It doesn’t bring any urgently needed speed bump and Apple supports everything TB4 supports with TB3 chips already.
Originally, TB4 was a fake news Intel released with Tiger Lake at CES2020 because they were caught with their pants down shortly after AMD released their wildly successful Renoir notebook processors and didn’t have anything new to show.
Ignore it and wait for TB5.
It would seem that Thunderbolt didn’t want to be stuck at 3 when USB has gone to 4 because 4 is bigger than 3.
At least they didn’t go with any of that 3.1 Gen 2 felgercarb.
People forget that Apple and Intel developed TB TOGETHER. It’s not like a PROTOCOL is depending on a specific CPU 🤦🏻♂️
See, this is interesting. Apple is saying that they developed it together. But shortly after the technology became out, Intel said that it wasn’t true. They said that Apple came to them with the idea of a fast port, but that Intel did all the work, and that Apple had nothing to do with the development. So this statement is interesting.
additionally, as far as TB 4 is concerned, VT-D is the reason I was concerned. While it’s true that TB is part of PCIE, VT-D is a technology inside Intel’s’ x86 chipsets. My concern was how Apple would implement an x86 technology. I guess we’ll find out.
VT-d is just IOMMU. It’s hardly proprietary, just something Apple hasn’t had to implement in their own chips because they have only ever provided PCIe (or equivalent) on-die. IOMMU cores exist for practically any popular processor architecture, and Apple can always design their own entirely in-house.
Just like the reason the DTK doesn’t have Thunderbolt. Like I guessed in another thread, it’s simply because the A12 never needed external PCIe, so the pins just don’t exist for a Thunderbolt controller to connect to.
You’re wrong. It is proprietary.
VT-d as branding is proprietary. The specific implementation may be proprietary, though the VT-d specification (PDF) has been published since 2007. Sure, a published specification doesn't make something non-proprietary, but it is a strong indicator the owning entity is open to interoperability. You don't publish something you don't want other people to learn from.
IOMMU as a technology is not proprietary. VT-d is one implementation, but IBM's mainframes have had their own since before the POWER4 architecture. I haven't looked into it in detail, but I'm pretty sure at least some variants of the System/370 had it back in the 70s. That's the only way I could see some features of their hypervisor working.
ARM has an IOMMU implementation called SMMU. No telling if Apple would just use that directly or if they would want to make their own implementation.
They have to publish it so that computer manufacturers can use it. If they keep their hardware secret, then no one will understand what to do with it. Apple can do this because they dont sell on reduction to OEMs. But when you do, you must give complete specs and even training to those using it.
nothing else you mentioned has anything to do with this.
People forget that Apple and Intel developed TB TOGETHER. It’s not like a PROTOCOL is depending on a specific CPU 🤦🏻♂️
See, this is interesting. Apple is saying that they developed it together. But shortly after the technology became out, Intel said that it wasn’t true. They said that Apple came to them with the idea of a fast port, but that Intel did all the work, and that Apple had nothing to do with the development. So this statement is interesting.
If a business analyst approaches me for a new feature idea in my app, and provides needed operating parameters for it, and then I implement it...did I do all the work? Or did we work on it together? Both could be considered fairly accurate. Don’t know the particulars, of course, but just a thought.
If I come to you and say that we should do something, and this is what I’d like to see, and then you do all the hardware and software, did you do all the work? Yes, you did.
People forget that Apple and Intel developed TB TOGETHER. It’s not like a PROTOCOL is depending on a specific CPU 🤦🏻♂️
See, this is interesting. Apple is saying that they developed it together. But shortly after the technology became out, Intel said that it wasn’t true. They said that Apple came to them with the idea of a fast port, but that Intel did all the work, and that Apple had nothing to do with the development. So this statement is interesting.
If a business analyst approaches me for a new feature idea in my app, and provides needed operating parameters for it, and then I implement it...did I do all the work? Or did we work on it together? Both could be considered fairly accurate. Don’t know the particulars, of course, but just a thought.
