Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Intel Looks to Broaden Thunderbolt 3 Adoption by Integrating Into Future CPUs, Eliminating Royalties
- Thread starter MacRumors
- Start date
- Sort by reaction score
A 5K display or two can easily chew up 40GiB/s, depending on the refresh rate.
It should, but it probably doesn't. -Some might work and some definitely won't, so -more than ever- try before you buy. What was that deal with counterfeit Lightning cables and fires? (Avoid cheap knock-off cables or you might melt your new $5000 vMP.) Anyway, add all this up and you get why people are so pissed about Apple yanking USB-3 and those tiny, but very powerful Thunderbolt 2 ports. It was obvious what worked with what. Like the ethernet port and the Mag-Safe port.
God help us all.
Bonus Question: Who is the patron saint of dongles?
Really I don't see the use for fast processors at all for anyone who isn't doing heavy work or intense gaming since the current low-end CPUs already do most tasks well. Hobbyist gamers don't use Macs, and work applications oughtta be parallelized anyway, so I'd be excited about Ryzen if I did more research. I've heard they require special motherboards. Intel currently charges a ton for Xeon, so it would be good to have some competition there.would be epic to see a Zen based CPU in a mac. They are really much better for creatives, with a focus on high core count over high Ghz (after all apple never let you overclock so what is the point of an overclockable intel cpu)
It could bring down the price of the top end iMac with Ryzen chips being much cheaper for the same performance + you get then benefit of ECC memorywithout going for the mad range Xeon prices.
Also for a mac pro at the moment the new AMD epic looks much better suited than intel's offering (we don't know about prices yet through)
BTW, GHz doesn't mean anything by itself. AMD in the past has had gimmicky high clock speeds on CPUs that actually suck. You can really only measure performance using some benchmark. Looks like the Ryzen CPUs have a better price/performance than comparable Intel, based on the benchmarks I've seen.
Last edited:
Since we are talking about a licensing fee, quoted as $/unit there certainly is no such thing as economies of scale for the licensee. For the licenser there certainly is but that will certainly do nothing with unit costs for the one obtaining that license.
Also, people here are thinking that if there is a license fee of 0,25$/unit, then the final price is only affected by $0,25$ per sold unit does not understand how things work.
To add a Thunderbolt port to your device, you will need to
a) pay the license fee
b) buy the "Alpine Ridge" chip that is only available from one single manufacturer
c) incorporate that chip into your existing device, ie re-model the entire motherboard
d) obtain the (currently non-disclosed ) protocol specifications of thunderbolt, then understand them and incorporate them into your software
All these things add to complexity and to cost far beyond what ever the licensing fee is. Cost and complexity is something most manufacturers are avoiding and thus we will not see a broad adoption of TB3.
The problems in b-d will go away once intel integrates the controller into a large number of their CPUs. How long will that take? 1-2 years?
[doublepost=1495814627][/doublepost]And just a reality check for you all. Intel has listed all current TB3 constructions in existence:
99 computer models or motherboards
56 devices, mostly disk cases (with a lot of overlap, counting each different storage size as a unique device) and one display
23 cables
That is the entire TB3 universe today.
Last edited:
Incorrect. Apple's role has been far smaller than you make it out to be. Intel already had a feasible and fully working product which they demoed to the world as Lightpeak. The only thing they didn't have was copyrights and a connector. The only thing Apple did was help Intel with that after Intel asking them for help. So no concept here but a fully working product that Intel developed entirely on their own. Let's not forget that Intel has far more experience with these kind of things than Apple does.
Which is what they are doing so your point being?
That would be you because Intel has stated this multiple times in the past. The current announcement however seems to be far more specific than previous ones (previously it was "we want to", now it is "we are going to").
No that would be far worse than this news. Intel now entirely dropped the royalties so implementers don't pay anything. Free is better than paying for it like you want them to do.
It's not because Thunderbolt is already adopted in the multimedia world and for those having high demands when it comes to their direct attached storage solutions. This is not a technology specific for the average consumer but one for the high end (professional) user. It is far better than FibreChannel, Firewire and all the other ones mostly because it in fact is cheaper.
Nope, the opposite, Thunderbolt is very cheap compared to many of the alternatives which is why an awful lot of people have adopted it instead of any of the alternatives. However, that doesn't apply to your average consumer who doesn't even benefit from USB 3.1 Gen 2.
