Intel Z690 Chipset for 12th Gen Alder Lake Leaked

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.


Recently, an image of an Intel Z690 chipset was leaked by a member of the Chiphell forums. The Z690 will be the foundation for the next generation of Intel motherboards. It is significantly larger than the current Z590 chipset, but how that will affect its performance remains to be seen. Launch for these next-generation products are expected around November 19th, but that is not set in stone.

Intel Z590 and Z690
Intel Z590 and Z690


Here is what we currently do know about this latest chipset. The Z690 will use 8 lanes of Direct Media Interface (DMI) 4.0 to communicate directly with the CPU. Up to 6 SATA 6 GB ports and 40 USB devices (ranging from USB 2.0 through USB 3.2 Gen 2×2) can be used simultaneously. A single x16 PCIe 5.0 or 2×8 PCIe 5.0 will be directly connected to the CPU while up to 12 lanes of PCIe 4.0 and 16 lanes of PCIe 3.0 will be routed through the chipset. Below is a complete details chart as we know it today.

Z690 Details
Z690 Details

Stay tuned as we will report additional information as it becomes available.

-John Nester (Blaylock)


Leave a Reply

Your email address will not be published.


  1. There are a few articles that reference this date, though it has not been confirmed and is merely speculation.

    This latest speculative date was started by WCCFTECH I think, though I have no solid evidence. Earlier speculation was October 5th. That obviously has passed.
    Earlier speculation was October 5th. That obviously has passed.