If I come to you and say that we should do something, and this is what I’d like to see, and then you do all the hardware and software, did you do all the work? Yes, you did.
Intel didn’t have a ship able product in light peak months later it was very different in thunderbolt.
Problem is also the product never transitioned into Intels vision that they were pushing just months earlier. If they drove development it should be what they were selling by now. Instead it is every thing Apple wanted from USBc and thunderbolt.. it even looks like it will kill off lightening as well which they were clearly working on parallel to thunderbolt and willing to bet the farm on.
From the outside it sure looks like Apple lead a lost project from the side to success.
People forget that Apple and Intel developed TB TOGETHER. It’s not like a PROTOCOL is depending on a specific CPU 🤦🏻♂️
See, this is interesting. Apple is saying that they developed it together. But shortly after the technology became out, Intel said that it wasn’t true. They said that Apple came to them with the idea of a fast port, but that Intel did all the work, and that Apple had nothing to do with the development. So this statement is interesting.
additionally, as far as TB 4 is concerned, VT-D is the reason I was concerned. While it’s true that TB is part of PCIE, VT-D is a technology inside Intel’s’ x86 chipsets. My concern was how Apple would implement an x86 technology. I guess we’ll find out.
VT-d is just IOMMU. It’s hardly proprietary, just something Apple hasn’t had to implement in their own chips because they have only ever provided PCIe (or equivalent) on-die. IOMMU cores exist for practically any popular processor architecture, and Apple can always design their own entirely in-house.
Just like the reason the DTK doesn’t have Thunderbolt. Like I guessed in another thread, it’s simply because the A12 never needed external PCIe, so the pins just don’t exist for a Thunderbolt controller to connect to.
You’re wrong. It is proprietary.
VT-d as branding is proprietary. The specific implementation may be proprietary, though the VT-d specification (PDF) has been published since 2007. Sure, a published specification doesn't make something non-proprietary, but it is a strong indicator the owning entity is open to interoperability. You don't publish something you don't want other people to learn from.
IOMMU as a technology is not proprietary. VT-d is one implementation, but IBM's mainframes have had their own since before the POWER4 architecture. I haven't looked into it in detail, but I'm pretty sure at least some variants of the System/370 had it back in the 70s. That's the only way I could see some features of their hypervisor working.
ARM has an IOMMU implementation called SMMU. No telling if Apple would just use that directly or if they would want to make their own implementation.
They have to publish it so that computer manufacturers can use it. If they keep their hardware secret, then no one will understand what to do with it. Apple can do this because they dont sell on reduction to OEMs. But when you do, you must give complete specs and even training to those using it.
nothing else you mentioned has anything to do with this.
No, they really don't. They just have to publish instructions and what those instructions are meant to do. Intel goes beyond that and discusses implementation details which would only be relevant to others wanting to implement similar technologies.
Again, VT-d is just Intel's IOMMU implementation. Any moderately-capable IOMMU can fill the requirements for Thunderbolt. Intel mentions only VT-d because that's their technology, and they are the ones publishing the Thunderbolt spec. Intel themselves say the requirement is actually DMA protection, that they use VT-d to meet this requirement, and how other vendors meet it is up to them:
"Thunderbolt is open to non-Intel-based systems. Like any other system, devices must pass Thunderbolt certification and end-to-end testing conducted by third-party labs. Thunderbolt 4 requirements include Intel VT-d based or an equivalent DMA protection technology that provides IO virtualization (often referred to as IO Memory Management Unit or IOMMU), as well as OS implementation support. If the equivalent technology supports prevention against physical attacks, then that should meet the requirement."
So in short, you do not know what you're talking about and have not done the required reading.
Although I suspected Apple would somehow support Thunderbolt going forward on Apple silicon because of their historical involvement in pushing it with Macs, I was worried when I heard that the DTK (Developer Transition Kit) didn’t include Thunderbolt ports.
Now that’s cleared up, for me, it’ll be a disappointment unless they support Thunderbolt 4 as I believe PCs will clearly cheap out with USB4 implementations (and probably can’t blame them), but for Apple premium pricing, I think we should be able to expect them to keep leading the Thunderbolt charge!
