The STAC-TS benchmarks for measuring timestamp accuracy have been designed by a working group of trading organisations, timing-related vendors, STAC, and Metamako, who envisage them becoming the industry standard. The partners highlight the importance of being able to verify timing to ever-tighter precision, for financial and other industries that have more and more stringent rules around the precise timing of actionable events.
Metamako presents itself as the first vendor to have a product measured with these benchmarks, which require laboratory-levels of time and frequency metrology. To ensure the most accurate results possible for the absolute accuracy benchmark, Metamako used high-resolution picosecond-level timestamping equipment.
Metamako’s MetaWatch application, with a timestamp resolution of 1 nanosecond, is the first product which has been through the STAC-TS benchmarks for network timestamping. The results demonstrated that for one of the device ports, 100% of the timestamps were accurate to 0.5 ±2.5 nanoseconds relative to an external atomic reference timing source.
The challenge in measuring timestamp accuracy has – the company continues – always been that it is entirely dependent on the accuracy of the measurement device being used. The new benchmarks provide a reference facility for the industry and consist of three stages: (i) the accuracy of a port relative to a reference time source, (ii) the accuracy of ports relative to each other on the same device and, (iii) the accuracy of ports relative to each other across different identical devices.
MetaWatch was proven to be accurate to, and far exceeding, any current regulatory standards. This powerful application, designed for Metamako’s high-performance hardware, combines several aspects of network monitoring functionality in a single device. These include tapping, aggregation, deep buffering (up to 32 gigabytes), nanosecond timestamping and time synchronisation. MetaWatch replaces up to 30 passive taps, an aggregation and timestamping switch, along with media converters, patch panels, and all other layer 1 switch use cases.
Dr. Dave Snowdon, founder and CTO of Metamako, says: “Until now vendors had no way of proving that the timestamps their devices apply to captured packets are as accurate as they claim. Nor was there a good way to compare the capabilities of different devices: there has been no independent way of verifying this. With these new STAC benchmarks there is a new gold standard. It’s a really important development for the industry, not least because there is now a way to demonstrate to regulators that the timestamps are indeed what they claim to be and that financial institutions are compliant with requirements such as MiFID II.”
Peter Lankford, Director of STAC, commented: “Vendors have traditionally expected clients to trust their claims of packet-timestamp accuracy. But major trading organisations in the STAC Benchmark Council, who have both business and regulatory needs for accuracy, have been keen to establish vendor-independent benchmark standards. Accurate network timestamps are important for many reportable trading events as well as for time-sync testing of other hardware and software components. Developed in conjunction with trading firms, Metamako and the rest of the STAC-TS Working Group, STAC-TS provides the most accurate and comprehensive benchmarks available and are now ready for use on other products.”