This is not a "Pro" machine.

[deconstruct60]

But we don't know how many ports the controllers have. How can x4 lanes be used to feed two ports with 20Gb/s bandwidth?

For the question above, easy. It is a switch. That's how.

" ... Thunderbolt 2 provides that solution. By combining the channels together, Thunderbolt 2 enables two 20Gbps bi-direction channels instead of two sets of 10Gbps channels. There's no overall increase in bandwidth, but the solution is now more capable ...
....
Thunderbolt 2/Falcon Ridge still feed off of the same x4 PCIe 2.0 interface as the previous generation designs. Backwards compatibility is also maintained with existing Thunderbolt devices since the underlying architecture doesn't really change. ... "
http://www.anandtech.com/show/7049/intel-thunderbolt-2-everything-you-need-to-know


All current cables work just as well driven by v1 or v2 controllers. It is the same bandwidth as before with a different switch allocation policy. The is a bandwidth increase in DisplayPort v1.2 data traffic.

Changing the signaling rate to what PCIe v3.0 offers would increase the bandwidth yes.

There is no additional bandwidth capacity on the TB network so that would be a deeply flawed move. Until affordable fiber and/or IOHub chipsets go to PCI-e v3 it isn't going to happen.

 

[subsonix]

No. Subsonix hand waving is deeply and fundamentally flawed.

The exact per port bandwidth, is really a tangent to what we are discussing however.

Yeah. ... more smoke with clearly flawed understanding of how Thunderbolt works.

Are you saying that it's not possible to add 6 devices like that drawing, one per port?

What would be the maximum bandwidth available per port in that scenario?

Pure and utter misdirection . Hiliarious that this get two up votes because it is pure smoke. The straw man is your bogus scenarios which purposely suppress the inherit nature of the thunderbolt controller being grounded in a switch. It is a switch so may example puts it into a switch based context.

The switch enables connection of downstream devices. If there are no downstream devices, as in the scenario presented then it's irrelevant to the discussion. I fully agree that if several devices is added in a daisy chain configuration you don't get added bandwidth in that chain, but that is not what we are discussing.

----------

All current cables work just as well driven by v1 or v2 controllers. It is the same bandwidth as before with a different switch allocation policy. The is a bandwidth increase in DisplayPort v1.2 data traffic.

Take a look at that cable cross section. It clearly shows a 20Gb/s capability in one cable. If a controller have two ports backed by x4 PCIe v2, then utilizing 20Gb/s in one cable would completely starve the other port.

There is no additional bandwidth capacity on the TB network so that would be a deeply flawed move. Until affordable fiber and/or IOHub chipsets go to PCI-e v3 it isn't going to happen.

How about we stop talking about the network? I agree that devices that share a daisy chain will not gain bandwidth magically, that is not what we are discussing however, but a one device per host port scenario.

 

[-hh]

If you're saying we need to put all our drives externally to reduce heat, I don't buy it. That's what fans are for. With low power drives, I'm not sure how much heat the extra heat could really contribute.

That's not what I'm saying, no. Just the obvious fact that the more one adds internally the faster the fans need to go and thus the more noise they make.

That's a fair enough statement if we assume an "All Other Factors Being Equal" perspective.

However, the marketplace reality is that the current MP has a large diameter fan(s) and the typical external enclosures have small diameter fans.

Without delving into the thermodynamic and gas dynamics engineering weeds, to achieve the same air mass for heat transfer capability, a smaller diameter always has to move the air at a higher velocity versus a larger diameter conduit, and higher air velocities 'always' (all other factors equal) results in more noise.

So while it is a good thing that the new Tube has a large diameter fan, the problem is that from a Systems perspective, the change vs the old MP is that it is less likely to be the 'only' fan in the system anymore, due to the context of more external expansion boxes being added to the system.


-hh

 

[deconstruct60]

Except that TB is only capable of 2GBps and there are only 6 of them on the new Mac Pro.

