From your link "According to the organization that sets the standards for the USB interface, discrete USB 3.2 controllers capable of supporting the standard's new 20 Gb/s Type-C mode will be available this year." - those chips have to turn up in motherboards, those motherboards have to turn up in systems, then those systems actually have to ship.
Not really. For example Sonnet has a way to add USB 3.1 gen 2 to some Macs Pros directly.
https://www.sonnettech.com/product/allegro-usbc-pcie.html
(the is a 4 port version as well but not going to get max out at 40Gb/s over a x4 PCI-e v3 link. Even more so a v2 link in a Mac Pro 2009-2012. )
USB 3.2 default to the motherboard would be an essential component to getting higher deployments out in a quicker time frame. However, it isn't the only mechanism.
And that's just higher-end desktops that use discrete controllers rather than cheaper systems and laptops that will have to wait for USB 3.2 in Intel/AMD chipsets. So, I'd say "a year off" is fairly reasonable, certainly before there is any significant demand for 3.2 peripherals...
USB 3.2 isn't going to swamp out Thunderbolt v3 in the next 1-2 years for sure. In part, because there is far more natural high synerigies with Thunderbolt v3 in the mid-high end laptop space than in the mid-high end desktop space. So the volume of the former is much higher TBv3 will probably win out in volume. But that just means USB 3.2 will be smaller than TBv3 not necessarily too small at all.
USB 3.2 isn't a great TBv3 'killer'. It is even less of the 'killer' now that TBv3 is merged into USB4. It is though a better "previous USB versions" port though. The topology matches and if narrow view to USB 2.0-3.0 like capabilities (using socket to provision a Type-A connector to a Type-A peripheral ) it works efficiently.
But the point is that USB4 == Thunderbolt 3 for most practical purposes - USB-IF have said that it's going to be compatible. There are already systems of all types with Thunderbolt 3 in circulation who can buy your TB3 devices... and you know what they won't be able to use? USB 3.2x2.
You won't be able to use the USB 3.2x2 device on TBv3 in 3.2x2 mode, but the USB 3.2x2 peripherals will work with TBv3 systems in USB 3.2 gen2 (3.1 gen 2 ) mode. They can possibly work
faster on the USB3.2x2 systems deployed. Those 3.2x2 peripherals will work with a USB 2.0 host system too (just super duper slow).
Many of the Thunderbolt v3 peripherals won't work with USB 3.0/3.1 host systems (that lack Thunderbolt). The USB3.2x2 peripherals will. USB3.2x2 peripherals will work on USB 2, 3, and 4 systems.... which adds up to the billion of systems range. That's the USB inertia advantage that Thunderbolt doesn't have.
It looks like part of Intel's strategy has been to make their TB3 chipset the "go to" discrete USB-C controller, with TB3 capability as a bonus, which has helped build a (maybe) critical mass of TB3 devices. I expect they're partly betting on having a 'first-mover advantage' when competing USB4 controllers actually appear.
Yes since Intel merged USB into the Thunderbolt implementation it was far easier to give it up for "adoption" to the USB folks. If TBv3 had not been an inherent USB promoter (and more so a USB competitor) then it would have been much harder to get them to adopt it. If they can't get USB-IF to adopt it then there is no "first mover" advantage, because it is relatively dead in the water. ( Intel could try to form a rival open standard but getting folks to sign up with Intel and Apple have a war chest to beat the crap out of you if they don't get their way would be hard. )
However, it as only more so been the "go to" USB-C controller for those who wanted the Alt modes too. For the folks who wanted to completely eschew the Alt modes ( for costs , complexity ,etc issues) or wanted specialized controller ( USB3-SATA for external drives , USB3-DSP , etc. ) Intel really has
not been the "go to" vendor. In the latter, case Intel has mostly been a "one size fits all" when it comes to controllers. 1-2 ports and about the same size (and approximately the same power). For two port peripherals, the host and peripherals have often by using the same chip just set to different mode.
