Thunderbolt 4 Arrives In 2020, But USB Will Remain the King of PC Ports (cnet.com) 161
Intel announced Thunderbolt 4 this week at CES, saying it will arrive in PCs later this year with Intel's new Tiger Lake processor. But, as CNET reports, "the all-purpose port won't be any faster at transferring data than the 4-year-old Thunderbolt 3." From the report: The chipmaker promised it would be four times faster than today's USB, then clarified it was talking about the USB 3.1 version at 10 gigabits per second. Thunderbolt 3, though, already can transfer data at 40Gbps. Still, you can expect other changes. "It standardizes PC platform requirements and adds the latest Thunderbolt innovations," Intel spokeswoman Sarah Kane said in a statement, adding that Intel plans to share more about Thunderbolt 4 later.
Thunderbolt, embraced first by Apple in 2011 and later by some Windows PC makers, has proved popular in high-end computing situations demanding a multipurpose connector. A single Thunderbolt port can link to external monitors, network adapters, storage systems and more. But Intel's years-long ambition to make Thunderbolt mainstream hasn't succeeded. Instead, USB remains the workhorse port.
Thunderbolt, embraced first by Apple in 2011 and later by some Windows PC makers, has proved popular in high-end computing situations demanding a multipurpose connector. A single Thunderbolt port can link to external monitors, network adapters, storage systems and more. But Intel's years-long ambition to make Thunderbolt mainstream hasn't succeeded. Instead, USB remains the workhorse port.
USB soup (Score:3)
Earth to Intel: we don't want more USB branding soup, especially if Apple is involved. Just ditch the stupid branding wankery already and concentrate on supporting USB-C. Earth to Intel: anything that sounds even a bit like "lightning" is going to drive techies and design wins away. And any connector/protocol not supported by AMD is dead in the water.
Re:USB soup (Score:5, Informative)
Thunderbolt 3 and up DO use a USB-C connector, but can transport Thunderbolt, DisplayPort, and USB signaling all over the same cable simultaneously. You can plug USB-C devices into the port just fine. And as the summary mentions, you can plug monitors into it as well. And with USB-C Power Delivery, it can be used to charge devices as well.
Re:USB soup (Score:4, Insightful)
That's great that it does that, but.... everyone still thinks of Thunderbolt as being "That Apple connector". Teaming up with Apple was a huge mistake as Apple is well known in the public as having it's own cables (I still hear basic users when asking for charging cables as asking "Do you need a normal charger or an Apple charger?").
Sure, techies may know the difference, but average Joe doesn't. They know that USB is in everything, and Thunderbolt as an Apple only thing (even if that hasn't been the case for years)
Re: (Score:2, Insightful)
Teaming up with Apple was a huge mistake
First off, Apple and Intel cooperatively developed the standard that eventually became Thunderbolt.
Second, without a major PC OEM like Apple going all-in on Thunderbolt, (like they did with USB), Thunderbolt would have languished (like USB did until Apple came along and saved it from oblivion), and likely died before it ever go to the (much) more useful Thunderbolt 3 version.
But don't let the facts (and history) interfere with your already-settled world view.
Re:USB soup (Score:5, Interesting)
But don't let the facts (and history) interfere with your already-settled world view.
I don't have that problem, but you clearly do.
But let me humour you for a minute. If what you are claiming is correct, that it would take a major PC (average Joe doesn't see Apple as a PC, but whatever) OEM like Apple going all-in on Thunderbolt, then after 8 years.... it should be THE standard now..... right?...
But hey, lets play a game here. I want you to pretend you are average Joe. You've never heard of "Thunderbolt". So, like anyone else, you search google of Thunderbolt (try it now). All you'll see is a Wiki page full of tech jargon, an Intel site with more tech jargon, maybe the notice that Thunderbolt 4 is coming and... stuff about Thunderbolt and Apple. Nothing about it from anyone else. Just Apple. Same if you look on Amazon, it has a bunch of ads (on my end) for Apple computers. Average Joe still gets to see from other sites that Thunderbolt = Apple.
So who isn't facing the facts still?
Re: (Score:2)
But don't let the facts (and history) interfere with your already-settled world view.
I don't have that problem, but you clearly do.
But let me humour you for a minute. If what you are claiming is correct, that it would take a major PC (average Joe doesn't see Apple as a PC, but whatever) OEM like Apple going all-in on Thunderbolt, then after 8 years.... it should be THE standard now..... right?
It is for anybody who needs blistering transfer speeds. Now go hate Microsoft for a while.
Re: (Score:3)
Average Joe still gets to see from other sites that Thunderbolt = Apple.
Intel was actually the biggest impediment to Thunderbolt's early uptake. They had to Approve projects, and extracted a License Fee.
Fortunately, they dropped both of those requirements a couple of years ago, and Thunderbolt devices and computers are finally beginning to become a thing.
