Apple Preps for Thunderbolt Display Release with MacBook Air EFI Firmware Update

[BC2009]

The Apple Displays look so nice, but it's hard to justify purchasing a monitor for $1000 when there's Samsung monitors for $300. I have other Apple hardware, but a display seems just like any other display. Can somebody tell me what the big draw is? (besides obvious design aesthetics!)

That $300 Samsung monitor is an HDTV, not a true computer monitor. Simply put, it does not support the 2560x1440 resolution that 27-inch LCD monitors typically support. If you add "2560x1440" to your Google shopping search, you will see the competition is in the $800 and up range. The best you will get out of that $300 Samsung is 1920x1080 since it is 1080p. There may be other differences that I am not aware of. Mind you the Samsung will give an HDTV digital tuner and speakers, but one is a TV and one is a monitor.

The competitors are offering comparably priced monitors to Apple when you start matching up specs, and the competitors don't offer Thunderbolt. Anyway, those are some differences you might want to consider.

Thanks.

 

[guzhogi]

So it seems that Windows XP worked on machines where you were legally allowed to install it, and MacOS X worked on machines where you were legally allowed to install it. So what's the difference?

Honest question: where does it say it's illegal to install Mac OS X on a hackintosh? My understanding is the Mac OS X EULA (and Windows' for that matter) are not legal documents. I didn't know that if the police/FBI found out that you installed Mac OS X on a hackintosh, they'd arrest you. I just thought that Apple wouldn't support, or maybe make a civil suit against you. That's hardly illegal as far as I know.

Besides, I think the person was saying anyways that some Apple software seems to only work for products has already shipped, but not future proof the software. How many times have we needed to upgrade iTunes to support a new iPod/iPhone model? Especially for small updates like bigger hard drive or something? I would think a lot of time, the method of putting songs/videos onto an iPod would stay pretty much the same.

 

[RoboCop001]

Software Update.

That didn't work because I'm still having the problems.

 

[Lennyvalentin]

What ridiculous is that those ports are not on an expensive Mac laptop.

You'd think an expensive Mac laptop would have all these ports as standard, but it's not just a matter of putting ports along the sides of the casing. You gotta have ROOM for them also.

Check out a teardown of the macbook air; there literally isn't ANY more space in that thing for anything else than what's already in there. The battery cells take up most of the internal volume, and the tiny logic board is already very crowded.

Most likely, fitting a gigabit PHY and a firewire controller chip would have been impossible, or at least not without totally reworking the logic board by adding more layers to the PCB (which is very costly and labor intensive). Then where do you add the ports themselves?

Also, adding more features to a product will bump up the price tag. Isn't the Air expensive enough as it is, you think...?

 

[gullySn0wCat]

What ridiculous is that those ports are not on an expensive Mac laptop.

So how do you suggest Apple fit an ethernet port into an MBA?!

 

[brdeveloper]

So how do you suggest Apple fit an ethernet port into an MBA?!

Apple loves the adapter approach. One port, thousand expensive adapters. I'm not sure if it isn't the right approach. Sometimes it annoys me when I have to pay premium prices for a cheap cable adapter and sometimes it makes a device look cleaner and elegant.

 

[deconstruct60]

2.
It should be obvious that requiring host firmware updates to support external devices is a very bad support model.

They by be external but these look to the system as though they are internal. PCI-e is largely used as an internal expansion bus.
What you have are MBA that can now boot off of wired Ethernet (where could not before), boot off of FireWire (where could not before), etc.

Not sure how customized Apple makes the EFI firmware for each Mac, but if they do it would not be surprising if the MBA couldn't boot off FW. I doesn't have FW so you can easily see how some engineering may have put that feature on the chopping block before inclusion into the MBA. Or included it, but never tested it so it is buggy. Kind of hard to test a FW device when don't have FW socket.


What appears to be lame in the support model is that Apple doesn't already have a variety of TB dongles ( for FW , ethernet , etc.) internally so can test these TB machines with a battery of tests before they hit the market. TB certainly seems to be immature, even inside of the Apple. Let alone out in the real production contexts.

Apple is certainly pushing hard to get TB deployed faster than the infrastructure is ready. If they didn't though it probably would get snuffed out by the more deeply entrenched alternatives.

 

[AidenShaw]

They by be external but these look to the system as though they are internal. PCI-e is largely used as an internal expansion bus.

What you have are MBA that can now boot off of wired Ethernet (where could not before), boot off of FireWire (where could not before), etc.

