using NFS-based configurations are often complex and inefficient, blocking the path to value because they do not scale to the size of the challenges.”
Supporting Resources:
- More STAC Solution Brief
- More on DDN A3I, EXAScaler
1 https://www.stacresearch.com/news/KDB211014
2 https://www.stacresearch.com/news/KDB200915
3 https://www.stacresearch.com/news/KDB210929
4 https://www.stacresearch.com/news/KDB210507
source: DDN