But you should have tried a different DDG search. Typing in "Laptops with Thunderbolt 3" pops up (among Dell and LG laptop ads), this handy list:
https://www.ultrabookreview.co... [ultrabookreview.com]
I didn't count but it looks like about 70-80 lapt
Re: (Score:3)
The irony of you calling someone else a Luddite...
The problem with Thunderbolt is that it required integration of video, which at the time was difficult and costly. PC manufacturers did not see value in it and the expense grew out of the required integration.
Thunderbolt was first and foremost an Apple concept, not an Intel one. Intel agreed to support it because it wanted Apple's business. PC manufacturers didn't want it because they needed to support external video solutions which were not easily routab
Re: (Score:2)
IMHO, the problem has largely been that as USB has gotten better over time, the number of devices that can either take advantage of or benefit from the new standard have not picked up on it.
I remember when USB 2 came out and months later PCs with USB 2 ports were still shipping with a mix of USB 1 and 2 ports, not all 2 ports. The same thing happened with USB 3.0 -- it was like what the fuck, why even bother putting USB 2 ports on at all when you could have made them all 3.0? It wasn't even like the plu
Re:USB soup (Score:4, Informative)
Teaming up with Apple was a huge mistake
First off, Apple and Intel cooperatively developed the standard that eventually became Thunderbolt.
They didn't develop shit.
It's external PCIe, which already existed. They just slapped a wrapper around it.
Re:USB soup (Score:5, Informative)
Teaming up with Apple was a huge mistake
First off, Apple and Intel cooperatively developed the standard that eventually became Thunderbolt.
They didn't develop shit.
It's external PCIe, which already existed. They just slapped a wrapper around it.
You're right that Thunderbolt is, in part at least, a serialized version of PCIe. This did already exist, albeit in Parallel form, as the CardBus standard (which Apple also supported back in the day).
But you gotta admit, as far as a peripheral bus goes, Thunderbolt is a helluva lot more convenient for the user than CardBus, and a helluva lot more capable.
Then we have to add DisplayPort and USB, and now we have something unique.
So, it does seem that Intel and Apple did develop something new afterall, eh?
Re: (Score:2)
You're right that Thunderbolt is, in part at least, a serialized version of PCIe. This did already exist, albeit in Parallel form, as the CardBus standard (which Apple also supported back in the day).
PCIe is a serial protocol itself. It can have more than one line (e.g. 2, 4, 8, 16) but each line is serial and the lines are not synchronized between each other at the bit level. Thunderbolt has two lines as PCIe x2. Claiming that Thunderbolt is serialized PCIe is misleading. CardBus is not a PCIe bus but PCI bus (which is in fact a parallel interface). Maybe you wanted to say ExpressCard which can work as USB2 or PCIe. But that is a serial interface.
Re: (Score:2)
You're right that Thunderbolt is, in part at least, a serialized version of PCIe. This did already exist, albeit in Parallel form, as the CardBus standard (which Apple also supported back in the day).
PCIe is a serial protocol itself. It can have more than one line (e.g. 2, 4, 8, 16) but each line is serial and the lines are not synchronized between each other at the bit level. Thunderbolt has two lines as PCIe x2. Claiming that Thunderbolt is serialized PCIe is misleading. CardBus is not a PCIe bus but PCI bus (which is in fact a parallel interface). Maybe you wanted to say ExpressCard which can work as USB2 or PCIe. But that is a serial interface.
You are right, I stand corrected. I was mis-remembering. I did mean ExpressCard (which I think is what Apple actually supported, sorry!)
But, a minor correction to your correction: Thunderbolt can either use 2 or 4 PCIe lanes. Apple uses 4 for its laptops. Some Windows TB3 laptops (of which there are many) use 2 lanes, and some use 4:
https://www.ultrabookreview.co... [ultrabookreview.com]
Re: (Score:2)
Re: (Score:2)
Re: USB soup (Score:2)
Thatâ(TM)s how Iâ(TM)ve heard it described.
Thanks for the research!
Re: (Score:2)
Re: (Score:2)
Convenient for PC makers, where you can NEVER add a "Thunderbolt" card to an old PC or motherboard.
Have you considered why this is? Can you explain a case someone would want a Thunderbolt add-in card?
Add in cards for things like USB, parallel, serial, Ethernet, and so on were for ports slower than the slot it plugged into, often by a wide margin. ThunderBolt is a PCIe implementation for hotplugging external devices, anything needing this could just as easily use ExpressCard, eSATA, or other such high speed interfaces.
ThunderBolt ports on an add-in card is meaningless if the computer you are adding it t
Re: (Score:2)
I'm guessing that Apple gambled on the use of ThunderBolt for adding peripherals to their cylinder Mac Pro and then quickly learned this was a bad idea, because people wanted more than 40Gbps for many things.
I agree with you that the 2013 Mac Pro (cylinder) was designed with wide and rapid adoption of Thunderbolt (2) in mind. As we all know, that sadly didn't happen.
