China Develops HDMI Alternative: 192 Gbps Speeds and 480 W Power Delivery

[Crackong]

This is known physics, and it's very easy to test with a bench power supply and a load. 16ga would be the smallest size for 480W and you typically want your cable to be able to safely carry more. Using smaller cables then the load requires is one reason why cables heat up and melt. The cables in this picture do not have the combined size to even meet the requirement.
Electricity doesn't except nationalities.

Wire Sizing Chart for 12V/24V/48V DC Systems - DIY Solar Resources

Information on wire sizing and a universal AWG/mm2 wire sizing chart to help in designing a 12V, 24V, or 48V DC renewable energy system.

www.altestore.com

True.

Put safety in mind it might even need 14 AWG wires.
A typical cable with 14 AWG wires looks like this.

It is just way too thick, stiff and bulky as a 'mobile' cable.

 

[TheLostSwede]

I know it's ampere that could cause harm, it's the same issue with the 12VPWR thingy on RTX cards that causing melted cables. I mean 48V might be the usual range for PoE but not PC peripherals, if they wanted to make this 'General Purpose'. And they wanted to shove 480W through this thing?

View attachment 394113

Yeah, I looked at the Type-B cable and dismissed it as insane straight away.

Have you tested it?

As others have pointed out, there's no need to test it, at least not based on the currently available information.
Also, xinese companies are known for cutting corners (no pun intended), especially when it comes to using correct wire gauge, which is why there often are recalls, as some "smart guy" at the factory figured out they could save 1 cent per device by using cheaper wires.

 

[Tek-Check]

Also, xinese companies are known for cutting corners (no pun intended), especially when it comes to using correct wire gauge, which is why there often are recalls, as some "smart guy" at the factory figured out they could save 1 cent per device by using cheaper wires.

I know, but this happens, not only in China. Intel is also capable of producing CPUs that died around the world. We know the story... Even if we always buy certified and tested gear from known brands, it's not a guarantee.

There is no reason to put advanced negative spin on something we don't know how it's going to look like in its final version. We simply do not know at this point. Besides, as top 50 local and global consumer electronics companies are behind it, plus regulatory bodies, in one way or the other they will need to get it right. We will judge it when it arrives.

 

[TheLostSwede]

I know, but this happens, not only in China. Intel is also capable of producing CPUs that died around the world. We know the story... Even if we always buy certified and tested gear from known brands, it's not a guarantee.

There is no reason to put advanced negative spin on something we don't know how it's going to look like in its final version. We simply do not know at this point. Besides, as top 50 local and global consumer electronics companies are behind it, plus regulatory bodies, in one way or the other they will need to get it right. We will judge it when it arrives.

Well, then the companies behind the new standard shouldn't show ridiculous mockups of cables that are not going to deliver the claimed specs. They could've done this in so many better ways. On top of that, 40 V at 10 A is lethal, a broken shielding on one of these cables will kill people.

This just feels like one of those things that were thrown together to show that we are better, without much thought having gone into it.

Also, I'm no proponent of HDMI, but if we're going to make new standards, they should be global and royalty free. No more of this crap that only works with one type of devices and that you can't use with other brands, that time is over.

 

[igormp]

On top of that, 40 V at 10 A is lethal, a broken shielding on one of these cables will kill people.

Maybe if you try to lick the wire. 48V DC is considered below the danger threshold by pretty much all standards out there, no matter the current.
USB-PD already allows 48V @ 5A, and I haven't heard about any safety hazards with it so far.


Just as a quick physics recap to folks, you need a certain voltage level to break through the skin's resistance, current is pretty much irrelevant at this point. The threshold for DC is also way higher than for AC.
After this threshold has been "broken", you could die with less than 1A. In fact, 0.1A is already enough to kill if the current goes through your heart.
I have written about this years ago:

Igor Moura - What kills, voltage or current?

And also a nice video with a more practical explanation:

 

[TheLostSwede]

Maybe if you try to lick the wire. 48V DC is considered below the danger threshold by pretty much all standards out there, no matter the current.
USB-PD already allows 48V @ 5A, and I haven't heard about any safety hazards with it so far.


