News for hackers. Stuff that matters.
Re:With USB 4, Intel is finally making good on the promise o... by jameskilton on Wednesday March 06, 2019 @11:22AM (#19320056)
Do note that the correct name for this iteration is "USB4" with no space[1].

Because naming things is hard.

[1] https://usb.org/sites/default/files/2019-02/USB_PG_USB4_DevU...

Re:With USB 4, Intel is finally making good on the promise o... by dylan-m on Wednesday March 06, 2019 @12:26PM (#19320718)
I'd be willing to bet the next version is USB4 2.0.
Re:With USB 4, Intel is finally making good on the promise o... by wlesieutre on Wednesday March 06, 2019 @12:46PM (#19320902)
If they stick with the current system, next will be USB4.1 Gen 2, while USB4 gets retroactively renamed to USB4.1 Gen 1.

But that's assuming they don't find a creative new way to name USB4 versions poorly.

Re:With USB 4, Intel is finally making good on the promise o... by ihuman on Wednesday March 06, 2019 @11:24AM (#19320071)
How come USB4 is one word, when USB 2.0, USB 3.0, and so on are 2? Is there an actual answer at all?
Re:With USB 4, Intel is finally making good on the promise o... by out_of_protocol on Wednesday March 06, 2019 @12:21PM (#19320688)
Latest version name is USB 3.2 Gen 2×2

You see, lines up perfectly

Re:With USB 4, Intel is finally making good on the promise o... by Aardwolf on Wednesday March 06, 2019 @12:46PM (#19320891)
"4" means bad luck in some countries, so maybe they should name it "USB5" from "3 + 2"
Re:With USB 4, Intel is finally making good on the promise o... by ConceptJunkie on Wednesday March 06, 2019 @11:34AM (#19320168)
Arbitrary and capricious marketing.
Re:With USB 4, Intel is finally making good on the promise o... by magduf on Wednesday March 06, 2019 @11:36AM (#19320183)
It's kinda like the Pentium4, where Intel switched from Roman numbers ("Pentium II", "Pentium III") to an Arabic numberal with no space.

Hopefully USB4 won't be the energy-guzzling disaster that Pentium4 was.

Re:With USB 4, Intel is finally making good on the promise o... by pbz on Wednesday March 06, 2019 @11:28AM (#19320114)
Naming is not (that) hard. What's usually hard is ignoring the ego that tells us to leave a mark or that we know better than others; i.e. putting your self before the whole.
Re:With USB 4, Intel is finally making good on the promise o... by basch on Wednesday March 06, 2019 @11:34AM (#19320164)
when can we start calling it the C4 port?
Re:With USB 4, Intel is finally making good on the promise o... by linarism on Wednesday March 06, 2019 @12:06PM (#19320544)
This is definitely one of the enabling technologies for Apple to use their own ARM-based CPUs on MacBooks
Re:With USB 4, Intel is finally making good on the promise o... by SlowRobotAhead on Wednesday March 06, 2019 @01:00PM (#19321008)
Royalty USB4 sure seems like the time that iPhone would adopt a USB-C connector like their laptops.
Re:With USB 4, Intel is finally making good on the promise o... by sitkack on Wednesday March 06, 2019 @11:12AM (#19319977)
Intel implements firewire immediately after killing firewire.
Re:With USB 4, Intel is finally making good on the promise o... by berbec on Wednesday March 06, 2019 @11:38AM (#19320194)
All I want is a Ryzen laptop with access to an eGPU
Re:With USB 4, Intel is finally making good on the promise o... by Crosseye_Jack on Wednesday March 06, 2019 @11:58AM (#19320435)
It _MIGHT_ be possible today... Kind of... With an ugly hack.

If said laptop has a m.2 port that exposes PCI-e lanes you might be able to use a m.2 to PCI-e adaptor. Dump the Thunderbolt binaries from a Intel UEFI, init a Thunderbolt PCI-e card on an intel system (to load the firmware onto that Thunderbolt Card) and then get Thunderbolt at least under Linux.

Wendel from Level1Techs was able to get Thunderbolt working on a Desktop Threadripper[0].

Obv it would be an ugly (both in looks, as you will have things hanging off the laptop and software as you will have to be injecting code into your UEFI). And if you have a M.2 port with PCIe then might as well just hook the GPU directly to the M.2 <-> PCIe adaptor[1]). Either way you would be limited to 4 PCIe lanes.

And with either method, you still might not get it to work. You might want to check out the eGPU Forums[2].

[0] https://www.youtube.com/watch?v=uOlQbP63lDQ

[1] https://www.youtube.com/watch?v=GDyL2tPyXFA

[2] https://egpu.io/forums/

Re:With USB 4, Intel is finally making good on the promise o... by dogma1138 on Wednesday March 06, 2019 @01:46PM (#19321418)
if you are going to use an m.2 to PCIe adapter might as well dump the middle man and just plug in a PCIe GPU straight into the x4 slot since that would work just fine.

Have a PCIe x4 cable going from the laptop to a box instead of a Thunderbolt cable the only thing you might "lose" is hotswap but technically even that is mainly an electrical problem if you don't use the PCIe power pins hotswap can work just fine.

Re:With USB 4, Intel is finally making good on the promise o... by Crosseye_Jack on Wednesday March 06, 2019 @03:23PM (#19322393)
> if you are going to use an m.2 to PCIe adapter might as well dump the middle man and just plug in a PCIe GPU straight into the x4 slot since that would work just fine.

I did say that and even linked a video of someone trying to do just that (The part about using the M.2 port is about 5 mins into the video).

> And if you have a M.2 port with PCIe then might as well just hook the GPU directly to the M.2 <-> PCIe adaptor[1]). Either way you would be limited to 4 PCIe lanes.

It does depend on what bios settings you are running and if you bios allows you to change those settings to get the card seen by the laptop. Linus (from LTT not Linux) wasn't able to get it to work, but then again it is Linus and his projects often don't go to plan.

Re:With USB 4, Intel is finally making good on the promise o... by berbec on Thursday March 07, 2019 @09:53AM (#19328464)
I'm aware of hacks involving m.2 slots. I find them interesting, but not helpful for a machine i use every day.

eGPU means my laptop is a mullet: Business in the front, party in the back. I want to be able to take it with me to presentations, server rooms, client offices and not have RGBBS dripping out of my keyboard. Then, when I get home, I plug it in and have a gaming rig. TB3/USB4 does this while other solutions are lacking.

Re:With USB 4, Intel is finally making good on the promise o... by ksec on Wednesday March 06, 2019 @01:33PM (#19321305)
We already have dozen of other similar submissions.

We have one that has 270 comments here [1].

[1] https://news.ycombinator.com/item?id=19304857

Re:With USB 4, Intel is finally making good on the promise o... by rasz on Wednesday March 06, 2019 @10:49PM (#19325382)
>40 Gbps

40 or "40"?, because currently: "dual-port '40 Gbps' Thunderbolt controller has 32 Gbps of PCIe bandwidth total, and each port supports 22 Gbps"

Re:With USB 4, Intel is finally making good on the promise o... by examancer on Wednesday March 06, 2019 @11:46AM (#19320279)
This is good for everyone. Hopefully the adoption is swift and we can pretend the myriad Thunderbolt/USB-C/USB 3.0 issues, differences, and confusion never happened.