The point of standards is to bring an industry together, to avoid duplication of effort, and to reduce risks associated with adoption of technology that may lock a user into a single vendor. These are some of the reasons why Breker was glad to see the creation of the Portable Stimulus working group within Accellera and actively participated in it since its inception. We donated technology and invested more time and effort, as a percentage of company size, than any other player. We were also glad to see the release of version 1.0 of the standard at DAC in 2018 – a huge step for the industry.
But was it enough? Some standards are ratified having been developed and refined by a single company and successfully proliferated across an industry, prior to donation to a standards body. Others are designed by committee and therefore run the risk of an unproven body of work captured as a hard to change standard. Sometimes this works well, other times, less so. It is only after the fact that you know if the committee got it right. So far the Portable Stimulus Standard is being tested by a relatively large number of companies with success, but there are still missing elements for a scalable solution.
A standards group, such as Accellera, has to walk a fine line. It relies on funds coming from member companies to pay for aspects of the development work. While a vast majority of the efforts are provided on a volunteer basis – effectively funded by member companies – committees often require rooms to meet, or professional writers to get documents completed. Member companies expect to get something in return for the time and effort involved, and thus there are restrictions placed on non-members.
Accellera is one of the most open standards bodies in the semiconductor industry. Many of its standards, especially those that have been donated to the IEEE, are available for free to the community. What it retains for its members is the ability to participate in the work of the committees, to be able to influence direction and to have early access to donated materials.
One of the initial requirements for the Portable Stimulus standard was for a hardware-software interface (HSI) and both Breker and Vayavya donated technology to form the basis of this. Unfortunately, it did not make it into the rather accelerated version 1.0 of the standard. Breker believes this concept is too important to not be part of a Portable Stimulus solution and offers a full version as an extension within our products.
The early adoption of the HSI interface with Breker’s tools may produce feedback useful to the committee once it again starts to consider extensions to the standard, and we will be delighted to share this with the committee.
Other versions of an HSI are also apparent in the industry. Mentor seems to have light version and now Agnisys, a company that was not part of the original PSWG and, therefore, had no access to donated material, has produced a similar technology. Although this targets a slightly different purpose, it also provides some HSI capability.
So now we have multiple versions of the HSI beginning to permeate the industry. The Portable Stimulus committee is again discussing requirements for the next release of the standard, but it is unlikely that this path will provide a quick resolution of this issue. Some other mechanism is required that solidifies this solution and delivers it to the market quickly.
One possibility is for the Verification 3.0 companies to get together and agree on this outside of the committee. This may well enable the interface to be properly tested in the field and go a long way toward being a de facto standard before the committee even considers it. Verification 3.0 will hold its first Innovation Summit on March 19th and this may be one of the hot topics for discussion.