Not sure how customized Apple makes the EFI firmware for each Mac, but if they do it would not be surprising if the MBA couldn't boot off FW. I doesn't have FW so you can easily see how some engineering may have put that feature on the chopping block before inclusion into the MBA. Or included it, but never tested it so it is buggy. Kind of hard to test a FW device when don't have FW socket.


What appears to be lame in the support model is that Apple doesn't already have a variety of TB dongles ( for FW , ethernet , etc.) internally so can test these TB machines with a battery of tests before they hit the market. TB certainly seems to be immature, even inside of the Apple. Let alone out in the real production contexts.

Apple is certainly pushing hard to get TB deployed faster than the infrastructure is ready. If they didn't though it probably would get snuffed out by the more deeply entrenched alternatives.

I agree completely - this step (firmware change) is a warning that TBolt may end up as a support headache.

Note also that in addition to the TBolt controller, every TBolt device will need one or more kernel drivers.

When the TBolt display comes out, someone should post their PCI maps with and without the display connected. I'll bet that when it's connected you see an additional USB root controller, a 1394 controller, a GbE NIC, and possibly a "sound card" - all with their own kernel device drivers. The video cam is probably a USB device, and the audio controller might be USB as well.

 

[deconstruct60]

Note also that in addition to the TBolt controller, every TBolt device will need one or more kernel drivers.

Not really.

I'll bet that when it's connected you see an additional USB root controller, a 1394 controller, a GbE NIC, and possibly a "sound card" - all with their own kernel device drivers.

Only if they are different that the controllers that ship inside of other Mac models. Otherwise they are just the same. The minor tweak the drivers (for both) need is code to deal with hot plug/unplug. You would have thought Apple would have fixed that while had Mac OS X is open surgery mode for Lion updates. Of course, I thought they'd get off their duff and put USB 3.0 driver support into the OS also. Hot plug/unplogging PCI-e cards has same effect so they didn't need "finished" TB devices to work on that.

The video cam is probably a USB device, and the audio controller might be USB as well.

Yes and the USB layer already has plug/unplug-and-play support in it so won't be a problem. PCI-e hot plug support isn't required so that is what is likely missing.

 

[AidenShaw]

Only if they are different than the controllers that ship inside of other Mac models. Otherwise they are just the same. The minor tweak the drivers (for both) need is code to deal with hot plug/unplug. You would have thought Apple would have fixed that while had Mac OS X is open surgery mode for Lion updates. Of course, I thought they'd get off their duff and put USB 3.0 driver support into the OS also. Hot plug/unplogging PCI-e cards has same effect so they didn't need "finished" TB devices to work on that.

Actually, your point that "it just works if all of the devices are identical to the devices on the motherboard" nicely illustrates how big the problem could be.

Many are hoping for a TBolt->USB 3.0 hub. Of course, the driver for this will not be an Apple driver - since Apple doesn't use any USB 3.0 controllers like Intel and other vendors do.

SATA RAID will be another popular option - again, not using an Apple driver.

If Apple can't ship a monitor without requiring a firmware upgrade to the host, what are third party hardware suppliers going to do? Will Apple upgrade the host firmware to make your Promise or Sonnettech or Belkin or... work?

And what about next year, when the Apples have a newer version of some of the controller chips than the display or third party device? Will Apple's drivers work with both the old and new versions of the controller on the same system?

Apple's consumer computers have been avoiding third party driver issues for years because they're un-expandable. Now Apple's opened a Pandora's Box by extending the PCIe bus outside of the box, where anyone can connect 'most anything to it. We may find lots of hits if we search for "'Thunderbolt' AND ('plug' OR 'unplug') AND 'kernel panic'".

This could be a TBolt problem, not an Apple problem - a search for "'Thunderbolt' AND ('plug' OR 'unplug') AND ('blue screen of death' OR 'BSOD')" could get hits as well. The Windows ecosystem, however, is pretty good at supporting third party devices.

Yes and the USB layer already has plug/unplug-and-play support in it so won't be a problem. PCI-e hot plug support isn't required so that is what is likely missing.

Of course, that's why I mentioned USB separately and did not suggest that the devices attached to USB needed special drivers because they were on a TBolt USB controller.

 

[sennekuyl]

Lost and lost of interesting stuff

They get that, but there are still people popping up after those same answers suggesting the $300 monitors are a better deal or that this is a particularly expensive monitor.

 

[SeaFox]

So it seems that Windows XP worked on machines where you were legally allowed to install it, and MacOS X worked on machines where you were legally allowed to install it. So what's the difference?

