Intel Introduces Thunderbolt 5 Connectivity Standard | Hacker News Reader

Intel Introduces Thunderbolt 5 Connectivity Standard

intel.com

49 pointsrobin_reala67 comments

I know cable standards are better today (esp. Cell phone chargers) but I miss being able to know what a cable or port does by the shape of it.

Now I have a drawer of USB c style connectors with tape on them I write what the cable can do or not.

Same. It turns out physical port standardization was a red herring. If I can’t charge my laptop with a cable, it does me no good to have the same physical interface.

Maybe we can at least switch to a color standard like we had with the block inside USB 2 v 3 (where 3 was blue). Many manufacturers didn’t/don’t honor it but it’s nice when they do. Color could be added to solve this without having to retool cable factories. The only problem I guess is that USB C is such a mess and has so many possible combinations of features the colors may be hard to remember. Maybe there would be common ones?

It could be like resistors, where there are separate stripes for voltage, wattage, data transfer speed, DisplayPort version, thunderbolt support, etc.

So you’d know a cable with red-green-black-blue-yellow was good for up to 60w @ 20v, USB 3.2 transfer speeds, DisplayPort 1.4, no thunderbolt, no active electronics.

I would immediately install an app which tells me what the code means.

Jokes on you, USB-IF would change the meaning of colors annually, and prohibit cables from labeling what version of the code they used.

For charging a laptop, it's usually easy to tell by the thickness of the cable.

But it does do some good? You can use any power adapter and cable that are up to the task of charging a laptop. And some devices do get it right anyway, all the ports can be used for charging. And they can all be used to charge something else.

Hard disagree: I think it's awesome how all our recent 20V chargers double as perfectly fine (even if wildly overbuilt) chargers for simple one-cell devices like mice or Gopros or, well, anything you throw at them. It's awesome, so much better than having a zoo of chargers standing at the ready.

The only thing I miss from the micro-USB days are really lightweight cables (and plugs)

Do they? My laptop charger can't fast charge my phone. And well, its sorta dumb in that regard and is something like the normal 5V 2A max of pre PD charging for nearly everything that can't negotiate the 15V (or is it 20) the laptop wants and the little charger it comes with provides.

The aftermarket chargers are much better at supporting a wide variety of voltages, and even QC/etc nonstandard charge modes. But even then, I have a charge dock on my desk which has different colored ports because only one can do 100W PD, two can do QC charging, and some others are 30W limited, etc. Its a crazy mismatch and probably more than half of the time, people are just getting 5V charging.

Plenty of devices are perfectly fine with plain 5V charging. That's my entire point: instead of a zoo of chargers, you have that one charge dock on your desktop that does it all. I call that victory, not defeat.

The USB-A -> USB-C cable that charges my sennheiser headphones can't charge my iPad, but a longer USB-A -> USB-C cable can (using the same USB-A port on the same charger).

Also, the majority of USB chargers that came with my Apple devices can't charge my M2 laptop (even when it is suspended), assuming it's not a cable incompatibility issue.

Devices that consider themselves special enough to deserve a specific fast charge technology can't be expected to charge on just about any USB A, what's complicated about that?

Your Sennheiser will fill up just fine in a pinch on the iPad's charger or on the M1 charger. Will the iPad charge on the charger for the M1? Likely, unless Apple sells noncompliant PD chargers or if the iPad uses some oddball fastcharge alternative that isn't PD. PD is carefully specced as a pyramid of "bigger sources can serve all lower level sinks and not just a subset" (yes, cables need to encode their level)

As an end user with unlabeled cables, chargers and devices, that encoding may as well not exist.

I have 10 chargers, 10 cables and 10 devices. It could be that only one charger and one cable provides the highest level, so now I have to try 100 permutations to charge something.

Also, many cables get flaky after a month or so, so I might have to try the winning combination multiple times.

Finally, having the cables form a compatibility hierarchy is some strange theoretical property. We could have every device come with a cable that is mutually incompatible, but have one special cable that works with all devices in a vault at nist next to the standard kilogram, and then we would have a pyramid of compatibility.

For instance, I’ve encountered situations where I need two different USB-C cables if I want video out to HDMI to work on two different laptops.

Why not think about it as being able to charge other devices with your laptop cable?

I wonder if the EU is going to amend the USB-C charger requirement to address this.

I can charge my laptop, my PS5 controller, my headset and my Kindle with a single cable next to my couch. It does some good.

I don’t care that I can use any cable for charging, I care that if I have a charging cable I can use it with any of my devices.

I wish someone made a cheap cable tester, sort of like the chip tester pro but for cables.

These days I just toss out all the random USB-C cables I get and replace them known (expensive) cables.

I was also curious, so I web-searched it for you and came up with a couple of promising items:

A USB Voltmeter with E-marker and protocol testing: https://www.amazon.com/ChargerLAB-KM003C-Portable-Charging-V...

A very basic but maker-friendly signal/capability tester: https://www.tindie.com/products/petl/usb-c-cable-tester-c2c-...

That's actually one of the goals of thunderbolt. With thunderbolt you can be assured of certain capabilities whereas usb-c is increasingly the Wild West.

but how do you distinguish a Thunderbolt-compliant cable from a very basic USB-C cable?

By a logo. Same as for any other type of cable.

And when i see the lightening bolt how do i know its 20, 40, or 80 gbps? Count the jaggies?

Thunderbolt certification is mandatory, not optional as it is with USB, to be permitted to use the Thunderbolt logo on your connectors.

The certified cables are listed here: https://www.thunderbolttechnology.net/products (select Cables, then Thunderbolt 4 or similar)

>but how do you distinguish a Thunderbolt-compliant cable from a very basic USB-C cable?