As for TB4 itself, I find the multiple port ability a little odd as I always thought you can either go FireWire-like daisy-chaining OR USB-like hubs, but not both?
I myself always thought the multiple ports was a non-issue because by time you use one dock or display (with built-in dock), you’ve probably used up all the bandwidth available on that link and a lot of docks I saw had a USB3.1 port in addition to a downstream TB3 port, so nullified the argument that there wasn’t enough high-speed ports available.
Now that I’m writing this though and thinking about it, maybe others have complained based on the idea that you want to be able to plug-in with just one cable? I thought that’s why as long the downstream device had a daisy-chaining port you can’t complain but I guess most TB devices only have a single port.
My main gripe with TB4 though is where’s the doubling of bandwidth? I see it more of a consolidating revision rather than a major upgrade. 80Gbps for me would mean eGPUs and PCIe 4.0 x4 SSDs would have the bandwidth they need and at the same time, mean that single port-docking could actually be much more of a reality because there’s just that much more bandwidth to play with on a single link, it’s less likely you’ll need it all at once (unless you lay out for the latest and greatest all the time and in that case, you’re already used to using multiple TB ports).
People forget that Apple and Intel developed TB TOGETHER. It’s not like a PROTOCOL is depending on a specific CPU 🤦🏻♂️
See, this is interesting. Apple is saying that they developed it together. But shortly after the technology became out, Intel said that it wasn’t true. They said that Apple came to them with the idea of a fast port, but that Intel did all the work, and that Apple had nothing to do with the development. So this statement is interesting.
additionally, as far as TB 4 is concerned, VT-D is the reason I was concerned. While it’s true that TB is part of PCIE, VT-D is a technology inside Intel’s’ x86 chipsets. My concern was how Apple would implement an x86 technology. I guess we’ll find out.
VT-d is just IOMMU. It’s hardly proprietary, just something Apple hasn’t had to implement in their own chips because they have only ever provided PCIe (or equivalent) on-die. IOMMU cores exist for practically any popular processor architecture, and Apple can always design their own entirely in-house.
Just like the reason the DTK doesn’t have Thunderbolt. Like I guessed in another thread, it’s simply because the A12 never needed external PCIe, so the pins just don’t exist for a Thunderbolt controller to connect to.
You’re wrong. It is proprietary.
VT-d as branding is proprietary. The specific implementation may be proprietary, though the VT-d specification (PDF) has been published since 2007. Sure, a published specification doesn't make something non-proprietary, but it is a strong indicator the owning entity is open to interoperability. You don't publish something you don't want other people to learn from.
IOMMU as a technology is not proprietary. VT-d is one implementation, but IBM's mainframes have had their own since before the POWER4 architecture. I haven't looked into it in detail, but I'm pretty sure at least some variants of the System/370 had it back in the 70s. That's the only way I could see some features of their hypervisor working.
ARM has an IOMMU implementation called SMMU. No telling if Apple would just use that directly or if they would want to make their own implementation.
They have to publish it so that computer manufacturers can use it. If they keep their hardware secret, then no one will understand what to do with it. Apple can do this because they dont sell on reduction to OEMs. But when you do, you must give complete specs and even training to those using it.
nothing else you mentioned has anything to do with this.
No, they really don't. They just have to publish instructions and what those instructions are meant to do. Intel goes beyond that and discusses implementation details which would only be relevant to others wanting to implement similar technologies.
Again, VT-d is just Intel's IOMMU implementation. Any moderately-capable IOMMU can fill the requirements for Thunderbolt. Intel mentions only VT-d because that's their technology, and they are the ones publishing the Thunderbolt spec. Intel themselves say the requirement is actually DMA protection, that they use VT-d to meet this requirement, and how other vendors meet it is up to them:
"Thunderbolt is open to non-Intel-based systems. Like any other system, devices must pass Thunderbolt certification and end-to-end testing conducted by third-party labs. Thunderbolt 4 requirements include Intel VT-d based or an equivalent DMA protection technology that provides IO virtualization (often referred to as IO Memory Management Unit or IOMMU), as well as OS implementation support. If the equivalent technology supports prevention against physical attacks, then that should meet the requirement."