Nice red herring, there.

What does "legality" have to do with my argument to begin with? According to your logic, OSX should not have run on Core2 Duo Hachintoshes either.

Let's remember that all Intel CPUs going back quite a ways have had the same x86 (and later x64) instruction sets, with newer stuff added later. If a piece of software doesn't support $newinstruction, it's simply going to use what it does support from before. Old-Hat-Windows still works because the older support is still there on the i7. There's no reason OSX-written-with-C2D-in-mind can't run on an Ci7 hardware-wise. Something would have to be added to say "check CPU ID and if it's not on this list, don't run".

 

[mijail]

There's no reason OSX-written-with-C2D-in-mind can't run on an Ci7 hardware-wise. Something would have to be added to say "check CPU ID and if it's not on this list, don't run".

As in "check if we are running in one of our official platforms"?
Boy, would that be surprising!

Why exactly are you expecting it not to?
Why is it called "HACKINtosh" to begin with?

 

[gullySn0wCat]

Apple loves the adapter approach. One port, thousand expensive adapters. I'm not sure if it isn't the right approach. Sometimes it annoys me when I have to pay premium prices for a cheap cable adapter and sometimes it makes a device look cleaner and elegant.

I mean, physically fit. Ethernet is quite a tall port (it barely fits into an MBP)

 

[frozencarbonite]

Could they do a firmware update for MacBook Pros (Mid 2010) to use the display port on it as a Thunderbolt connection? It's the same exact size and everything. Looks like they could do that.

 

[mijail]

And what about next year, when the Apples have a newer version of some of the controller chips than the display or third party device? Will Apple's drivers work with both the old and new versions of the controller on the same system?

You mentioned, but then mostly forgot, about the possibility of TBolt simply not being a very mature technology - which it isn't, of course. Maybe there is some real infrastructure flaw as you say,.. or maybe "simply" some problem was discovered and fixed with this update. Who knows.

The Windows ecosystem, however, is pretty good at supporting third party devices.

What does that even mean? "Good" manufacturers do publish drivers for their
devices, and maintain them. "Less good" manufacturers don't publish them, or don't maintain them. Usually manufacturers care less about OS X than about Windows, true. But what has "the Windows ecosystem" to do with that?
If anything, I have come to expect more community support for OS X than for Windows.

 

[deconstruct60]

Many are hoping for a TBolt->USB 3.0 hub. Of course, the driver for this will not be an Apple driver - since Apple doesn't use any USB 3.0 controllers like Intel and other vendors do.

TB is not USB. It is more a complement to inernal PCI-e cards. A USB 3.0 controller PCI-e has the exact same "problem". If it has unknown, non-kernel supported hardware it needs to come with a driver.

As long as TB has the $50 cable and higher than average price costs this should present a major problem. If the TB devices are not engaged in 'race to the bottom' price wars then the costs of licensing and/or building a driver can be incorporated into the cost of the device. It is a problem if want 'cheapest' devices.

This is exactly why TB is not a major competitor to USB. Basic USB support should be incorporated into the basic OS. Windows 8 has USB 3.0 support. Apple probably will be embressed enough soon to roll out USB 3.0 and XHCI support for Mac OS X soon. If they want to avoid the "screwy 3rd party driver" problem they should have done this already, but still some time to fix the problem ( especially if the new Mac Pro rolls out with USB 3.0 .... which is should. There is no excuse why it shouldn't.).

SATA RAID will be another popular option - again, not using an Apple driver.

Hasn't reall impeded the ExpressCard or Mac Pro PCI-e card market for these. It will only get bigger with TB.

If Apple can't ship a monitor without requiring a firmware upgrade to the host, what are third party hardware suppliers going to do?

For non boot devices the firmware is probably a non issue. If want the firmware to hand off the OS boot process to the device then may need to fix the firmware. if only want to collect the PCI-e devices and hand that list to the OS to boot from then all really need is proper drivers.

Will Apple upgrade the host firmware to make your Promise or Sonnettech or Belkin or... work?

I think Apple is going to make the docking port vendors self manage on this ponit. Use Apple parts ( which indirectly makes Apple's parts cheaper... most customers ... lower price) or have a device users can't boot from. If can't boot Apple's dock will so their devices will be at competive disadvantage. Apple is not going to loose sleep over that. This is not the Windows market. If non support leaves Apple with a hardware competitive advantage their is no reason to do it. As compared to a market where Microsoft doesn't sell 95% of the hardware.