Just as a quick physics recap to folks, you need a certain voltage level to break through the skin's resistance, current is pretty much irrelevant at this point. The threshold for DC is also way higher than for AC.
After this threshold has been "broken", you could die with less than 1A. In fact, 0.1A is already enough to kill if the current goes through your heart.
I have written about this years ago:

Igor Moura - What kills, voltage or current?

And also a nice video with a more practical explanation:

Technically 7 mA across your heart can kill you...
48 Volts is really on the borderline where you can die if shocked, much more so than lower Voltages.
You also assume dry skin, keep in mind that in Asia, many people still don't have AC in their homes and it's quite humid in many parts of Asia, which mean people's skin won't be dry.

And yes, technically speaking USB PD, once we get 48 V chargers, could be nearly as bad, but there's a reason the USB-IF didn't go with higher current than 5 A.

 

[igormp]

You also assume dry skin, keep in mind that in Asia, many people still don't have AC in their homes and it's quite humid in many parts of Asia, which mean people's skin won't be dry.

You're overblowing the skin resistance due to humidity, it won't be soaking wet because of that. I do speak from experience given that I do live in a place that has 90%+ humidity constantly (82% as of this exact moment). The skin's resistance does decrease, but it won't suddenly make 48V DC life threatening unless you're sitting in a tub filled with water.

And yes, technically speaking USB PD, once we get 48 V chargers, could be nearly as bad, but there's a reason the USB-IF didn't go with higher current than 5 A.

We already have plenty of 48V PoE stuff around, and I've never heard of any health hazards due to it.
Higher current than 5A is not due to health hazards, but rather fires and wire gauging. As you well said, low currents are enough to kill you, so even 0.5A would impose such risks.
The 5A or higher thing has nothing to do with shocking dangers.

48V, no matter if at 0.5 or 100A, is not lethal from a shock risk perspective, period.

 

[R-T-B]

Maybe this will force HDMI Forum into breaking from making stupid decisions like forbidding AMD from releasing Linux support for HDMI 2.1...

To be fair, they didn't forbid it outright. They just refuse to allow an OSS driver to have it. The nvidia linux binary driver supports hdmi 2.1 fine.

That said, it's still a major asshole move.

no matter the current.

lolno. Current is a big part of the entire "safety" equation.

Of course its safe if you assume you will never be shocked but we just can't realistically do that.

 

[igormp]

To be fair, they didn't forbid it outright. They just refuse to allow an OSS driver to have it. The nvidia linux binary driver supports hdmi 2.1 fine.

Issue isn't the OSS driver, the Nvidia open source driver also supports hdmi 2.1.
The catch is that such support from Nvidia is built into the firmware, unlike AMD.

lolno. Current is a big part of the entire "safety" equation.

Of course its safe if you assume you will never be shocked but we just can't realistically do that.

I did overstate it, but my point is that there's no difference if it's 5A or 10A, such low voltage won't impose shocking risks.
If we were talking about voltages over 60V AC, discussing 10A would be moot given that even at 1A you could be facing actual lethal shocks.

 

[nageme]

No more of this crap that only works with one type of devices and that you can't use with other brands, that time is over.

With the potentially looming tariff wars, and China working toward more self-sufficiency in tech (for example, CPUs using LoongArch),
it's not impossible that we're on the verge of tech standards becoming more fragmented internationally.

 

[R-T-B]

The catch is that such support from Nvidia is built into the firmware, unlike AMD.

Well yeah that would solve it because the firmware is closed source. It's still a OSS vs closed binary blob concern.

 

[igormp]

Well yeah that would solve it because the firmware is closed source. It's still a OSS vs closed binary blob concern.

Yeah, but same applies to AMD. I'm not sure why they're insisting in not placing such support within their exitisting binary blob.

 

[R-T-B]

Yeah, but same applies to AMD. I'm not sure why they're insisting in not placing such support within their exitisting binary blob.

They probably will in UDNA. Rewriting your firmware doesn't happen overnight.

 

[igormp]

They probably will in UDNA. Rewriting your firmware doesn't happen overnight.

For sure. And hindsight is 20/20, but I wonder why such a basic feature was not in the firmware to begin with.
But as you said, we shall see how it goes with UDNA.