Flawed. It 2GBps of PCI-e data per controller. As there are no 6 physical port controllers it is extremely doubtful there is just one here in the Mac Pro 2013 model. Likely there are three. The 6 ports represent access of over 6GBps which is hardly limiting.

That means any LGA2011 with a few PCIe 3.0 slots technically has more expandability than the New Mac Pro.

Bandwidth wise, just the v3.0 lanes of the E5 Xeon means more bandwidth. But not necessarily an increase in devices. Firmware and reconfiguration allows some bandwidth reallocated. That is flexibility.

On a metric of expandability where talking about more devices then the Mac Pro 6 ports allows to add 36 additional PCI-e switches to the system. A given mainboard has a fixed number of switches. The bandwidth is diluted as add more devices but the device count goes up.

We are comparing thunderbolt to PCIe, since it was expressed to me (by you, among others) that it was an adequate substitution.

Where it makes sense. You'll be hard pressed to find a post where I'm cheerleading moving x16 or x8 cards to Thunderbolt.

The question with the MP 2013 is far more so is whether trading two x4 PCI-e v3.0 slots is better than 4 Thunderbolt ports based on the same two lane bundles. I think that more so depends upon what folks were going to toss into those two x4 slots.

If those cards are primarily current v2.0 cards then the downshift to v2.0 with Thunderbolt isn't a big loss. Users were going to do that anyway. If this is mainly lower-mid range eSATA , audio, FW , USB , etc. cards there is probably as much need in the rest of the Mac line up as there is with the Mac Pro's. So going to format that whole Mac line up can share tips the scales relative to some distant future x4 v3.0 cards that might show up.

It's reasonable to say that a board with 5 slots can each run at 4 times TB2 without being throttled is superior in terms of bandwidth. ......a board like that allows for more total bandwidth than the new Mac Pro's TB ports all combined.

The MP 2013 has no need to ship all the PCIe bandwidth out since there are two x32 worth of PCIe bandwidth that is needed for the GPU cards ( and likely the PCIe SSD also ).

It is just an Apples to Oranges comparison to snarf up all the lanes when all the lanes aren't in question.

As far as PCIe 3.0 16x, my opinion is that it is currently an overkill for everything up to and including GPU--8x (8GBps) seems adequate for the cards on the market. ...--the benchmarks of the 7970 (the "first PCIe 3.0 16x card) running at PCIE 2.0 16x (8GBps) prove that.

You are benchmarking with software optimized for x8 v3.0 worth of bandwidth it should not be too surprising to find out that it doesn't really take advantage of much more. Most of these benchmarks themselves are deeply flawed in evaluating the utility of x16 of v3.0 bandwidth because they have no concept of that kind of bandwidth.

Plenty unless you want more than one SAS port on your new Mac Pro--then you have to buy another $900 SAS controller instead of a single PCIe card with 2-4 ports.

There no good reason a single ( or even dual ) port SAS card would cost $900. In the current Mac Pro, there is no place to put another $900 SAS controller card (presuming don't evict the GPU). The MP 2013 isn't try to expand into markets the current Mac Pro isn't even in.

There are some folks would probably would have preferred the current Mac Pro looked more like the HP z800 , but it doesn't. The MP 2013 moves even further away but example that outstrip the current Mac Pro aren't really very relvant about the MP 2012 - MP 2013 transition.

Yes, it allows me to be a jackass and make the motherboard share lanes, but only if I start utilizing more than 40 lanes. I contend that this is a better option than having two non-replaceable proprietary video cards and six thunderbolt ports with no PCIe expansion--that's been my fundamental point this whole time.

4 x8 slots is a different market than even the current Mac Pro. It isn't going to be suprising if this is a mismatch to the MP 2013.

Also, I like how you were just saying 8GBps PCIe would "throttle" a four port SAS card... I guess we now agree that's not the case?

You are throttling GPGPUs far more than SAS. Still doesn't get past where you labeled these as x16 slots which was only physical.

I think we agree, I'm not sure where either of us got the idea we didn't. There are many on this board (goMac included, IIRC) that stated that Apple had no choice but to remove PCIe due to the addition of thunderbolt.