I figured that when that computer came out, along with multiple Thunderbolt equipped audio and video interfaces and such (of which a few did actually happen), there would quickly be external PCIe card-cages, much like the eGPU cages you see now (but less specialized), and that those cages would allow those who wanted to plug all manner of bog-standa
Re: USB soup (Score:2)
Convenient for PC makers, where you can NEVER add a "Thunderbolt" card to an old PC or motherboard.
Well, maybe not to an old PC motherboard (although Iâ(TM)m not entirely sure that is true); but the new Mac Pro certainly demonstrates you can design a new PCIe slot-based PC with plenty of Thunderbolt 3/USB-C ports.
Now what?
Re: (Score:2)
LOL was does "serialized version of PCIe" mean? You really are unqualified to talk on this subject.
Taking a clue from your equally stupid "Cardbus" reference, it appears you mean external. PCIe already had external solutions prior to Thunderbolt, but you wouldn't know that.
USB was not added to Thunderbolt originally, but DisplayPort was, and that was the cause of its cost and initial lack of success (but also the reason for Apple's interest). You don't know this either, judging by your other ignorant pos
Re: USB soup (Score:2)
Apple did NOT develop Thunderbolt, it created demand for it. Intel did the development, then failed to achieve broad adoption with customers other than Apple.
Sorry, youâ(TM)re wrong.
LightPeak was co-developed by Intel and Apple, and Appleâ(TM)s design and specification work was also responsible for the copper-wire based interface that we know as Thinderbolt.
Did Intel do the heavy-lifting on the hardware side? Undoubtedly! But Appleâ(TM)s ginger was definitely in the pie on the specification side from the very beginning.
And although Iâ(TM)ll admit I havenâ(TM)t paid much attention to the evolution of the PCI and PCIe standards, I actually
Re: (Score:2)
Debatable. The same year as the iMac ditched legacy ports we also guy Windows 98 with native USB support.
Anyway, it would have been better if they had implemented Thunderbolt as part of USB from the start.
Re: (Score:2)
Apple contributed nothing to USB, nothing to its technical development and nothing to its success. USB's success was secure, it was merely delayed due to Microsoft's lateness to market. Every PC maker was fully supporting USB prior to Apple's late adoption of it.
Your "already-settled world view" is notably absent historical facts.
Re: (Score:2)
Apple contributed nothing to USB, nothing to its technical development and nothing to its success. USB's success was secure, it was merely delayed due to Microsoft's lateness to market. Every PC maker was fully supporting USB prior to Apple's late adoption of it.
Your "already-settled world view" is notably absent historical facts.
While you are technically correct (see what I did there?) that Apple contributed nothing to the initial development of USB (the "committee" consisted of Compaq, NEC, Microsoft and Intel, IIRC), no one can say with a straight face that Apple (and more specifically, the iMac) contributed nothing to its success. That's patently absurd.
The fact that some Wintel motherboards had a useless USB riser on-board from 1996-on, was nothing more than a cruel joke on the consumer, who was forced to purchase hardware that
Re:USB soup (Score:5, Insightful)
There are technical differences between what a thunderbolt connection does vs what a USB connection does, and these differences is why there are Thunderbolt connections. I am glad however that we have USB. Thunderbolt would be almost impossible to integrate into most devices that are built on SoC silicon. Cameras, speakers, etc... would require huge amounts of circuitry and CPU power to handle thunderbolt connections because they require 4 lanes of PCI right into the CPU. SoC don't have that.
As for Apple, no technology company has a better track record of introducing a standard as Apple does. They were the first to implement USB too, lest you forget
Re:USB soup (Score:4, Informative)
They were the first to implement USB too, lest you forget :)
I must have forgetten that Apple made the Toshiba Satellite 220CS in 1997.
https://www.manualslib.com/manual/328836/Toshiba-Satellite-220cs.html [manualslib.com]
I know they are the first to go ONLY USB in September 1998 (they ditched other style connectors), but they weren't the first to implement USB. Nice try to re-write history though....
Re: (Score:2)
I know they are the first to go ONLY USB in September 1998 (they ditched other style connectors), but they weren't the first to implement USB. Nice try to re-write history though....
I remember motherboards from that era with USB ports on them, That came about the same time as the original iMac, and probably before the iMac as well. These ports were useless because the dominant OS's at the time (Windows NT/95/98) did not have native PnP USB support.
Re: (Score:2)
These ports were useless because the dominant OS's at the time (Windows NT/95/98) did not have native PnP USB support.
Actually, USB support was added to Windows 95 as an update in August 1997. I still have an OEM CD from the time labelled "Windows 95 with USB support". Windows 98 had support from the beginning. In fact, the famous crash that happened on stage with Bill Gates when Windows 98 was first demonstrated was prompted by the insertion of a USB scanner.
You are right that Windows NT 4.0 did not have USB support. That was introduced with Windows 2000.
Re: (Score:2)
In fact, the famous crash that happened on stage with Bill Gates when Windows 98 was first demonstrated was prompted by the insertion of a USB scanner.