The biggest advantage of Thunderbolt is that it in fact makes computers cheaper because an OEM can now make 1 device for a far bigger group. Previously they had to design and manufacturer 2 systems: 1 for consumers, 1 for the high end users. This will be accelerated when Intel integrates it into their CPUs. It makes designing new machines simpler and thus cheaper. The bigger audience also means that they can drive down the price even more.
USB-C alone does absolutely nothing as it is only a cable and a connector. It's all about the protocol that is carried over that physical connection. In this case that would be Thunderbolt 3 which is PCIe, DisplayPort, Ethernet and USB into 1 package. If you were to use just USB than you'd lose a lot of that.
Plugging in a device is not going to change the notebook. It is still the same machine and thus portable. The entire setup itself isn't and that is a subtle but very important detail. When you connect a display to it, the setup isn't portable either. So the conclusion here would be that this is only useable at a desk (at home, the office).
Then you'd not factor in all the other costs there are. It is maintaining 1 machine (plus all of your data and apps) vs 2 machines (where goes what, where is what data?) so there is also this thing called complexity.
You are dead wrong because you do not understand what USB-C is. As stated earlier it is no more than the physical hardware (cable and connector) and in itself meaningless. What is important here is the protocol being used over that physical connection. By using Thunderbolt 3 you can have all the advantages of both Thunderbolt and USB. If you were to only use USB than you'd only have what USB offers you. Right now that would mean that you cannot use any 5K display nor dual 4k displays or a 4k display and still have USB3.0 speeds. The biggest advantage however is that Thunderbolt 3 ports can do anything so you no longer have to think about whether or not your computer is going to support it. When you use USB you have to go read the specs each and every time to determine whether or not it is going to work with your machine.
Also, Thunderbolt 1 is 10Gbit/s so no faster than USB because USB is max 10Gbit/s also
It's exactly what they have on the 2016 model (Thunderbolt 3 controllers have an in-built USB controller capable of doing USB2.0 through USB3.1 Gen 2). The only difference is that you need a different connector on one side of the cable (USB-C instead of USB-A). People are overreacting and exaggerating the entire USB-C thing here. After a certain amount of time you'd want or even have to buy a new cable for your device anyway (especially how most people are treating their cables).
No you are simply forgetting the fact that Thunderbolt 3 is backwards compatible with versions 1 and 2.
The fact is that you have very little understanding of the Thunderbolt 3 and USB-C.
- Thunderbolt 3 integrates USB; the Thunderbolt 3 controllers come with an in-built USB controller capable of doing USB2.0 through USB3.1 Gen 2.
- USB-C is a universal standard for both USB and Thunderbolt 3
- Only USB-C cables meant for Thunderbolt 3 are not compatible with device only support USB; there is 1 exception and that would be the passive 0.5m USB-C cable which can be used for both protocols
- USB-C cables meant for USB are compatible with Thunderbolt 3 due to the first bullet point
If it is a passive USB-C cable it is going to work, if it is an active cable it will be Thunderbolt 3 only (and thus not work at all on a USB computer). Since the only passive cable compatible with USB and Thunderbolt is the 0.5m one that would mean that nearly all of the Thunderbolt 3 cables are not going to work. The info is in the USB-C spec you can get from the USB-IF website as well as the Intel overview brief.
Let me see if I get you correctly: Because TB3 is backwards with TB1 and 2, vendors will rapidly adopt a technology that requires a separate chip on a protocol that isn't published at a rate that will make all computers having any other ports obsolete in 2-3 years?
Or you didn't bother to read what DanielDD wrote before you went in with all your smarts?
Just wondering who benefits or learns something from this list. Without version info, there's actually no useful info.
hdmi2.1?
dp1.3?
dp1.4?
And I'm sad to report that you didn't get that correctly at all due to completely failing to understand the point. I was pointing at the classic case of the chicken and the egg.
There are many Thunderbolt devices out there as well as USB devices. Since Thunderbolt 3 is fully compatible with all of those there is a whole big world out there that is fully compatible with these Thunderbolt 3 ports on the MBP. It is compatible right now and it is compatible in 10 years time.
Because of the many devices that are compatible with Thunderbolt 3 and the fact that they'd also need a separate chip for USB 3.1 Gen 2 (with Thunderbolt 3 they also gain that protocol) it makes more sense for OEMs to outfit their notebooks with Thunderbolt 3. And that is exactly what they are doing. There are far more non-Apple computers with Thunderbolt 3 than with Thunderbolt 1 and 2 combined and it is still increasing.