Apple didn't remove PCIe. They removed standard PCIe sockets and a standard card cage, but PCIe is still there. Don't confuse form for function. The function is still there. The question is whether the form is required.

While Apple could have used just one TB controllers the dominos falls like this. TB means needs at least one embedded GPU. If Apple wants the best possible DisplayPort signal out the TB port then it is a very large embedded GPU.

Next domino fall is need to sell more embedded GPUs to get volume ( not just shifting an iMac GPU over ). A single W7000-W9000 class could supply DisplayPort input for two TB controllers. Another could definitely do two controllers. So that is one x4 slot on the chopping block. Another embedded GPU and could cover both and the whole card cage disappears.

----------

Take a look at that cable cross section. It clearly shows a 20Gb/s capability in one cable.

It is a logical data allocation not a physical one. On the left side there are actually two physical wires in each of those 10Gb's strands. This not a physical wire diagram on either side.

Try looking up link aggregation. https://en.wikipedia.org/wiki/Link_aggregation That is really the only major change to TB v2.



A single device connected on a Thunderbolt port is a daisy chain network. With no devices Thunderbolt serves no purpose. With two Thunderbolt devices you have a network. Hence you have a switch involved. You can hand wave till the cows come home, but it won't change the truth.

 

[subsonix]

It is a logical data allocation not a physical one. On the left side there are actually two physical wires in each of those 10Gb's strands. This not a physical wire diagram on either side.

Of course, but it shows 20Gb/s of traffic in the cable, how many physical wires it takes is irrelevant.

With two Thunderbolt devices you have a network. Hence you have a switch involved. You can hand wave till the cows come home, but it won't change the truth.

Yes of course! We are in agreement of this I don't know why you keep bringing it up, it's irrelevant to our discussion. You are diverting the subject into this discussion about daisy chains, it's irrelevant to the question if the 6 host ports in the Mac Pro can be used for a striped RAID.


I will ask these two questions again.

1.) Are you saying that it's not possible to add 6 devices like my illustration, one per port?

2.) What would be the maximum bandwidth available per port in that scenario?
​

 

[barmann]

So while it is a good thing that the new Tube has a large diameter fan, the problem is that from a Systems perspective, the change vs the old MP is that it is less likely to be the 'only' fan in the system anymore, due to the context of more external expansion boxes being added to the system.


-hh

Very good point .
Vibrations can be an issue too .

 

[slughead]

Flawed. It 2GBps of PCI-e data per controller. As there are no 6 physical port controllers it is extremely doubtful there is just one here in the Mac Pro 2013 model. Likely there are three. The 6 ports represent access of over 6GBps which is hardly limiting.

Actually I assuming there were 12 channels of Thunderbolt on the new mac pro (for 12GBps)--bandwidth which is easily trounced by two PCIe 8x 3.0 slots. You're saying that the three TB2 controllers are only capable of a combined 6GBps? That's even less than I thought! What a waste of the LGA2011 platform!

Bandwidth wise, just the v3.0 lanes of the E5 Xeon means more bandwidth. But not necessarily an increase in devices. Firmware and reconfiguration allows some bandwidth reallocated. That is flexibility.

On a metric of expandability where talking about more devices then the Mac Pro 6 ports allows to add 36 additional PCI-e switches to the system. A given mainboard has a fixed number of switches. The bandwidth is diluted as add more devices but the device count goes up.

Yes, bandwidth wise. I think it more likely for people to have a handful of high-bandwidth devices+controllers than 36 different low bandwidth devices. Maybe I'm wrong.

Where it makes sense. You'll be hard pressed to find a post where I'm cheerleading moving x16 or x8 cards to Thunderbolt.

The question with the MP 2013 is far more so is whether trading two x4 PCI-e v3.0 slots is better than 4 Thunderbolt ports based on the same two lane bundles. I think that more so depends upon what folks were going to toss into those two x4 slots.