And you want to call that "Support"?
Re: (Score:2)
Yes, bugs can appear in prerelease software. That is hardly news.
Re: (Score:2)
Actually, USB support was added to Windows 95 as an update in August 1997. I still have an OEM CD from the time labelled "Windows 95 with USB support".
It was earlier then that, back in April or May of '97. The version you're talking about is OSR2 or 2.1 with USB support, which was the OEMs version for system builders all in a shiny system disc. There were other ways you could get it too, if you had internet access you could spend hours downloading it...but that was kinda silly, especially with the lack of USB devices. 'Cause those were the days when internet access costs were tied to time(10-20h blocks for $20-80). You could also get it from MS's old BB
Re:USB soup (Score:4, Informative)
They were the first to implement USB too, lest you forget :)
I must have forgetten that Apple made the Toshiba Satellite 220CS in 1997.
https://www.manualslib.com/manual/328836/Toshiba-Satellite-220cs.html [manualslib.com]
I know they are the first to go ONLY USB in September 1998 (they ditched other style connectors), but they weren't the first to implement USB. Nice try to re-write history though....
When I was building white-box PCs around 1996-97, all the motherboards came with a port-plate that had these weird little rectangular connectors on them. The documentation said they were "Universal Serial Bus" connectors; but since there wasn't anything that seemed to use them, I generally just left the port-plate in the box with the rest of the motherboard goodies I didn't need.
So yes, many PCs had the connector (good on the OEM Intel Sales Reps in winning those motherboard "sockets"); but until the iMac came out in 1998, there were like one or two obscure devices that supported the USB standard.
Oh, and it didn't help that Windows didn't have USB support of any kind until Windows 98, and didn't have usable USB support until at least Windows 98 SE, which didn't debut until nearly a year after the iMac.
And talk about embarassing: Linux didn't officially gain USB support until kernel 2.4 in 2001 (!!!) (although there was work on USB support starting in 1999, in kernel 2.2).
https://en.wikipedia.org/wiki/... [wikipedia.org]
So sure, there were plenty of computers running around before the iMac with these then-useless USB connectors on them. Big Deal. But the iMac was undeniably the first computer that actually made use of USB.
And it undeniably changed the world.
In 1998, you could hardly find a printer with a USB connector. In 1999, you could hardly find a printer without one (but most also had a Serial and/or Parallel connector, too). In fact, I had an Epson inkjet printer from that time that had all three ports! And by 2000, it was getting hard to find a printer with anything but a USB connector.
Re: (Score:2)
It was a good laugh back then because NetBSD had USB support before Linux.
Re: (Score:2)
So now it's clear, you were a PC assembler at that time. LOL
At that time, I was a developer with a Fortune 100 PC manufacturer doing firmware work. I was personal friends with the core USB development teams and the teams that validated USB. I have intimate knowledge of USB development history, although it was not my personal work. I also knew those responsible for marketing strategy and the trail of tears that resulted from MS's long delays adding USB to Windows.
You are wrong on literally everything bec
Re: (Score:3, Informative)
Wikipedia is hardly a reliable source, child.
I had USB in 1996 on my Pentium Pro system and every system on that before Apple ever released the iMac. It was already widely-used in the PC world around 1997, despite needing to install an update and drivers due to no native PnP until 98SE, which is what REALLY popularized it, because the iMac wasn't shit sales wise compared to the PC market in those days.
Don't rely upon a constantly-edited and heavily warred-upon site for information.
Re: (Score:2)
Re: (Score:2)
Wikipedia is hardly a reliable source, child.
I had USB in 1996 on my Pentium Pro system and every system on that before Apple ever released the iMac. It was already widely-used in the PC world around 1997, despite needing to install an update and drivers due to no native PnP until 98SE, which is what REALLY popularized it, because the iMac wasn't shit sales wise compared to the PC market in those days.
Don't rely upon a constantly-edited and heavily warred-upon site for information.
Widely used?
For what?!?
Re: USB soup (Score:4, Informative)
I bought one of the very first consumer boards with USB available.
I bought an I-Will socket 7 board, that I originally put a K6-166 in, eventually moving it up to a K6-233.
It was one of the most reliable, long-lived boards I ever bought, which I hated it for because I bought other I-Will boards in the future as a result of that, all of which were garbage.
The board in question had a riser socket to accommodate the USB riser you had to order at a future date since the connector standard wasn't finalized when the board was released even though the protocol/electrical standards were.
I never ordered the riser, because despite actively looking for USB devices to plug into it, everything was still PS/2, serial, and parallel for the most part, maybe SCSI while I had that board. I found some expensive mice and scanners I could special order, but they cost a lot more than the rest of what was out there, I couldn't justify it "just to use the new port".
The first USB device I bought was a scanner years later. I fully give credit to Apple for making that port happen. Even though I had several systems with USB before the first-gen iMac came out, I couldn't find anything to plug into those ports.
