The politics of standards

Standards are rarely created for the benefit of the industry, although that can be a useful byproduct. More often, they run an industry for someone else’s benefit.

popularity

Standards are often seen as an industry coming together to agree on a common solution to a common problem, but there are times when this couldn’t be further from the truth. Having been involved in standards at every level, from participant to chair of various standards, some have succeeded, some have not, some are experiencing significant adoption while others are withering along the way, and some are being blocked. for those who are disadvantaged. for them. The truth of the matter is that standards can be political hornets’ nests that are designed to both divide and unite.

There are many types of standards, and the most successful are probably the de facto standards that have become open standards. In our industry, one of the biggest success stories was Verilog. Originally developed by Gateway Design Automation, it was acquired by Cadence and eventually became an open standard. Since then, it has been expanded and built. But this only happened because it was threatened by the release of VHDL, a DARPA-defined language that began to see widespread global adoption. The threat was enough to drive standardization into an open standard.

There have been other cases where a standard has been stalled to prevent a vendor from having a market advantage. This often happens when a company has robust technology that is offered as standard. While the industry may agree on its merits, they want to delay ratifying the standard until they have an implementation that they can announce at the same time as others. At worst, they do this by inserting something into the standard that makes it difficult for the donor company to implement, since they must maintain backwards compatibility with their existing capability. In some extreme cases, this can result in the donor company being late in announcing a fully supported implementation.

I know of an example where there was significant disagreement over a detail of a standard. It turned out that a member of the standards group was unable to implement a specific feature. This was not disclosed by that member and therefore, even when a compromise was offered so that both alternatives could be used, the compromised vendor attempted to delay a vote on the issue. Rather than accept the result, they left the standards group and dropped all support for the standard and worked towards an alternative.

I have also seen examples where the creation of a standard was done to disrupt a market where one or more companies did not have a competitive product. By partially excluding the leading market share company, it can be designed in such a way that it has to catch up and lose market share in the process. Many variations of this have happened, and in some cases have completely blindsided a competitor.

Why am I writing this now? While conducting interviews for my UCIe story, I never did an investigation where more people asked to be off the record. They want to show support for one standard and denigrate the other, but they don’t want their companies to be associated with those comments. This tells me that there is the party line, and then there is the emotional aspect of the race between two competing standards. In some cases, it is clear that they want to preserve the investment they have previously made and are concerned that that investment will be lost. In other cases, it’s clear they don’t like the way large amounts of money can buy anything, even if the solution isn’t that great. Think back to the VHS-BetaMax wars of the past.

Companies do not participate in standardization efforts for altruistic purposes. They do it because it will expand their market or benefit them in some way. Participating in standards is expensive and decisions are based on capital allocation, just like any other product development. Engaging in standards often requires a lot of politicking, a lot of time, a lot of patience, and sometimes the ability to shut up.

brian bailey

brian bailey

(all messages)

Brian Bailey is Technology Editor/EDA for Semiconductor Engineering.

Leave a Reply

Your email address will not be published. Required fields are marked *