USB4 Version 2.0 announced with up to double the standard's original speed

[Redaktion]

The USB Promoter Group has unveiled specifications for USB4 Version 2.0, the next-gen standard for type-C ports that might increase data-transfer speeds by 2x, even through some existing cables. It is also rated as compatible with its PCIe and DisplayPort counterparts, and may improve performance for some USB 3.2 devices.

https://www.notebookcheck.net/USB4-Version-2-0-announced-with-up-to-double-the-standard-s-original-speed.647198.0.html

[vertigo]

More speed is almost always good, even though I doubt most people come close to saturating the current speeds. What I'd really like to see is the problem of RF interference by USB 3. It's a known issue for affecting WiFi, and I just had to unplug my USB 3 SSD enclosure so I could get more than 2-3 meters from my computer before my headset started cutting out. And that was without even doing any active reading from or writing to the drive.

[_MT_]

Seriously? What kind of a genius came up with a designation like "USB4 Version 2.0". As if it's not a big enough mess already.

Just to be clear, 80 Gb/s will require the new active cables. Yes, it is backward compatible with older USB4 cables, but of course at appropriate bandwidth. The improvement to tunnelled USB 3.2 is also interesting as it would be silly to expect older USB 3.2 devices to suddenly get faster when their controllers were never designed for it. I struggle to see what that is about. Uplink of a hub? New devices with new controllers being able to use tunnelled USB 3.2 without that bandwidth cap?

[NikoB]

Both commentators simply point blank do not understand the essence of the change, and the funny thing is after I have been writing about this for quite a long time. The essence of the change is USB4 compatibility with Display Port 2.0, which requires the same 80Gb/s. This is all done solely in the form of preparation for the massive transition to 8k resolution in monitors. And for nothing more. At the same time, this transition, to the utter shame of the entire high-tech industry, should have happened back in 2014, when the first industrial 8k panels for monitors appeared. But only in 2019, DP2.0 was introduced as a standard for transmitting 8k signal under ONE cable. And again, to the shame of the entire high-tech industry, the introduction of DP2.0 was delayed, in fact, for 4 years! Given that video chips for discrete cards from NVidia/AMD cards, starting from the average level already in 2010-2011, had enough bandwidth of VRAM to serve 8k monitors. But alas, the PC's RAM could not then serve the requests of 8k monitors. It is still too slow for 8k resolution, which is why the hi-tech industry is dragging this out year after year due to the fact that they cannot make available to mass PC (even with i3/r3 level processors) memory with a bandwidth of 100 Gb/s+ in a conventional dual channel mode (and since many, due to technical illiteracy, are still sitting in single-channel mode, it is generally necessary that 100Gb/s+ be the norm even in this mode).

[NikoB]

The main stupidity of the situation with the introduction of new standards is that due to the monstrous illiteracy and low education even of people in the "developed" Western countries, the high-tech industry cannot switch to optical connections and cables for home use. Which is holding back progress. Because of the primitive thinking of the townsfolk, unable to follow the net connectors. They should be taught to wash their hands with soap and running water, and not dialed into the sink, otherwise even in the USA these wildness is still the norm in everyday life ...

No active cables are needed - the signal cable has long been (must) be optical and not copper. This changes everything dramatically. There is no problem today to transmit even 500 Gb / s over an optical core over distances up to 100 m.

Those. the transition to optical links at home is hampered by the monstrous level of the primitive mentality of the inhabitants, who are simply unable to comply with the rules for using optical cables for computer equipment, if we talk about the vast majority of the inhabitants. There are no technical obstacles for a long time already. Chips-transmitters and receivers for such speeds (100Gb Ethernet for example) have long been really worth a penny in bulk. Those. all this stupid fuss with different cable quality and constant problems could have been solved long ago with a single optical link in which multimode transmission takes place at speeds above 100 Gb / s.

For example, many projector owners (like me) understand the problems of long-distance routing of HDMI cables, in which there is a complete mess with quality and complete fraud by manufacturers.

Remember the same eGPUs - there is a hard limit on 32Gb / s pci-e 3.0 x4. Guess why? Copper cables.. And very short only 0.6-1m maximum. And people need an eGPU on pci-e 4.0-5.0 x16, and at a distance of up to 50-100m (for example, to take a super-noisy video card to the back room in an apartment or house and sit in complete silence during the game!). All this will provide us with an optical signal. But they pull with him ...