The entire industry was doing what I now refer to as "Logiteching". I use that term because Logitech has been very conspicuous in pretending Bluetooth isn't important and pushing their own standard - which is basically adapter-level Bluetooth anyways, in an effort to lock people in.
Thanks to USB standards being a real thing now, I actively refuse to buy anything that has a proprietary data connector - and increasingly I'm looking very hard at proprietary power connectors as well. Sorry Apple, despite making the USB standard happen your stupid-assed lightning and it's 30 pin predecessor are proprietary and remove your products from my consideration.
Re: (Score:2, Interesting)
It was integrated into EVERY modern PC chipset when the iMac was introduced. Some systems elected to omit the connectors and not advertise the feature due to lack of Windows support.
Every one of those systems could support mouse and keyboard, though dual PS/2-USB devices existed because USB was designed to make that easy, not by Apple but by the people who actually developed USB.
USB IO, storage, hubs, null-modems and networking devices all existed prior to the iMac, but without Windows support they were no
Re: (Score:3)
I'm curious as to what peripherals you used. It wasn't until Apple's core demographic with their disposable income needed USB that we started seeing USB mice, printers, etc.
Hell, the Apple puck mouse probably got more mice developed than Windows getting USB support.
How long until PC OEMs stopped shipping PS/2 devices and ports?
Re: (Score:2)
Yeah tons of boards has USB back then. How many peripherals were you plugging in? Maybe a mouse but that was a rarity. It wasn't until Apple made a device with USB only did companies start making keyboards/mice/webcams etc etc.
Re: (Score:2)
I guess english isn't your native language because you seem to not know that "first widely used" and "first to implement" are two wildly different thing.
Either that or you're lying to push an agenda. Which do you prefer?
Re: (Score:2)
Written by people with an agenda that did not work in the industry, as is common with Apple-related content on Wikipedia.
Re: (Score:3)
Apple was NOT the first to "implement USB"! First, Apple's USB was implemented by Intel who was providing the same chipset to other manufacturers and they were doing the testing!
Many manufacturers were already shipping USB when the iMac was introduced. USB was of little use because Microsoft was behind with support, that is all.
It is amazing how Apple marketing lies were so easily adopted as fact. Apple contributed absolutely nothing to USB, it merely exploited it after the work was done by everyone else
Re: (Score:2)
Thunderbolt would be almost impossible to integrate into most devices that are built on SoC silicon.
I wouldn't be so hasty with that pronouncement.
I'd posit that we're only about 1 or 2 generations of iPad Pros before that USB-C connector grows Thunderbolt capabilities.
Re:USB soup (Score:4, Insightful)
Apple contributes to the confusion (and high-pricery of Apple's commodity products) by shipping a USB-C cable with every Apple notebook WHICH DOES NOT FULLY SUPPORT USB-C.
If you use the wire in the box (which is used for charging) to plug in your external hard drive then that drive will not run at full speed.
And if you use that cable to plug your Retina iMac into your MacBook Pro then the Pro will not charge at full speed.
Not simple enough.
Re: (Score:2)
I seem to recall USB cables not able to support USB 2.0 speeds, this was rare at the time, and unheard of now (because people threw out those shitty cables).
I recall having to double check that ribbon cables for IDE, ATA, and SCSI had to be checked for the speed that they supported. Oh, and if they supported cable select for setting master/slave.
I recall parallel port cables being of various quality to support different speeds. And there being at least three different connectors for the port.
Then there is
USB4 == ThunderBolt 3 (Re:USB soup) (Score:2)
Sure, techies may know the difference, but average Joe doesn't. They know that USB is in everything, and Thunderbolt as an Apple only thing (even if that hasn't been the case for years)
That will change with USB4. People that still think of ThunderBolt 3 as "that Apple thing" will think nothing of using the same devices on their non-Apple computers if it's got "USB4" slapped on it. They can still live in their ignorance while gaining the advantages of ThunderBolt speeds and features.
(I still hear basic users when asking for charging cables as asking "Do you need a normal charger or an Apple charger?").
When it comes to phones there's still plenty using the old micro-B and vendor specific additions (well, "violations" might be more precise) to their implementations of USB that it seems to me that people acce
Re: (Score:3)
Thunderbolt 3 and up DO use a USB-C connector, but can transport Thunderbolt, DisplayPort, and USB signaling all over the same cable simultaneously...
Yah, we don't want that. I thought I was clear. It sounds like "proprietary extensions to USB-C", and that is what we don't want. Repeat: we don't want that, please just quit beating that dead horse and get back to what the world does want. If the formal USB-C isn't up to the needs of HPC, then fix it. Don't rebrand it, that strategy is doomed.
Re:USB soup (Score:4, Insightful)
Yah, we don't want that.
No, you dont want that. The rest of us think it's pretty cool (if annoyingly named).
Re: (Score:2)
I think what he really means is: We want all those features as standard in USB-C - in every downstream USB C port, please.