And what about next year, when the Apples have a newer version of some of the controller chips than the display or third party device?

A dubious question since Apple supports the hardware they sold for at least 2-3 years. If there are millions of mac out there in active use and a high fraction covered with Applecare with the older chip it will get support.

The only "flaw" here is that the 3rd party vendors have to wait to see which "next gen" choice Apple makes before moving if want to stay underneath the umbrella. It shouldn't be too hard to make good guesses though.

Now Apple's opened a Pandora's Box by extending the PCIe bus outside of the box, where anyone can connect 'most anything to it. We may find lots of hits if we search for "'Thunderbolt' AND ('plug' OR 'unplug') AND 'kernel panic'".

The Mac PRo and ExpressCard macs already opened that box so this isn't "new". The only issue is whether it will be more widespread issue. Not so sure it will be. The cheapest route for peripheral vendors is to ride in Apple's wake as docking station vendors. For parts that Apple doesn't do ( Fiber Channel , SATA RAID , etc. ) the higher TB prices give some air cover for paying for drivers.

The cheapest option for the lower margin players it just do USB 3.0. When Apple adds it and their device just needs basic XHCI driver they don't have to do anything. .

I do think it will have slight impact on Apple's margins. Support costs are going to go up a bit. But it is a price they were going to have to pay anyway to stay competitive in the overall PC market. Average PC costs are going down. There is no way Apple can charge the price premium and have zero PCI-e expansion. That has worked for last several years but the pricing gap is continuing to grow. Throw on top Windows 8 does not look like another Vista and Apple is almost required to make a move to stay competitive.

If Apple can establish a "higher than USB 3.0 prices" market for TB it won't be that big of an impact.

 

[ro050408]

If Apple can't ship a monitor without requiring a firmware upgrade to the host, what are third party hardware suppliers going to do? Will Apple upgrade the host firmware to make your Promise or Sonnettech or Belkin or... work?

And what about next year, when the Apples have a newer version of some of the controller chips than the display or third party device? Will Apple's drivers work with both the old and new versions of the controller on the same system?

Apple's consumer computers have been avoiding third party driver issues for years because they're un-expandable. Now Apple's opened a Pandora's Box by extending the PCIe bus outside of the box, where anyone can connect 'most anything to it. We may find lots of hits if we search for "'Thunderbolt' AND ('plug' OR 'unplug') AND 'kernel panic'".

This could be a TBolt problem, not an Apple problem - a search for "'Thunderbolt' AND ('plug' OR 'unplug') AND ('blue screen of death' OR 'BSOD')" could get hits as well. The Windows ecosystem, however, is pretty good at supporting third party devices.

This is the best post on this thread so far and is precisely what I have been thinking over the past few months of non-stop vaporware announcements of various thunderbolt accessories. So far, after one year, we just have the Pegasus external hard disk array. NOTHING ELSE.

There is a reason for needing a firmware update to get their display working, and we will likely never know precisely why. We can hypothetically guess all day long though. What we do know is SOMETHING was incompatible with the new display, and SOMETHING has caused the constant delays in its shipment. Apple is not known to delay shipment ETA's over and over, so something is VERY wrong with this display. Until this firmware announcement, we were in the dark. Now we know it is thunderbolt related.

AidenShaw is quite correct in that this is pointing towards there being compatibility issues with Thunderbolt that may require Apple to issue firmware updates each time a new piece of hardware is released. That is something Apple has never done previously and likely will not want to do. My guess is they will rely on consumers to purchase USB, Firewire, etc devices that will be compatible with their $1k port expander (aka monitor).

From my own experience of using mini displayport cables for various monitors and various MacBook's over the past few years, there are huge backward compatibility issues with the new multi-use Thunderbolt/Mini Displayport connector. On both of my systems with Thunderbolt, external monitors are no longer auto-detected upon connection/disconnecting the cable. I have to manually make the computer detect the screen each time I connect or disconnect the monitor. If I put the computer to sleep, I have to disconnect and reconnect the cable to get it going again. According to Apple, this is due to the new Thunderbolt interface and will be fixed in a firmware update. Months later....NOTHING. Other people are having severe flickering issues and are being told the same thing.

 

[Mr. Wonderful]

This is the best post on this thread so far and is precisely what I have been thinking over the past few months of non-stop vaporware announcements of various thunderbolt accessories. So far, after one year, we just have the Pegasus external hard disk array. NOTHING ELSE.

You mean, after six months.

 

[ro050408]

You mean, after six months.

Apple announced Thunderbolt a year ago.