Image: Intel

Intel’s upcoming Z490 (Socket 1200) motherboards were designed with PCIe 4.0 support in mind, but despite having the necessary components to do so, they won’t be flaunting those speeds. Sources claim that Intel had trouble implementing PCIe 4.0 into the Comet Lake chipset due to unacceptable amounts of jitter, so retail Z490 motherboards will perform no better than PCIe 3.0.

This is a very bad showing for Intel, being that AMD’s X570 chipset offers full PCIe 4.0 support. Enthusiasts who own a Zen 2-based Ryzen processor and flagship motherboard can enjoy breakneck transfer speeds of over 4 GB/s , but evidently, prospective Comet Lake-S owners won’t be getting the same privilege. It’s an interesting turn of events for a company who initially berated the standard but was forced to adopt it due to competition.

Motherboard manufacturers may not be too happy, either. As noted above, Z490 boards were engineered specifically for next-generation PCIe, but that was all for naught. Vendors are expected to amend current designs (i.e., remove certain components) in order to recuperate costs.

Intel has remained tight-lipped on the release date of its Comet Lake-S processors and Z490 motherboards, but rumors have pointed to a Q2 debut.

Join the Conversation

11 Comments

  1. Yea that sucks for Intel. I could even see Intel not bothering to release this to the consumer line and wait until they have something feature complete to release. But then all of those motherboard partners would **** a brick at lost sales and sue Intel. They are in a hard spot that’s for sure.
  2. I thought Intel were on record several months ago (before the Ryzen 3000 announcements) that they were skipping Gen 4 and going straight for Gen 5?

    Side note. Why does the front page want me to sign in now instead of taking me to the forums? That’s annoying…

  3. I thought Intel was skipping PCIe 4.0 as well. Additionally, this isn’t surprising if they had planned on it as it would be costly to implement. Right now, they need to be as cheap as possible to move units.
  4. Don’t see the point of skipping 4.0 assuming 5 is more costly and complicated. 4 is barely starting out, doesn’t make a lick of sense to jump to 5 for no reason.. servers on the other hand im guessing there is more use there.
    I think this is more a demonstration of profound disfuncion at Intel… This is looking a lot less than a "sleeping giant" or resting in laurels or some such ****… It looks like straight up disfuncion in a functionally broken company and it hasn’t resolved yet obviously.
    Intel gets way too many passes, with this performance their shares should be clobbered… Their sales should have tank brutally…. Yet they performing freaking near catastrophically, and make billions… Something to say about size an inertia alone.
  5. Don’t see the point of skipping 4.0 assuming 5 is more costly and complicated. 4 is barely starting out, doesn’t make a lick of sense to jump to 5 for no reason.. servers on the other hand im guessing there is more use there.

    Well, if 4.0 comes out in 2019, and 5.0 is ready to roll (the PCI 5.0 standard is done).. then why waste your time even fooling with 4.0?

  6. But the issue is that isn’t what they are doing. I could see the 4000 line of new CPU’s from AMD coming actually have PCIE 5.x AND better performance than Intel. that would be some real pain for Intel if AMD is 2 GENERATIONS ahead on I/O.
  7. Well, if 4.0 comes out in 2019, and 5.0 is ready to roll (the PCI 5.0 standard is done).. then why waste your time even fooling with 4.0?

    Well if its as easy, and same cost, then sure. That seems unlikely.
    This bullshit of ddr5 and and pcie 5 is pure , oh look , wait for meeee! Its Fud, they have nothing.
    God, what would we be saying of any other company that FAILED to implement an already old standard. This is not s slip up it is not a little thing.. its a company that can’t execute for ****. And its looking real bad for real long. If Keller quits Intel is ****ed. If Keller is strangled by a **** company culture Intel is ****ed, is Keller sets the future to high and too complex and fails, Intel is ****ed. And and yes if he only delivers an increment , Intel is ****ed.

  8. Well if its as easy, and same cost, then sure. That seems unlikely.
    This bullshit of ddr5 and and pcie 5 is pure , oh look , wait for meeee! Its Fud, they have nothing.
    God, what would we be saying of any other company that FAILED to implement an already old standard. This is not s slip up it is not a little thing.. its a company that can’t execute for ****. And its looking real bad for real long. If Keller quits Intel is ****ed. If Keller is strangled by a **** company culture Intel is ****ed, is Keller sets the future to high and too complex and fails, Intel is ****ed. And and yes if he only delivers an increment , Intel is ****ed.

    You’re 100% being over dramatic. Current server builds with Intel farms are still Intel. It’s going to take something truly massive to make enterprise users switch off of Intel in Mass. (Like… you know Dell having model parity between Intel and AMD CPU’s on their website. Yep that would probably do it.

  9. You’re 100% being over dramatic. Current server builds with Intel farms are still Intel. It’s going to take something truly massive to make enterprise users switch off of Intel in Mass. (Like… you know Dell having model parity between Intel and AMD CPU’s on their website. Yep that would probably do it.

    ****ed in execution.
    I am not understating inertia in the market place hence they making money off of products that are crap.
    Matter of fact, i have doubts AMD will EVER grow to any significant size for any significant time with the market as is. AMD is too dependent on so many other companies to move their products.. companies with their own inertia. Might not be possible but I can only see AMD growing to any meaningful size would be for them to be an all solution vendor, whole servers, whole super computer, whole everything… Intel just needs to perform a little better (which shouldn’t be hard considering their current performance, but apparently it is since no pcie 4) and inertia will take care of the rest. Well.. Maybe if AMD executes flawlessly for the next 5 years or more, well maybe they will grow to some significant size.
    What I think its being understated how much of a red flag this "slip up" is. You never even heard ? about pcie 4 being difficult to implement or some ****, yet Intel, just could not do it. An old, late standard. None of these things are priced in their stock… I guess another benefit of blind faith.

  10. I don’t think it was that Intel couldn’t do PCI 4.0 because it was beyond their capabilities.

    I think they just looked at it practically – how much would it delay product to market, how much would it cost to implement new motherboards and chipsets, how much gain does it present

    Add all that up and they just didn’t see it being worthwhile at this time. Doesn’t mean next gen won’t have it… but having to rush it out an already suffering product line that needs to get out to market ASAP, yeah I can understand why they didn’t do it this go around just to have what amounts to a marketing bulletpoint.

Leave a comment