Abstract
Financial Market IT solutions increasingly depend on ultra low latency message processing and target microseconds latencies in order to provide traders with a competitive advantages over their peers. Some solutions are available on the market, ranging from general purpose systems with advanced network cards to specialized hardware solutions based on FPGA. The novel IBM PowerENTM “Edge of Network” processor integrates network interfaces with functional accelerators and multi-threaded cores. This paper describes the design, implementation and performance evaluation of a Market Data Feed solution based on the PowerEN processor. The paper details the SoC mechanism designed to reduce latency and increase throughput and shows how these can be used to build a Market Data Feed solutions that achieves performance figures usually obtainable only by special hardware. This prototype described achieves an average latency of 6.7 μs for an OPRA v2 feed and is able to process 1M updates per second on 2 hardware thread, achieving more than 16M updates per second on a single chip solution without consuming all the resources and thus being able to also run customer software.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
FAST for OPRA. URL http://www.opradata.com/
Agarwal V, Bader DA, Dan L, Liu LK, Pasetto D, Perrone M, Petrini F (2009) Faster fast: multicore acceleration of streaming financial data. Comput Sci Res Dev 23(3–4):249–257
Bho: Integrated virtual Ethernet adapter technical overview and introduction. IBM Redbooks (2007)
Exergy: URL http://www.exegy.com/
Exergy STAC-M1: URL http://www.stacresearch.com/
Flume Data Inc: URL http://www.flumedata.com/
Gavrilovska A (2009) In: Attaining high performance communications: a vertical approach, 1st edn, chap 13. Chapman & Hall, London, pp 305–327
Franke H, Xenidis J, Bass B, Basso C, Woodward S, Brown JD, Johnson CL (2010) Introduction to the wirespeed architecture and processor. IBM J Res Develop 54(1)
InRush STAC-M1: URL http://www.stacresearch.com/
Johnson C, Allen DH, Brown J, Vanderwiel S, Hoover R, Achilles H, Cher CY, May GA, Franke H, Xenedis J, Basso C (2010) A wire-speed PowerTM processor: 2.3 GHz 45 nm SOI with 16 cores and 64 threads. In: Proceedings of the IEEE int solid-state circuits conf, pp 104–105
van Lunteren J (2006) High-performance pattern-matching for intrusion detection. In: Proceedings of the IEEE INFOCOM, pp 1–13
Mok AK, Konana P, Liu G, Lee CG, Woo H (2004) Specifying timing constraints and composite events: an application in the design of electronic brokerages. IEEE Trans Softw Eng 30:841–858
Morris GW, Thomas DB, Luk W (2009) In: High-performance interconnects, symposium on Fpga accelerated low-latency market data feed processing, pp 83–89. http://doi.ieeecomputersociety.org/10.1109/HOTI.2009.17
NYSE L5673: URL http://www.nyxdata.com/
Sadoghi M, Labrecque M, Singh H, Shum W, Jacobsen HA (2010) Efficient event processing through reconfigurable hardware for algorithmic trading. Proc VLDB Endow 3:1525–1528
Schneidert S, Andrade H, Gedik B, Wu KL, Nikolopoulos DS (2010) Evaluation of streaming aggregation on parallel hardware architectures. In: Proceedings of the fourth ACM international conference on distributed event-based systems, DEBS ’10. ACM, New York, pp 248–257
Subramoni H, Petrini F, Agarwal V, Pasetto D (2010) Intra-socket and inter-socket communication in multi-core systems. Comput Archit Lett 9(1):13–16
Zhang XJ, Andrade H, Gedik B, King R, Morar J, Nathan S, Park Y, Pavuluri R, Pring E, Schnier R, Selo P, Spicer M, Uhlig V, Venkatramani C (2009) Implementing a high-volu low-latency market data processing system on commodity hardware using ibm middleware. In: Proceedings of the 2nd workshop on high performance computational finance, WHPCF ’09. ACM, New York, pp 7:1–7:8
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Pasetto, D., Lynch, K., Tucker, R. et al. Ultra low latency market data feed on IBM PowerENTM . Comput Sci Res Dev 26, 307–315 (2011). https://doi.org/10.1007/s00450-011-0166-0
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00450-011-0166-0