    Isn't Oct. 5th the date for Win11 release? Given Alder Lake really needs Win11 to make the most of it, I think that sets the earliest date. How much would MS and Intel work together to launch on the same date? Of course, later in the year remains open. IMO avoiding the time around US thanksgiving and christmas would seem logical, and that November date is getting close to Thanksgiving.
    Correct. I meant speculation for launch on October 5th has passed and mid to late November is the new speculation. Nothing has been announced so we won't know for sure until we get closer to the actual launch date.
    I heard that DDR5 samples will be ready for the premiere in the 2nd half of Nov. This is what two vendors told me. On the other hand, someone else promised motherboard samples for the second half of Dec/beginning of Jan so there can be a delay or a "paper premiere". I guess we will know for sure in ~2-3 weeks.
    The 5Gb LAN is interesting to me. What's the point? I said the same thing about 2.5G but, seriously, who is going to bother putting it down on the board with a PHY? Plus 5Gb is a weird half step that some network gear might have issues with. Anyone that cares is going to have 10GbE gear already imo.
    Lots of higher end mobos already have 2.5G, and the cost difference to set up a home network at 2.5G is a LOT cheaper than at 10G if you want more than 2 copper ports. This has been a long time frustration of mine, that there are no 10G switches that don't cost well into the hundreds to get more than 2 ports, which would be needed for anything other than a point to point connection. There is a 5 port 2.5G switch for around a hundred or so. That'll do for a higher performance part of a home network. Once 5G start rolling out I'd guess it'll start replacing 2.5G while still far below 10G pricing. Also, if the mobo doesn't have a high speed NIC, 10G copper is about 3x the cost of 2.5G too. If you really need 10G, you'll have to pay for it. 2.5G isn't quite no cost over gigabit, but it is a lot closer.
    BTW 2.5G is fast enough for single HD network transfers to not be a limit, which is my personal use case. Moving stuff to bulk storage. 5G would be about the sweet spot for SATA SSD transfers.
    10GbE is on the market for a couple of years. I really have no idea why we could see 10GbE even in desktop motherboards some years ago, and later "high-end" series had 2.5GbE. Now is 5GbE ... what popular network hardware is even supporting that? A big part of the consumer market moved to 2.5GbE in cheaper solutions and 10GbE in more expensive. If 10GbE was more popular then it would be already much cheaper than it is right now. Instead of pushing 10GbE, we can see some weird solutions like 2.5 or 5Gb.
    I had at least 3-4 motherboards with 10GbE LAN ~3 years ago. After that, "refreshed" versions came out with 2.5GbE. So I made a 10GbE network at home and new motherboards have slower LAN. 5GbE is another stupid standard that isn't even supported by most devices. Most are like 1GbE/10GbE. New cheaper are like 1GbE/2.5GbE. Many that support 10GbE don't support anything between 1GbE and 10GbE so you can have a 10GbE switch that won't support 2.5GbE or 5GbE.
    Business series are already going for 50/100Gb while the new great desktop standard will be 5? ... For me this is a stupid idea as maybe it's faster but requires replacing all network devices to make it work and this standard is not something that will last for long years. Like 2.5Gb started to be quite popular maybe a year ago so those who recently replaced all devices at home won't be happy.
    Not to mention that most motherboards with 2.5GbE Intel LAN had faulty controllers and were working at 1GbE anyway. It was fixed only in some part. I hope that Intel won't repeat that mistake.
    I think I found the reason for 5G to exist in addition to 10G. Cable support. Apparently you can run 5G full distance over cat 5e, whereas 10G requires cat 6/6a depending on distance. For a home user replacing a few cables this probably isn't a big hurdle, but there's probably a lot of sites where 5e cabling was put in a long time ago and replacing it will not be trivial. 2.5G/5G kit can rotate in over time and keep costs down compared to making the jump straight to 10G.
    BTW I see these connection speeds as more for enabling better client access, not necessarily for infrastructure. I think the industry had sat at gigabit ethernet on endpoint devices for far too long that even wifi has overtaken it.
    Depending on the distance you may need cat 6/6e on 5Gb too. Somehow, good cat 5 and 6 cables are about the same. This is just my experience working with this stuff. I have clients who moved to 10G having 10 year+ old infrastructure and there are no issues. Of course, it's not guaranteed but at least cables are no issue in a typical home environment. We are still talking about desktop solutions so I'm not even thinking about replacing the whole network for a larger company. In business, no one is even thinking about 2.5/5Gbps. All are moving from 1Gbps to 10Gbps.
    Another thing is the availability of 5GbE devices. Check the market for 5GbE routers and switches. Good luck finding anything good. Most are 2.5GbE or jump to expensive 10GbE as anything with more than two 10GbE ports is business-class hardware. So we may have 5Gbps LAN cards that won't have support on the other devices side. My point is that instead of using one standard, then we already have four that are not always fully compatible (depends on the network devices).
    To my understanding from searching earlier, cat 5e supports 5G to 100m. Cat 6 supports 10G to 55m, and 6a to 100m. If it works better than that, great. In my box of cables I have a mix of 5, 5e and 6. I did have trouble getting 2.5G until I realised some were cat 5 cables. Replaced with 5e and it was fine. And these are short lengths too.
    I think part of the problem is that 10G is an ancient standard in comparison, and was originally only really used for infrastructure, not client access. These newer offerings I feel are more client focused. Yes, it will take some time to become available. I recall the time between 2.5G being announced and affordable hardware was longer than I like, and even now I'd like to see cost go down more as it still far from gigabit cost. 5G will likely take some time to be adopted.