Not sure why - if a full-stack discrete USB 3.2 controller is going to need 4xPCIe, a DisplayPort stream and a hook-up to the power supply then the implementation and wiring is going to be much the same as a Thunderbolt 3 controller, so its just down to the cost of the chip.
Nope. Just look at the sonnert two port card above. It only has a 1x PCI-e v3 connection ( 8Gbps so not going to get the USB advertized 10Gbss ). [ Actually that is a major disconnect between Thunderbolt and USB. Thunderbolt tends to talk about the real data throughput of the TB network. USB historically always quoted theoretical speeds you'll never reach but doesn't include read overhead for data. ]
Quite likely several USB 3.2 controllers are going to be under provisioned ( just a x2 connection maybe a x4 when slapped into ancient PCI-e v2.0 system); the user is never going to see 20... just perhaps a little more than 10.
The DisplayPort stream is entirely options. The aftermarket add-in cards will highly likely come with zero Alt modes. So no DP hook up. Same with more than a few motherboards where it will probably be mostly used as "drop in" replacement for USB 3.1 gen 1 (2) controllers and perhaps bumped to x2 (from x1 ) PCi-e stream. The high end desktop motherboards are going to cater to GPU-less CPU packages so they don't want to deal with DisplayPort. They are highly likely to label that a huge punt.
The only context where looking at Alt modes for both controllers is probably mainly just in laptops. In that context though the USB controller in the PCH tends to have even more advantages then they do (as already paid for). If only one can be added to the laptop then yeah Thunderbolt has some upsides. (at least on the Intel systems. AMD not as much historical ground work done.)
At the moment, 3.2 controllers don't exist, the first few will probably be expensive and after that one assumes that Intel will adjust their prices to compete.
They do exist. Just not in volume. A "Related Reading" to the article you referenced (and I posted in another response). ASMedia was demoing a chip at Computex 2018 (almost a year ago). Vendors who wanted a sample to work with have had them since Summer-Fall 2018.
https://www.anandtech.com/show/12922/asmedia-demos-usb-32-gen-2x2-phy-usb-32-controller-due-in-2019
The time to market more so has to do when can get relatively complete engineering samples to the folks doing the building. There is very little to point to that didn't happen.
In part, new motherboards are waiting on new CPU/PCHs to trigger another major upgrade round in the motherboard market also. Those new boards are likely not 100% stuck on USB3.2x2 controller production flow. ( OS and driver updates too probably tie to one of OS drops if looking for stability there also. )
Meanwhile, Intel have also said that future CPUs will include TB3 on-chip... which will make it much easier and cheaper to implement TB3. The interesting question is - will Intel add USB 3.2x2 support at the same time? Not doing that could pretty much kill 3.2g2.
TBv3 isn't going to be completely on chip. The only "CPU" they will be on is one where the PCH is also looped into the package ( e.g., Core-m series. ). Most likely this is going to be somewhat like Intel implementation of Wi-Fi and Ethernet that are "on-chip" in the PCH. Technically not completely true. both the Wi-Fi and Ethernet chips also require a Intel PHYS chip to implement the port. Thunderbolt will probably fall into the same category where you'll need an incrementally cheaper TB PHY chip also. ( My guess is that they'll yank the USB3 part out and just feed that over the wire somehow when need it for "fall back" mode and move to a smaller process node for the PHYS chip ... both will make it smaller). There are placement himations for the TBv3 chips. It has to be just a couple of inches from the physical port ( part of its role is as a very low latency switch so don't the traces too long.) . Intel is unlikely going to saddle the CPU or the PCH packages with those sort of restrictions.
USB4 and eventual follow ons have at least the placement limitation problems if try to crank up the bandwidth on the network higher.
For the GPU-less CPU packages, the associated PCH probably will
not have TB weaved in the next versions at all. Or anything like USB4. 3.2x2 would be more likely because a relatively older standard than USB4 and easier to do. I don't think they gotten to 3.1 gen 2 so that would be a start. ( C422 usb 3.1 gen 1 C600 series just USB 3.0 ). They tend to trail about a 1-2 years back from the mainstream laptop/desktop ones. ( even more conservative). USB4 would be 2-4 years out.