 

[Scrizz]

With the potentially looming tariff wars, and China working toward more self-sufficiency in tech (for example, CPUs using LoongArch),
it's not impossible that we're on the verge of tech standards becoming more fragmented internationally.

There's a very relevant xkcd for that.

Standards

xkcd.com

 

[yfn_ratchet]

There's a very relevant xkcd for that.

In other news, XKCD just violently sneezed.

You claim-jumped my joke, man...

 

[Scrizz]

You claim-jumped my joke, man...

lol I missed that, but I said it in response to nageme.

 

[Caring1]

Well, then the companies behind the new standard shouldn't show ridiculous mockups of cables that are not going to deliver the claimed specs. They could've done this in so many better ways. On top of that, 40 V at 10 A is lethal, a broken shielding on one of these cables will kill people.

Can kill, not will kill.
A momentary shock generally won't kill a person (if the Amperage is low enough) it might hurt and throw them though.
I've had a High voltage transmission power line come down and hit me in the forehead, instantly knocking me down and out during a rain storm.

 

[ncrs]

They probably will in UDNA. Rewriting your firmware doesn't happen overnight.

Pushing this support into the firmware isn't an optimal solution either. Firmware can't be modified by anyone but AMD. The native kernel driver would have been able to be modified by anyone. Bugs would be fixed faster, possibly even new features added. The code could be shared among all users of HDMI 2.1 just like many Linux kernel features are shared between for example Intel and AMD GPUs.
AMD's Linux support while relatively good isn't perfect as I'm personally being affected by a 1-year old bug. It's not a big issue since it's purely visual, but the fix isn't complicated either.
I do not understand the reasoning for HDMI Forum's decision. Is the possibly of having parts of the 2.1 specification revealed by open-source driver code that much of an issue? If someone wanted to they could either pirate the specs or reverse-engineer it.

 

[R-T-B]

Pushing this support into the firmware isn't an optimal solution either.

Oh don't get me wrong, I'm of this opinion as well.

AMD's Linux support while relatively good isn't perfect as I'm personally being affected by a 1-year old bug.

You and me both, actually. Good thing it's a largely cosmetic bug.

 

[Tek-Check]

On top of that, 40 V at 10 A is lethal, a broken shielding on one of these cables will kill people.

I am sure they are aware of this.

I'm no proponent of HDMI, but if we're going to make new standards, they should be global and royalty free. No more of this crap that only works with one type of devices and that you can't use with other brands, that time is over.

I hear you. The consortium of leading TV and other companies behind HDMI love collecting royalties. It's them we need to thank for this new rebellious GPMI.

Besides, USB-C version of GPMI will be globally compatible with all devices registered with USBIF.

 

[trsttte]

~50V to break the skin barrier is a ballpark, it can vary wildly depending on conditions, skin type, if a tiny strand of copper pierces your skin, whatever. 48VDC and 10A is simply not wise, and point taken 48VDC 5A from USB-C PD isn't either.

Safety margin is a basic idea to understand, if the literature says ~50V is a tipical limit, you shouldn't go above ~33V, and if possible even like ~25V since the consequence can be death (1.5 and 2 respectively)

Can kill, not will kill.
A momentary shock generally won't kill a person (if the Amperage is low enough) it might hurt and throw them though.
I've had a High voltage transmission power line come down and hit me in the forehead, instantly knocking me down and out during a rain storm.

A big difference there is AC vs DC - with a power line there's 50 or 60 (depending where you are) times per second where the voltage and current is 0, not as in you can get lucky and be hit during that time, but as in you have a chance to let go of the cable while the voltage and current are droping, that doesn't happen with DC.

 

[nageme]

There's a very relevant xkcd for that.
https://xkcd.com/927/

Though it's different here.
I don't think the impetus for this new standard is to improve prior ones or to unify.
But rather, it's to become independent and not have to coordinate or mind foreign players.

 

[Scrizz]

Though it's different here.
I don't think the impetus for this new standard is to improve prior ones or to unify.
But rather, it's to become independent and not have to coordinate or mind foreign players.

Regardless of the impetus, it's another competing standard.

 

[Tek-Check]

Regardless of the impetus, it's another competing standard.

And what does USB-C part of it represent?