"Experts in the PCIe Electrical Workgroup carefully analyzed a number of target bit rates for the next generation of PCIe architecture, taking into consideration several key factors, including our ability to continue using low-cost materials," said PCI-SIG chairman Al Yanes in a statement. "We have concluded that 16 GT/s is a feasible technical solution that satisfies our member companies' requirements."
When The Reg spoke with PCI-SIG's Serial Communications Workgroup chair Ramin Neshati at this year's Intel Developers Forum, he told us "We think we have line of sight to get to 16 gig on copper – maybe even higher."
"Even higher" didn't happen – although, to be frank, there's little reason that it needed to. AT 8GT/s, PCIe 3.0 is as fast as normal humans will ever need.
Well, "ever" may be quite a long time, but Neshati was of the opinion that first, second, and third-generation PCIe would be "good enough for the broad spectrum of applications for a long, long time" – and when we asked him exactly what a long, long time meant, he told us: "Forever."
PCIe 4.0 – which Neshati called "a boutique-type application for very few topologies" – will, however, be welcomed by those building systems in the rarified world of high-performance computing, and may possibly also find its way into data centers tossing around obscene amounts of bits and bytes.
The PCIe 4.0 spec, which is expected to be released sometime in 2014 or 2015, will be backward-compatible with earlier PCIe architectures. The PCI-SIG also says that it is "technically feasible" that it can be implemented with today's run-of-the-mill, low-cost silicon technologies at "approximately" PCIe 3.0 power levels. ®
The PCI-SIG's announcement contained our favorite statistic of the day – and perhaps the month: "Approximately 24 billion lanes of PCIe have shipped in the marketplace since its introduction." Hmm ... that'd be about 3.4 PCIe lanes for every living human being. Are you using yours wisely?