The hardware support is so fragmented right now. Myself, I have been looking for a controller chip for a USB hub I'm designing, and I haven't yet found one that supports everything that is already promised in the base USB C spec, without any extensions like Thunderbolt.
Re: (Score:2)
I don't want that. Fuck Intel shitting the USB bed over and over and over.
Re: (Score:2)
Thanks for that DMA port right into your system. So much for your security.
And just how many random PCI (and ISA) peripheral cards also have/had DMA support?
So, do you actually have a point?
Re: (Score:3)
And just how many random PCI (and ISA) peripheral cards also have/had DMA support?
The big difference is that it's far easier to casually plug in a doctored USB device into an (exposed) USB port, than it is to unscrew the PC's case, insert a PCI card, and close it again.
And it's far easier to trick an unsuspecting user into inserting an untrusted USB stick than to trick him to open up his PC case...
Re: (Score:2)
The scenario the GP claiming is that someone comes across an unlabeled PCIe card and will likely decide to plug it into their computer which is obviously bullshit.
Its obviously bullshit because 99 out of 100 people would never EVER open up their computer to install ANYTHING, even stuff they purchased themselves!
Contrast that with coming across an unlabeled USB thumb.... thats getting slotted by 99 out of 100 people... the exact opposite!
Re: (Score:2)
ISA? ISA!?!? WTF are you talking about? Who gives a shit about ISA?
100% of "random" PCI cards have DMA, idiot.
You clearly do not understand the problem being communicated to you, nor does it seem you are capable of it. You don't walk up to a PC and plug in a PCI or ISA card while it is running.
Re: USB soup (Score:2)
You can't "fix" it. USB is a general purpose communication protocol. DisplayPort is specifically for largely unidirectionally shoving large chunks of data 60 (or more) times a second. Thunderbolt is basically plugging directly into the PCIe busâ"like writing to disk without a filesystem, managing byte offsets and references manually. There is a valid place for all three.
There's also no one-size-fits-all solution.
Comment removed (Score:5, Insightful)
Re: (Score:2)
Except nobody cares because Intel's stupid licensing made thunderbolt a non starter and now USB is more than "good enough" for the masses. Its the same stupidity that killed Firewire, when launched nobody could argue that USB was even close to what Firewire was doing but nasty licensing costs made the OEMs go "fuck that just add more USB ports"
Intel did screw up with their Licensing and "Approval" regimes for Thunderbolt. But they finally came to their senses about 3 or 4 years ago, and dropped both requirements. That may be why there are dozens of non-Apple TB3/USB-C equipped laptops out there, from pretty much every major laptop OEM, and the trend is steadily accelerating since 2017:
https://www.ultrabookreview.co... [ultrabookreview.com]
https://www.windowscentral.com... [windowscentral.com]
And the licensing for Firewire was a whopping 25 cents per unit. If that is the straw that broke t
Re: (Score:2)
Re:USB soup (Score:4, Informative)
Here you go with this red herring again...
25 cents is a dealbreaker for high volume PCs, but you wouldn't know that coming from a background of assembling PCs in your bedroom.
Re: (Score:2)
When people look back at thunderbolt I'm sure it will be looked at like Firewire, people will say "Yeah it was a good idea, too bad they fucked it up" because even my Apple customers don't use thunderbolt,its USB everything because USB devices are as common as air and cheap as dirt and can be found everywhere. The ship is done sailed Intel, sorry you missed the boat.
You do realize that USB4 is just USB 3.x with ThunderBolt 3 rolled in, don't you? People will be using ThunderBolt soon whether they realize it or not.
People started using USB because it was a singular connector that provided a data channel and power. As time went on people kept asking for more power and higher data rates. To the point that it became FireWire re-implemented badly. They tried to fix this with USB 3.x and USB-C, but that added to the confusion rather than remove it. They keep trying to m
Re: (Score:2)
Fine provided that it will be easy to plug in usb2 adapters and use almost new pro level hardware like a great many audio recording devices in most pro and semi pro studios. IF all of a sudden every piece of pro level hardware that has usb2 connectivity is obsoleted there will be hell to pay in the music industry.
You mean like how Apple did that with their newest computers with the T2 "security" chip?
https://www.youtube.com/watch?v=tpVNkglXc84&t=994s [youtube.com]
They will still be buying such hardware.....
Re: (Score:3)
Earth to Intel: we don't want more USB branding soup, especially if Apple is involved. Just ditch the stupid branding wankery already and concentrate on supporting USB-C. Earth to Intel: anything that sounds even a bit like "lightning" is going to drive techies and design wins away. And any connector/protocol not supported by AMD is dead in the water.
Apple doesn't have anything to do with the USB naming conventions.
And yes, I do agree that USB naming is a hopeless mess. It's just not Apple's mess, sorry.
Re: (Score:2)
Apple was involved, at least because of firewire, and also lightning. So I disagree with you, as does history. Apple's involvement tainted firewire, lightning and finally thunderbolt, just by association. Simple historical fact. Intel needs to get some of that learning from history thing.