The picture on the end of the cable - it will contain a picture of a lightning bolt, and a number indicating which version of thunderbolt the cable supports:

https://www.thunderbolttechnology.net/sites/default/files/CB...

I don't think it is as easy as you imply...

Belkin's Thunderbolt 3 cable supports 60W power delivery https://www.belkin.com/thunderbolt-3-cable-usb-c-to-usb-c-3....

Cables 2 Go's Thunderbolt 3 cable supports 100W https://www.cablestogo.com/usb-and-pc/usb-c-cables-adapters-...

StarTech says optical Thunderbolt 3 cables don't support power delivery at all https://www.startech.com/en-us/faq/thunderbolt-3-power-deliv...

Thunderbolt 3 has to give up to 15 watts of power delivery on copper cables, anything beyond that is out of scope for the standard.

More bandwidth is always good, but this also has universal cabling support for 240W charging [1].

My M1 MacBook Pro can use a single Thunderbolt 4 cable to an external monitor supporting 100w now. But if the battery is charging, occasionally the external screen blanks until the battery is fully charged or connected to external power.

1. https://www.tomshardware.com/news/thunderbolt-5-debuts-120-g...

Ok, how many Lbs of reference manuals do I have to read to practically use it?

In case you were wondering about the "bandwidth boost" thing alluded to in the press release:

> As Intel has demoed with a prototype in the past, the new Bandwidth Boost mode uses three of the 40 Gbps pipelines in transmit mode to deliver up to 120 Gbps of transmit speed, albeit at the cost of stepping down to a 40 Gbps receive rate.

Hopefully this means we can have an Apple 27" 5K or 32" 6K display with ProMotion some day.

3 USB-C's walk into a bar. They all order different drinks of different sizes. The bartender forgets who ordered what. Its a mess.

Now make the license costs bearable so it becomes ubiquitous.

Intel doesn't charge licensing fees. You might have to pay a lab to certify your device as Thunderbolt 5 compliant, but that's a one time cost and nothing that Intel charges for.

Didn't they already removed it years ago? At least for Thunderbolt 3:https://www.pcmag.com/news/to-spread-thunderbolt-intel-will-...

They removed the license costs on Thunderbolt nearly 6 years ago...

Well, in the PC space, Thunderbolt is intentionally market segregated by motherboard vendors who basically only put it on top tier motherboards for both Intel and AMD.

But even for phones it isn't easy to implement as you have to basically throw PCI lanes at a thunderbolt controller or roll your own thunderbolt IP internal to the processor.

Otherwise, Intel removed all fees and shifted to a royalty free license with Thunderbolt 3 in 2017.

In what way is this a standard?

Where can I read the specification?

TB3 and 4 have a max data transfer rate of 23 Gbps iirc when using an SSD.

I wonder what's the actual max data transfer speed of TB5.

The release materials mention 2x more bandwidth for external SSDs and that lines of with the mentioned PCIe gen 3 -> PCIe gen 4 upgrade also noted. On Thunderbolt 3/4 that was 32 Gb/s prior to encoding overhead and the not-a-perfect-world factor so, extending that to twice the raw bandwidth, I'd guess somewhere around ~55 Gb/s for NVMe SSDs.

As a consumer, TB4/USB4 is fast enough. What I want is more pci-e lanes, without getting into the HEDT processors.

Would be nice if systems offered more bifurcation. An NVMe SSD on pcie 4 could easily run off 2 lanes and be fine. 1 lane for 5.

You lose a bit of top end transfer that shows up in the benchmarks, but it's near impossible to tell.

Hopefully this means Promotion support for monitors for MacBooks. Being limited to 60hz isn't ideal.

So what version of USB is this? /s

No joking, it's actually USB4 V2....

gen1 or gen2?

USB 3.1 Gen 2 V2

I just love that USB-IF just goes all-in on naming chaos. They must have great parties.

Yeah, they don’t need a list of invitees either. Just ask some USB standards trivia at the door.

Was it. It is now USB 3 Gen 3

Ultra speed gen 128x256 v3.2

It's incredibly sad and frustrating that I can't tell if this is sarcasm or not. God, I hate the USB standards folks.

https://www.usb.org/document-library/usb4r-specification-v20

I wish it was a joke.

The version is meaningless to a consumer. Labelling and documentation guidance are to use "USB 80Gbps / 240W".

Apple uses this?

And today we have Apple announcements.

What iPhone use case would this much bandwidth be useful for?

Quickly transferring high resolution ProRes videos?

Importing 4K ProRes video for editing on a computer?

I feel that's a bit overkill. From practical perspective, how often would you be downloading 4K ProRes from your phone vs sharing a lower res version? Unless Apple are setting up for something related to Vision Pro I don't know if it's useful to include this on their phones. A regular USB-C connection would suffice for occasional backup.

iPhones are used for professional videography, so for those people, all the time.

Just announced that the 15 Pro can record directly to external storage

Monitoring 8K output from a Pro model.

possibly using the iPhone as a camera on a macbook I'd think. Also locally backing up an iOS device to a desktop with speed.

You can do the camera thing today over wifi though

40Gbit/sec should be quite enough for a _camera_.

iPhone camera has absolutely no need for 80 or 120Gbit for webcam usage. It works just fine even with standard USB2 bandwidth.

Backups maybe, but I suspect you're going to be limited by the speed of the storage on both sides. Existing thunderbolt 3 already supports 40 gigabit/s. Not much storage, even apples storage, supports 5000MB/s read/write except for very short periods of time.

Rumors are the iPhone 15 Pro will have Thunderbolt, but it's not going to be 5. Probably Thunderbolt 4, although 3 would be plenty if there was some reason (lower power?) to use that over 4.

Reply on news.ycombinator.com