So in short, you do not know what you're talking about and have not done the required reading.
Don’t be a wise ass. You dont understand the reality here. This is the important part, and why Apple may not support TB 4 now. TB 3 is different.
from that Anandtech, which I had read the day the article came out, but what you seem to have missed.
” While additional security protections are always a good thing, they ideally need to be based around open vettable standards, something which might limit Thunderbolt for another generation as an Intel-only technology (whereas USB is far more ubiquitous).”
Firstly, in regards to PCIE 4.0 support and Thunderbolt, since it appears Intel is still a driving influence in the spec, it probably won't support PCIE 4.0 until Intel's Chipsets and CPUs do. I've heard previously, that they will probably be moving off of PCIE 3.0 in the next few chipsets, but I've heard conflicting things, like they could skip PCIE 4.0 in favor of 5.0, and things like that. Do I believe they would do that, not really, but it's probably going to still be a year+ until Intel has native support for PCIE 4.0, then add another year or two to bring it to the "Thunderbolt Spec".
Now, in regards to Apple's products, and Thunderbolt, I imagine that all of the Macs will have Thunderbolt, natively, and for a very long time to come. The question would come in with the "smaller devices", such as iPad and iPhone. I imagine that the iPad Pro will be eventually getting Thunderbolt. It may not be making much sense, currently, for a tablet, but they've been positioning the iPad Pro as a laptop replacement, so I could easily see them adopting it.
However, I imagine the more consumer focused iPads (iPad Air and iPad whatever generation), along with the more "higher end" iPhones, such as the iPhone Pro, and the XL/XR(?) would get USB4, with the "higher data transfer speed", but with the more advanced USB PD spec.
The "low end iPhone", the iPhone 11, it's replacement, the iPhone 12, that segment of the iPhone lineup, would probably, eventually, go with USB-C for the port, and USB PD for the power charging, but the data transfer rate could stay at one of the USB 3 specs. The real question would be the iPhone SE, or at least it's next gen, would it stay as the only model with the Lightening Connector, or would it do the same as the "low end iPhone", and the USB-C and maybe USB PD for charging.
A couple of other things, firstly, I could see Apple maybe eventually "retiring" the Macbook Air, or their other, lowest end laptops, in favor of the higher end iPads. That being said, while I doubt they would ever actually make it, something I could see them doing, with the switch over to Apple Silicon is making a "hybrid device". That is, the screen is essentially a version of an iPad, running iPadOS, but it "plugs into" the base, probably using that "smart port" thing recent iPad models have had, which enables/switches it over to run MacOS, probably disabling the touchscreen, but then enabling it to function as a laptop.
Presumably, outside any needed functionality which a laptop would need, and an iPad wouldn't, most of the base would probably have a battery, giving the device a "very long" battery life. I can't recall the name, but I'm sure many of you remember when Microsoft initially released a "Surface Book", they used that hybrid device design, with the detachable screen. I'm assuming Apple could do better. That being said, I freely re-acknowledge that it's highly unlike Apple would ever do this, but I can dream, I suppose.
Comments
nothing else you mentioned has anything to do with this.
If I come to you and say that we should do something, and this is what I’d like to see, and then you do all the hardware and software, did you do all the work? Yes, you did.
Again, VT-d is just Intel's IOMMU implementation. Any moderately-capable IOMMU can fill the requirements for Thunderbolt. Intel mentions only VT-d because that's their technology, and they are the ones publishing the Thunderbolt spec. Intel themselves say the requirement is actually DMA protection, that they use VT-d to meet this requirement, and how other vendors meet it is up to them:
"Thunderbolt is open to non-Intel-based systems. Like any other system, devices must pass Thunderbolt certification and end-to-end testing conducted by third-party labs. Thunderbolt 4 requirements include Intel VT-d based or an equivalent DMA protection technology that provides IO virtualization (often referred to as IO Memory Management Unit or IOMMU), as well as OS implementation support. If the equivalent technology supports prevention against physical attacks, then that should meet the requirement."