It would have been different if there weren't that many devices that were compatible with Thunderbolt 3. The OEMs then wouldn't have had a reason for including it in their computers. Aka the classic case of chicken and egg.
If you actually bothered to read both his and mine reply you'd not even have submitted this reply.
You're not ensured from a failure after your warranty expires
And then what? Instead of just replacing a ssd you'll have to buy a new mac. nice
I find that highly unlikely. I think the writing is pretty clearly on the wall that USB-C as a connection standard is here for a while. (Too bad Apple is going to continue with Lightning on mobile devices. Gross.)
If you can't fix it yourself, you take it into Apple or some other repair service to fix the SSD. No need for a new Mac unless something more critical dies. But anyway, SSDs don't die after 3 years. HDDs vary more because they spin, and stuff goes wrong, but SSDs won't have issues so soon.You're not ensured from a failure after your warranty expires
[doublepost=1495874086][/doublepost]
The connector won't change, but the protocol will. And, for some reason, the cable you use actually matters. It's a mess.
Last edited:
Your point was not at all clear or even articulated in your post.
Let's re-cap. DanielDD said
To which I responded
which you then quoted and commented with
The line of discussion was whether or not anyone with old ports will regret having them in 2-3 years, and you supported the initial claim because "Thunderbolt 3 is backwards compatible with versions 1 and 2".
I then asked you if you are saying that this backwards compatibility is what is driving vendors to adopt TB3 in a rate where is is pretty much the only port used in 2-3 years, and you responded 'No' and that I am missing the point.
The original claim was that those with newly purchased computers with other ports than TB3 would regret their purchase in 2-3 years. The regrets were said to come only from the lack of TB3 ports, not from anything else. That may be true if and only if these buyers are finding themselves wanting but not able to connect to the devices they like to connect to in 2-3 years.
So what is actually claimed is that all devices people are generally wanting to use in 2-3 years time will not work with a computer without a TB3 port.
I find this to be a preposterous claim, and thus I said the poster is greatly overestimating the adoption rate of TB3.
You've come back to explain that you mean that because many USB devices (and older TB devices) can be used with a TB3 port, it will make sense for vendors to adopt TB3 even if there are very few TB3 devices. I agree that it makes sense, but only if TB3 is easily available and not more expensive than what ever they are putting in their machines today. A lot of indicators are to the opposite - for one TB3 is not integrated in any existing Intel CPU.
This is what the news article was about - Intel announced that they will integrate TB3 in the future and will release specifications next year. Not "now" like you claimed.
To sum up.
There are currently 99 computers/motherboards that has TB3. I don't know all these brands, but isn't it fair to say that of these 99 models, only 3 are dropping all other ports, and 96 out of 99 TB3 equipped models are keeping other ports alongside with TB3?
According to Intel, there are currently only 56 devices using TB3 in the entire universe.
Yes, there are many devices not using TB3 that can be connected to a TB3 equipped computer using one of the 23 existing adapters. But these devices can also be connected to their legacy port, which already exists on virtual all motherboard designs. The TB3 interface does not bring any benefit to these millions devices, while the need for an adapter bring costs and inconveniences.
So is it really fair to claim that everyone with a computer that has say a separate display port, an array of USB ports, a firewire port and/or a SD card slot will regret their purchase decision within 2-3 years?
Are we seeing a explosion on TB3 interfaces in the near future to a point where all other interfaces are going extinct in 2-3 years?
I think you know the answer, so just stop trying to cover up your crusade. You do have a point in how TB3 is compatible and can be used for many existing devices but here is no link to a high adoption rate from there.
Thunderbolt isn't only used in devices, it is also used in computers. In total the adoption rate is far higher mainly due to the computers and it is still rising. That's what happens if you have a piece of technology that also does USB (right up until the very latest version). If we look at USB-C in general the adoption rate is even higher. Many gaming notebooks are using it because they can allow for something that is called an eGPU.
Let me remind you that Thunderbolt 3 is able to connect nearly any kind of device because it is speaks that particular protocol. This has been the entire idea behind Thunderbolt since the very beginning: having 1 type of port/connection that piggybacks other protocols so users only have to connect 1 cable. This increases ease of use because when buying/using cables users don't have to think about what the right cable should be. You really need to start looking at Thunderbolt as what it is: a piggyback protocol. Thunderbolt itself is meaningless, the meaning is given by the protocol that it carries (right now: PCIe, USB, DisplayPort, Ethernet and USB PD).
