Link meaning what? Threadrippers are explicitly 60 lanes on the CPU and 4 from the chipset, totalling in 64 lanes. To my knowledge there aren't any more lanes than that, from anywhere - that's all you're getting (not like that's a small amount of lanes...). According to the article, Intel offers a total of 68 lanes, 44 from the CPU and 24 from the chipset, all of which operate at 3.0 speeds. I'm not sure what you mean by "24 @ the speed of 4". 24 PCIe 3.0 lanes means you could have a x16 slot and a x8 slot. If you were to combine the chipset's resources with the CPU, you could have four x16 slots and one M.2 (x4) slot. Note these would all be running at full bandwidth. Regardless - this is just Intel's way of saying "ha-ha we're better because we have slightly more of a number that makes literally no difference but people will buy into it anyway!" and yet AMD's Epyc is supposed to have 128 lanes. I thought 64 lanes was overkill.

Link meaning what? Threadrippers are explicitly 60 lanes on the CPU and 4 from the chipset, totalling in 64 lanes. To my knowledge there aren't any more lanes than that, from anywhere - that's all you're getting (not like that's a small amount of lanes...).

He means 24 lanes with a maximum of 3.0 x4 to the cpu, meaning that if you where to try and use all 24 lanes at full speed concurrently, the maximum bandwidth to the cpu is still 3.0 x4 total , each lane would have effectively 1/6th the bandwidth it would have if they were a "true" 24 lanes in this situation. If you aren't using more than 3.0x4 off the chipset concurrently, a single x4 slot will achieve 3.0x4 speeds. A possible situation where this would matter is if you were to try and use more than 1 3.0x4 nvme drive concurrently via the chipset , you would be limited to the speed of 1 drive.

Sorry, I'm not following. The chipset is advertised to have 28 3.0 lanes. Where eactly are the other 20 lanes going? The article mentions the x4 lane interconnect, but I get the impression that would cause the opposite scenario, where 4 lanes are lost and the chipset provides 20 that you can actually use.

" For a couple of generations, the Z-series chipsets have adopted the HSIO (High-Speed IO) topology, essentially allowing the chipset to act like a big PCIe switch. Using the DMI 3.0 link from the processor, which by any other name is essentially a PCIe 3.0 x4 link, the chipset can support 20-24 lanes of PCIe 3.0 from it in various combinations. With access to 20-24 PCIe lanes from the chipset, rather than having to bother the CPU for PCIe lanes and reduce potential graphics performance, motherboard manufacturers were able to go overboard with additional functionality. This has included more SATA ports, more networking ports, higher performance networking ports such as 10G, extra storage ports, elements such as PCIe storage in an M.2 form factor, or add special unique features specific to that motherboard manufacturer’s package. While there is technically an uplink bottleneck, that is rarely hit unless a user decides to play with multiple high-speed PCIe 3.0 x4 storage devices."

Turbo frequencies for X299 chips from Anandtech Hope this helps Thanks, Jura

Listing turbo speeds on a giant multicore CPU is just appealing to fagots with little knowledge and lots of money as the only "correct" way to use this platforms it to run them under almost 100% load 24/7, which in turn means base clocks :P

chipset connects to cpu via DMI 3.0 which is a pcie 3.0x4 link, chipset supports up to 24 lanes, the chipset acts as a pcie switch for io essentially. allows you to connect many devices over the 3.0x4 link. http://www.anandtech.com/show/11461/intel-announces-basin-falls-the-new-highend-desktop-platform-and-x299-chipset

Unfortunately without for how long those can be sustained the frequencies mean little. If a 18 core could sustain 3.4Ghz on all core indefinitely, that would be the base frequency and not the turbo.

Listing turbo speeds on a giant multicore CPU is just appealing to fagots with little knowledge and lots of money as the only "correct" way to use this platforms it to run them under almost 100% load 24/7, which in turn means base clocks :P

Fagots with money?? :3eyes: And clearly you are lacking knowledge, many mobos set to allow turbo speeds all of the time regardless of load.

According to what? You have to use an aftermarket cooler, and pretty much the only reason to get these CPUs vs a Xeon is the overclocking, in which case the turbo is somewhat irrelevant anyway. The main reason it's there is so you can squeeze extra performance without boosting every single core. Yeah... if you're going to be that crass about your opinion, you might want to do a little more fact checking.

I was replying to a factory clocks/turbos sheet, which means stock cooler and settings. Ofc u can do crazy things when tweaking/overclocking. But if setting the mobo to permanent all core turbo (so, effectively raising the base clock?) were possible with stock cooling, Intel would be doing it themselves.

My point is amd/intel advertising turbo speeds on giant multicore cpus is pretty much worthless, just an idea of oc capability at best.

Let me stress this again - the i9 series doesn't come with stock coolers. Turboing isn't the same thing as increasing the base clock. Yes, it is possible for all cores to reach turbo speeds simultaneously (which is stupid), but turboing also allows individual cores to reach different turbo speeds too (which isn't stupid). This is to allow you to achieve better performance on tasks that don't demand all cores without making all of them increase, which in turn results in better power consumption. For Intel, the turbo speed you get on all available cores is nothing more than a marketing gimmick and might as well be treated as the base clock. For anything involving less than 100% of the cores, turbo speeds are far from worthless.

So if I understand this correctly, this is works very similarly to how USB hubs work (I am aware there are distinct differences). How dumb, and misleading (sure fooled me anyway).