Hey dumbass:
Firewire: Jointly developed by Apple and Sony. Originally designed as a replacement for the SCSI bus. Technically, it was quite nice for its time. But time marches on...
Lightning: 100% designed by Apple. Originally designed as a (far superior) alternative to MicroUSB's shitty, shitty, shitty connectors. Never intended as an industry-standard; so it's not clear how Apple "tainted" Lightning.
Thunderbolt: Jointly developed by Intel and Apple (originally as an Optical-only standard, LightPeak). Init
Re: (Score:3)
All of this is also wrong. What a surprise.
Firewire was originally developed by Apple as a interprocessor comm link. Apple eventually abandoned it, then Sony picked up because it needed a digital interface for its new DV format. Sony productized Firewire and developed the original version that we know it today. Technically, Apple owns the Firewire name, so Sony's product was IEEE-1394. 1394 products FAR preceeded Firewire ones, for obvious reasons.
Lightning started as the iPod connector which encompass
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
So obviously Intel should just shut up about thunderbolt and call it USB 4. Just let thunderbolt die as the pointless branding exercise it is.
Re: (Score:2)
So obviously Intel should just shut up about thunderbolt and call it USB 4. Just let thunderbolt die as the pointless branding exercise it is.
We could do that, but then you would not be reminded of the fact that you are using dirty Apple tainted tech every time you use a USB 4 device.
Re: (Score:2)
If you worked with equipment that needs more than the USB-C standard then it isn't "just" double. 70 mph is "just double" 35 mph but try selling someone a car that can't cruise smoothly above 35 mph on that basis. An obvious example of this is display data. The reason high end laptops tend to include at least one thunderbolt port (even if it is USB-C format) is because you can't run high resolution high refresh monitors over
Re: (Score:2)
The Thunderbolt cables cost more because Intel never gave up on screwing customers for a few extra bucks.
What in holy hell does Intel have to do with the zillions of Thunderbolt cables you can buy on Amazon?
Longer TB3 cables that support high-speed have active electronics in them which makes them (kind of unnecessarily, IMHO) pricey; but that really isn't Intel's fault; blame it on the laws of physics.
You're an idiot.
Re: (Score:2)
You might be missing the point of TB -- it's not just about raw speed, it's the PCIE lanes it exposes to an external device, that USB does not.
eGPU's are a 'thing' solely because of thunderbolt. The fact that you can take a lightweight laptop with a modest CPU, and connect it to whatever desktop video card you wish, and get pretty fucking good performance + the ability to upgrade is not something to dismiss out of hand.
Yes laptops come with discrete on-board video but it's always so marginal in comparison
will intel have pci-e 4.0 and more lanes to make (Score:2)
will intel have pci-e 4.0 and more lanes to make use of this?? on desktops the same DMI + 16 (tied to video card)
and with apple shared with the flash storage on the DMI bus even on the high end imac pro and mac pro.
Ahhh, Firewire take 2 (Score:5, Interesting)
So I designed our product around USB.
Which sucked, we were doing video and Firewire was much better suited for what we wanted to do. But USB did the job, and I haven't looked at Firewire since (still got the 20 year old book on my shelf though, Just In Case (tm).
Comment removed (Score:5, Insightful)
Re: (Score:2)
The other colossal fuck to with FireWire was that they couldn't agree on a voltage.
Re: (Score:2)
The one good thing it used to have was no frame drops. Back in the days you were recording to DV tapes it had a much higher consistency while it seems every USB connection would glitch or hang for just long enough to lose frames. Of course this all became a moot point when we started recording to memory cards. And it was a one way competition to begin with, USB was for everything like keyboard and mouse. Firewire? You'd have at most one port. But it's not necessarily a bad idea to have one "superport" thoug
Re: (Score:2)
Re: (Score:2)
Re: (Score:3)
Which is why where money is no issue, like the US Military, they went with Firewire for the JSF's video system.
Re: (Score:2)
The cost of FireWire came from the fact that all devices were full peers. USB had a clear host/device separation and the USBDEV integration was severely limited so that you could afford putting it in a disposable mouse. Meanwhile, the dead hand of Gordon Moore has been pushing up the amount of integration you could put in a disposable device and now we are at a point where what was once a full peer FireWire controller is now disposable integration.
Re: (Score:2)
Back in 2000 or so I had to decide between Firewire and USB. Firewire was the clear winner. Then our suits talked to the Apple suits, our suits told me how big a bite the Apple tax was going to take.
So I designed our product around USB.
Which sucked, we were doing video and Firewire was much better suited for what we wanted to do. But USB did the job, and I haven't looked at Firewire since (still got the 20 year old book on my shelf though, Just In Case (tm).