So in short, you do not know what you're talking about and have not done the required reading.
Now that’s cleared up, for me, it’ll be a disappointment unless they support Thunderbolt 4 as I believe PCs will clearly cheap out with USB4 implementations (and probably can’t blame them), but for Apple premium pricing, I think we should be able to expect them to keep leading the Thunderbolt charge!
As for TB4 itself, I find the multiple port ability a little odd as I always thought you can either go FireWire-like daisy-chaining OR USB-like hubs, but not both?
I myself always thought the multiple ports was a non-issue because by time you use one dock or display (with built-in dock), you’ve probably used up all the bandwidth available on that link and a lot of docks I saw had a USB3.1 port in addition to a downstream TB3 port, so nullified the argument that there wasn’t enough high-speed ports available.
Now that I’m writing this though and thinking about it, maybe others have complained based on the idea that you want to be able to plug-in with just one cable? I thought that’s why as long the downstream device had a daisy-chaining port you can’t complain but I guess most TB devices only have a single port.
My main gripe with TB4 though is where’s the doubling of bandwidth? I see it more of a consolidating revision rather than a major upgrade. 80Gbps for me would mean eGPUs and PCIe 4.0 x4 SSDs would have the bandwidth they need and at the same time, mean that single port-docking could actually be much more of a reality because there’s just that much more bandwidth to play with on a single link, it’s less likely you’ll need it all at once (unless you lay out for the latest and greatest all the time and in that case, you’re already used to using multiple TB ports).
from that Anandtech, which I had read the day the article came out, but what you seem to have missed.
” While additional security protections are always a good thing, they ideally need to be based around open vettable standards, something which might limit Thunderbolt for another generation as an Intel-only technology (whereas USB is far more ubiquitous).”
Now, in regards to Apple's products, and Thunderbolt, I imagine that all of the Macs will have Thunderbolt, natively, and for a very long time to come. The question would come in with the "smaller devices", such as iPad and iPhone. I imagine that the iPad Pro will be eventually getting Thunderbolt. It may not be making much sense, currently, for a tablet, but they've been positioning the iPad Pro as a laptop replacement, so I could easily see them adopting it.
However, I imagine the more consumer focused iPads (iPad Air and iPad whatever generation), along with the more "higher end" iPhones, such as the iPhone Pro, and the XL/XR(?) would get USB4, with the "higher data transfer speed", but with the more advanced USB PD spec.
The "low end iPhone", the iPhone 11, it's replacement, the iPhone 12, that segment of the iPhone lineup, would probably, eventually, go with USB-C for the port, and USB PD for the power charging, but the data transfer rate could stay at one of the USB 3 specs. The real question would be the iPhone SE, or at least it's next gen, would it stay as the only model with the Lightening Connector, or would it do the same as the "low end iPhone", and the USB-C and maybe USB PD for charging.
A couple of other things, firstly, I could see Apple maybe eventually "retiring" the Macbook Air, or their other, lowest end laptops, in favor of the higher end iPads. That being said, while I doubt they would ever actually make it, something I could see them doing, with the switch over to Apple Silicon is making a "hybrid device". That is, the screen is essentially a version of an iPad, running iPadOS, but it "plugs into" the base, probably using that "smart port" thing recent iPad models have had, which enables/switches it over to run MacOS, probably disabling the touchscreen, but then enabling it to function as a laptop.
Presumably, outside any needed functionality which a laptop would need, and an iPad wouldn't, most of the base would probably have a battery, giving the device a "very long" battery life. I can't recall the name, but I'm sure many of you remember when Microsoft initially released a "Surface Book", they used that hybrid device design, with the detachable screen. I'm assuming Apple could do better. That being said, I freely re-acknowledge that it's highly unlike Apple would ever do this, but I can dream, I suppose.