Richard Solomon has heard the rumblings over the years. As vice president of PCI-SIG, the organization that controls the development of the PCI-Express
Well, the thing is that 3 years looks like "too long" but eventually the spec is held by the timeframe of having actual silicon. Even if it's not 1 year or 2, at least is not 5 or 7.
That's probably the problem of standards. Everyone has to agree to a new spec, instead of a company offering double the PCI Express bandwidth and latency that, low and behold, only works on their hardware and will charge for royalties.
3 years look like a lot, but it's cheaper than vendor lock-in, which everyone has afraid of since is in that moment your business is controlled by other business.
Well, the thing is that 3 years looks like "too long" but eventually the spec is held by the timeframe of having actual silicon. Even if it's not 1 year or 2, at least is not 5 or 7.
That's probably the problem of standards. Everyone has to agree to a new spec, instead of a company offering double the PCI Express bandwidth and latency that, low and behold, only works on their hardware and will charge for royalties.
3 years look like a lot, but it's cheaper than vendor lock-in, which everyone has afraid of since is in that moment your business is controlled by other business.