So your suits balked at a whole 25 cents royalty per unit on Firewire? You must not be a very persuasive advocate for your Designs. How much did your device sell for? $2.00?
https://en.wikipedia.org/wiki/... [wikipedia.org]
And oh, BTW, your suits were wrong. That wasn't an "Apple Tax". There were 10 Corporations that held the Patent Pool. Not just Apple.
Quit lying.
Re: (Score:2)
> So your suits balked at a whole 25 cents royalty per unit on Firewire?
Of course not - they knew that every computer would come with USB which was included with the chipset and very few would come with Firewire because mobo vendors don't spend extra money on parts most people don't want.
I had some nerdy mobos and even a few 1394 PCI cards, but most people didn't and his "suits" didn't want to hand over the sector to another company that chose to embrace the reality that by 2000 most computer owners were
Re: (Score:2)
Of course not - they knew that every computer would come with USB which was included with the chipset and very few would come with Firewire because mobo vendors don't spend extra money on parts most people don't want.
Oh, so now you change your story!
FIrst you said it was the "Apple Tax",
But when I laid that argument bare, now you switch to "Nothing had the connectors".
So which is it?
I think we're done here.
Re: (Score:2)
He's not changed his story - he's a different poster.
Re: (Score:2)
Your suits made the winning decision over engineers. That's rare, but it happens because engineers only fetishise technical superiority while attempting to ignore the world around them.
Misleading article. (Score:5, Informative)
The two standards (Thunderbolt and USB) are coming together to become USB v4. This is great because it means that AMD will be able to utilize the new standard in their upcoming chipsets. It will also make it easier for device and cable manufacturers to start making products. The USB standard is controlled by the USB Implementers Forum [wikipedia.org] - a much safer entity to work with then Intel.
But the article does not make it clear if Thunderbolt 4 is to become a new standard separate from USB v4 or if they are really just talking about USB v4. Considering the speed has not changed, one would assume that Intel is actually talking about USB v4. Some USB features will probably be adopted by Thunderbolt to validate the version update - such as USB PD (power delivery). It makes since when they are sharing the same cables.
Re: (Score:2)
I think this is the most important detail. When thunderbolt was new, I inquired about including a connector on an embedded system, not using an intel CPU or chipset. Not even an answer from them, so it was a void and useless technology to me. Becoming an open standard, usable on non-Intel platforms (Arm, RiscV etc as well as AMD), as USB can, is what will begin to make Thunderbolt at all relevant. And I have only had Intel based laptops or desktops for work or home for several years, but I do not have a si
Re: (Score:2)
I was under the impression that USB 4 will implement Thunderbolt 3 as standard. As to which "4" we are talking about now it's confusing as hell.
Far more interesting is that USB 4 makes supporting PD mandatory. No more pot luck as to whether your USB device supports PD, iQ, or some proprietary garbage.
Re: (Score:2)
Re: (Score:2)
USB v4 as the merging of current thunderbolt and current USB...
Technically, that's great, but I can't even imagine the branding nightmare it will be. Watch out, I think the versionning obfuscation division of the USB consortium is hiring.
Re: (Score:2)
USB v4 as the merging of current thunderbolt and current USB...
Technically, that's great, but I can't even imagine the branding nightmare it will be. Watch out, I think the versionning obfuscation division of the USB consortium is hiring.
I believe the opposite will be true. This should bring less confusion, not more.
One thing is that people have come to understand that new USB is backward compatible with old USB. Having something with "USB4" on it in big print and then in the small print put in that it is compatible with USB 3.x, ThunderBolt 3, and USB 2.0 if that also applies. People will soon find this a non-issue.
What is also happening is people are expecting faster and faster interfaces on their computers to connect hard drives and s
Thunderbolt 3+ (Score:2)
They should just name it Thunderbolt 3+, and then Thunderbolt3++ until they actually have some actual speed improvements.
I'm not at all surprised by the misleading Intel marketing, though. They're also comparing their 25W CPUs to their 15W CPUs, and touting higher performance, comparing to several generations old CPUs and again touting higher performance, comparing a 8C/16T CPU with RTX 2080 to a 4C/8T CPU with RTX 2060 and who knew, theirs was a winner!
Re: (Score:2)
They should just name it Thunderbolt 3+, and then Thunderbolt3++ until they actually have some actual speed improvements.
Or Thunderbolt 3.1, then 3.2, etc.
I'm confused on what ThunderBolt 4 is supposed to bring. It still uses the USB-C port, right? If there was a new port then I can understand the need for a new name but that doesn't sound like what's happening.
Best I can tell they are merely changing the licensing to make it easier to add to new chips not made by Intel, and are revealing a new implementation that's supposed to make it smaller and cheaper for people that keep buying silicon from Intel. I'm not sure that ca
Re: (Score:2)
Why thunderbolt is not ubiquitous (Score:2)
Re: (Score:2)
Me 2. Would love to commute in that.
Re: (Score:2)
Me 2. Would love to commute in that.
Actually, a Harrier is much more practical as a commuter-vehicle.
You can land it in your employer's parking lot (nevermind the melted asphalt, though!)