The Test and Measurement Industry Demands Enhancements to IEEE 488
A Letter to the Editors of Trade and Technical Publications
From Robert Canik, Chairperson of the High-Speed 488 IEEE Working Group
A Response to Hewlett-Packards' December 10, 1997 Press Release

December 15, 1997

Dear Editor,

Recently, Hewlett-Packard (HP) issued a press release lobbying IEEE members to oppose an industry supported performance update to IEEE 488 (GPIB). The proper name for the update they refer to as HS488 is high-speed 488.

Although I feel it is inappropriate to publicly comment on a standard that is in the balloting process, I feel compelled to stand up for the efforts of the working group and not allow a single company to attempt to influence the outcome of an IEEE vote.

For the past three years I have chaired the IEEE working group, under IEEE TC-8, responsible for developing high-speed 488. In 1994, driven by widespread industry and user interest, the IEEE endorsed the formation of a working group to improve data rates across GPIB.

Individual volunteers from Test and Measurement Companies, including Hewlett-Packard, Tektronix, CEC, Ines, National Instruments, Iotech, and others, set out to develop a standard that met two primary goals. First, to maintain complete compatibility with the installed base of IEEE 488 equipment/software and second, to provide an order of magnitude performance increase over GPIB.

After three years of hard work and detailed simulation and testing, this working group has successfully met those goals and, as a majority, has voted to submit this robust specification for formal IEEE approval. This specification is now in the hands of the IEEE voters.

I would like to clearly state that high-speed 488 delivers complete compatibility with the installed base while providing an 800% performance increase in data transfers. For example, users of HP instrumentation have great reason to celebrate high-speed 488. Recent benchmark testing has shown that the new HP Infinium Oscilloscope gains a 30x performance boost in raw GPIB data rates (7.5 Mbytes with high-speed 488 vs. 232 kbytes/s). Because the Infinium is essentially a PC, engineers were able to swap the standard GPIB interface for a high-speed compatible GPBI interface. One can imagine the benefits this offers for test and measurement systems.

Satisfied with the level of testing and robustness, seven instrument vendors have already endorsed high-speed 488 including LeCroy, Anritsu, Adtech, and Nicolet Technologies. High-speed 488 is now a field proven technology with six of those companies shipping compatible instruments with no reported incompatibilities or problems.

"LeCroy has always been on the forefront of data transfer technology in Oscilloscopes," states Mike Lauterbach, Director - Product Management. "We intend to stay there and high-speed 488 is a good way of achieving that end."

Three months ago, HP suddenly began a campaign against high-speed 488 even though they have been involved in the working group from the beginning.

"After breaking an IEEE bylaw in a failed attempt to stack the IEEE ballot, it is disturbing that HP now chooses to wage a marketing campaign as a further attempt to derail the high-speed 488 standardization and undermine user confidence and support", comments Marilyn Miner, TC-8 chairperson for IEEE.

In fact, HP's objections have already been resolved in the working group, as the IEEE is the appropriate forum for resolving technical issues. HP's public campaign is based on misleading and inaccurate statements.

It is ironic that despite this public campaign against high-speed 488, HP already publicly endorsed the high-speed 488 technology when they joined the VXIplug&play Systems Alliance in 1994.

High-speed 488 is an improvement that the industry has demanded. Three times in history, based on industry demand, the IEEE 488 standard has been updated to keep pace with technology and user needs. Continuing these improvements, high-speed 488 provides faster GPIB, requires no test software modification to use, and maintains compatibility with the installed base of GPIB equipment.

I think it is a given that new technologies will find their way into test and measurement. I support all initiatives for giving users alternatives. USB may fulfill a niche in instrumentation but its maximum speed to date is 12 Mbits/s (1.5 Mbytes/s). Firewire (1394) is at least a year away from commercial products. Of greater concern is the preservation of user investments in GPIB equipment and software. Are we to turn our backs on the installed base and force them to abandon their existing investments for new technologies - or do we have a responsibility to ensure continued success with IEEE 488?

The public lobbying of a single company based on what appears to be a non-technical agenda has no place in standards activities that clearly deliver demonstrated benefits to users. To maintain objectivity, IEEE does not consist of company memberships - it is a professional organization made up of individuals so standardization activities can be carried out on technical merit and without bias or political motivations. The future of IEEE 488 should not be dictated by the agenda of an individual company nor should we abandon IEEE 488 as HP alludes. The users will suffer the greatest.

Sincerely,

Robert Canik
Chairperson of the High-Speed 488 Working Group