    There's one more reason, which I don't give seriously for the existence of 5G. All the 5's! DDR5. PCIe 5. Why not 5G ethernet too? :D
    I highly doubt that 5Gb will be widely popular as right now there are barely any 5Gb devices and it takes time to release new products. It may move to 10Gb or something new as 10Gb isn't even anything new anymore. I wouldn't be surprised if AMD released their next-gen with 10Gb LAN to beat Intel. 2.5Gb jumped on the market without any earlier info and there were a lot of new motherboards with 2.5Gb LAN but no switches/routers. After a year+ a lot of cheaper devices were released so now there is a wider offer but it's still not popular.
    I can agree that 10Gb was designed as a faster business-standard. The same now is 50Gb and 100Gb. But in the same way, was 1Gb a looooong time ago. 2.5/5Gb are "home and small office" standards which were released after the 10Gb hit the mass market ... and I really have no idea why they exist. It's like we may already need 10Gb for some solutions and manufacturers give us 25-50% of what we could use. For me, a stupid idea is to invest in 2.5/5Gb infrastructure as it will be replaced soon. So yes, it can be a good option for a home or small office and that's all.
    Think that ISP in Poland are offering 1Gb or 10Gb LAN routers/modems as some of them already offer 10Gb fiber connection. It works as 8Gb but still more than 5Gb. If they already offer that then expect 2-3x higher services in 1-2 years. They are not offering 2.5Gb/5Gb devices at all and I highly doubt that ISP will invest in replacing thousands of devices.
    Pricing will always be a factor, and higher speeds mean higher cost.
    For example, 2.5G NICs were around £30 when I last picked some up. The 5 port 2.5G switch I have is £123 on Amazon right now.
    To compare, cheapest 10G NIC is ~£83 on Amazon, and I can't find any switches with more than 2 copper ports for under £300, and I haven't seen any with 5 or more copper ports <£500.
    I kinda accept that if 10G pricing comes down faster, it would largely negate the need for 5G to exist. But I think it is still a nice stepping stone that can be used as fall back if cabling or other equipment can't handle 10G.
    There also comes a point, people who really need higher performance will pay for it. If a consumer level product has to make a choice of increasing pricing to support 10G, or make the product cheaper at a lower speed, we have often seen lower pricing win over technical spec.
    To me 2.5G is something that makes a tangible difference over gigabit. Better than 2x transfer to my store is a benefit. 5G/10G wont be so much benefit unless at some point in distant future I go all SSD storage. Note I don't traditional raid HDs as it doesn't work well with smaller drive numbers. I'll certainly take higher speeds than needed if available for free, but if I have to pay for it, it will be a choice that has to be considered. The higher end exists, but there is also a much bigger spread in cost consideration.
    I have this one at home, £230, 5x 10GbE -
    8-9 10Gb port version costs £320 -
    I also had 3 PC with integrated 10GbE. One left as 2 others were with older motherboards that I sold. Everything new has 2.5Gb except the overpriced "professional" series.
    Btw, if you compare the performance then you will see that most 1Gb cards can make ~900mbps. Most 2.5Gbps somehow can make ~1.5Gbps. If you use Intel 2.5Gb NIC then often results are close to 1Gb NIC. I wasn't testing 5Gb as there were no such NIC anywhere. 10Gb can usually make ~8-9Gbps.
    I bet that Intel moves to 5Gb as they improved their faulty 2.5Gb NIC and wanted to change it anyway. Probably also as you said, everything 5 looks better ;)
    Ok, it is great 10G pricing seems to have moved down a little bit since I last looked for hardware, although the prices I see right now for the two items linked above are £280 for the 5 port from Amazon, and £440 for the 8 port from marketplace seller. The £320 model is "used" so not a fair comparison, although a potential bargain for someone.
    I've not had trouble capping gigabit just doing Windows file transfers. I've not fully tested at 2.5G since my transfers are HD limited to around 220GB/s, keeping in mind that Windows rates are user date, and network rate is for the link, including protocol overheads. Again, for my usage at least, 2.5G is a great stepping stone to beyond gigabit. I'm getting around 2x which helps a fair bunch. I never justified to myself spending much more for 10G to enable that.
    My understanding of the Intel 2.5G problem was some kind of timing problem which meant they didn't work with all connected hardware. Not all 3rd party hardware was affected. Since I never owned Intel 2.5G (all mine are Realtek) I didn't look into it much. It is unclear to me if it is an unfixable hardware fault, since it looks like many report it resolved through firmware + driver update of the NIC.
    On the earlier suggestion that AMD might try to jump Intel again by going straight to 10G for next gen, I do wonder, what is the relative price like on the interface hardware that will still be required? If higher perf comes at little to no extra cost, go for it. But I'm not sure that will be the case here.
    Z690 has everything I would need. I wouldn't mind a higher speed integrated NIC as I could run a direct line from a second LAN port from my rig to the 10GBe on my NAS in my office.
    I'll hold off on the DDR5 migration until the speeds and latencies for enthusiasts settle out, though. I'm not usually an early adopter on new memory tech. Also need to wait on the new big/little architectures to mesh well with Win11. I'm just getting used to Win10!
    Glad I got my upgrade in this year, so many changes coming up with the Intel little big cores , Ryzen 3d stuff , and new ddr5, I am sure prices aren't going to go down, it probably will be a good year to sit on the sidelines and see how it all turns out. :)