From that perspective the people with the new computers only have 1 cable to worry about while the ones with the older computers still have to deal with this cable mess. Not to mention that there is an increase in the amount of USB-C devices in general leading to being far more USB-C devices on the market in 2-3 years time. The ones with the older machines will have to deal with adapters and if you read on the forums here and elsewhere you can see that many are vehemently against using those.
That's a really bad argument since we can mention all the other components on the PCB that aren't in the CPU. In order to have it in the CPU it has to be technical feasible and right now it isn't. The problem here is that this also applies to USB 3.1 Gen 2. You either need that USB version or Thunderbolt to have these nice solutions, especially in business where you can save lots of money because you can now buy docking solutions that aren't tied to a specific model of a specific brand.
Look at what Thunderbolt 3 is and you'll quickly understand why an OEM wants it in their machine even though it requires a separate chip. The only thing having it integrated in the CPU will do is accelerate it or allow for even smaller/flatter machines.
In other words, you need to look at the entire picture and not at that one little detail you dug up.
That's funny since I never made any claim like that, in fact I stressed the point (and am doing so above once more) that this is a future thing, not something we have now.
Wouldn't matter as it is completely irrelevant to the discussion. After all, this is about Thunderbolt 3 adoption and as you've just pointed out, those computers/motherboards are equipped with that.
That's the point exactly: you can still use all the old stuff which is rather important considering the cost of quite a lot of the Thunderbolt devices (not many will easily buy a new storage solution setting them back another $2000).
Yep it really is fair to make such a claim as long as it is based upon actual facts. It is also fair to claim the exact opposite and again as long as it is based upon actual facts. It depends on several things such as being optimistic/pessimistic, kind of computing that is done and so on. Main thing here is that you actually use facts to base your claim/opinion on and respect someone else's even if it is very different from yours.
The answer is that people like you need to stop being childish and disrespectful towards others. Base things on facts and discuss these with facts instead of doing what you are doing: jump to personal attacks the moment someone disagrees. There is a basis for all of the claims whether you like it or not but it requires you to do your homework, really think about what everything actually means and actually like at the entire picture instead of the minute details. Some people seem to be seeing problems in everything. If you want to accuse me of something then accuse me of opposing the idea that there are problems in everything. I don't care about problems, those are the most unimportant things we have. Solutions are what you need to look for. Try doing that instead of fiercely shaking your head, stamping your feet and shouting "NO".
If you ask me personally then I'd say it is not going to be a problem just like USB-C only and single port computers have been a problem since they were released to the general public. People buy them, use adapters or get different cables and get on with their business/life.
Wow! Let's take a qwik-poll:
47. If all other ports are removed from the design of a computing device, how many USB-C/Thunderbolt 3 ports -built into the chassis- should such a computing device have?
Circle the best answer.
(a) One.
(b) Two.
(c) Four.
(d) Six.
(e) As many as can be engineered to fit a particular device.
(f) None. My laptop is disposable after 8 hours on Facebook.
(g) None. Bluetooth obviously has more bandwidth and sustained throughput than even USB-3. Duh.
47. If all other ports are removed from the design of a computing device, how many USB-C/Thunderbolt 3 ports -built into the chassis- should such a computing device have?
Circle the best answer.
(a) One.
(b) Two.
(c) Four.
(d) Six.
(e) As many as can be engineered to fit a particular device.
(f) None. My laptop is disposable after 8 hours on Facebook.
(g) None. Bluetooth obviously has more bandwidth and sustained throughput than even USB-3. Duh.
Well, it was you who decided to quote one single row of my post as a part of your crusade. I then just asked about your line of reasoning, and it kicked off a tirade of ever circular arguments. I don't think the discussion can be salvaged, so I'll just going to show you a few of your blind spots.
and what was actually said earlier:
So, did you really stress it was a future thing, not something we have now?
And by the way, this is what Intel said, just to make sure we are fact-based:
"...next year Intel plans to make the Thunderbolt protocol specification available to the industry under a nonexclusive, royalty-free license."
It is not my argument. It is the basis of the entire article we are discussing. "Intel Looks to Broaden Thunderbolt 3 Adoption by Integrating Into Future CPUs".
And who is it that is childish and disrespectful to others?