I don't understand why people would think that a new Mac Pro would have to have so few (and so low-bandwidth) PCIe slots. They would need to change the case a bit to accomodate more slots, but it hardly follows that it was either a choice between the iTube and a case designed 5 years ago.

If those cards are primarily current v2.0 cards then the downshift to v2.0 with Thunderbolt isn't a big loss. Users were going to do that anyway. If this is mainly lower-mid range eSATA , audio, FW , USB , etc. cards there is probably as much need in the rest of the Mac line up as there is with the Mac Pro's. So going to format that whole Mac line up can share tips the scales relative to some distant future x4 v3.0 cards that might show up.

I'm not sure why you are basically comparing future TB2 products (which do not exist) to current PCIe offerings. Eventually, most all PCIe cards will be 3.0 and higher. If you can say we will forever be doomed to 4x PCIe2.0 cards for drive controllers and such, I can say the new Mac Pro will forever be doomed to overpriced Thunderbolt 1 controllers. You're using theoretical examples such as low-cost thunderbolt2 solutions (which I grant will someday exist), I don't see why I don't get to do the same with PCIe 3.0 solutions (which I think are much more likely to come about).

The MP 2013 has no need to ship all the PCIe bandwidth out since there are two x32 worth of PCIe bandwidth that is needed for the GPU cards ( and likely the PCIe SSD also ).

And who asked Apple to do that? At present, most people don't require 16x PCIe 3.0 for their cards (for the reasons you pointed out as well has hardware limitations with the GPUs). Therefore it makes sense to allow users to run other configurations as would happen with *gasp* PCIe slots. The nMP eliminates choice.

You are benchmarking with software optimized for x8 v3.0 worth of bandwidth it should not be too surprising to find out that it doesn't really take advantage of much more. Most of these benchmarks themselves are deeply flawed in evaluating the utility of x16 of v3.0 bandwidth because they have no concept of that kind of bandwidth.

I admitted that this will change in time, but for now, why does it make sense to throw away nearly half the bandwidth of LGA2011 when it makes much more sense to allow users to just switch their GPU to 16x when the time comes that hardware and software utilize it (could be many years, depending on what users are utilizing it for).

There no good reason a single ( or even dual ) port SAS card would cost $900.

But they do. Maybe someday that will change, but currently Thunderbolt ->SAS is a total ripoff of nearly biblical proportions. Even if they do come down in price and become TB2 compatible (which they probably will), they will still require a ratio of roughly one TB2 port per SAS port. So if I want 2 ports, I need two controllers. Maybe they can create a 2 port SAS with two TB2 inputs? I believe you were saying though that TB doesn't work that way.

In the current Mac Pro, there is no place to put another $900 SAS controller card (presuming don't evict the GPU). The MP 2013 isn't try to expand into markets the current Mac Pro isn't even in.

The current Mac Pro is ancient. As far as what markets the current Mac Pro is in, maybe if Apple had kept up with the technology instead of farting around for the last 3 years, it would be in other markets. This is not a valid comparison. The only comparison is with commonly available Vs the New Mac pro--and even that favors the nMP as it uses hardware that hasn't even been released yet.

4 x8 slots is a different market than even the current Mac Pro. It isn't going to be suprising if this is a mismatch to the MP 2013.

... Six Thunderbolt ports is a different market than the current Mac Pro--at least with currently available thunderbolt options.

Apple didn't remove PCIe. They removed standard PCIe sockets and a standard card cage, but PCIe is still there. Don't confuse form for function. The function is still there. The question is whether the form is required.

The function is still there! *with use of expensive external PCIe Chassis throttled to 2GBps!

While Apple could have used just one TB controllers the dominos falls like this. TB means needs at least one embedded GPU. If Apple wants the best possible DisplayPort signal out the TB port then it is a very large embedded GPU.

Or they could've just ignored thunderbolt displays--something most pro users would have favored over dropping PCIe slots. I'll agree Apple would never have done that but I think it would've been more popular than not allowing for more GPU options.

 

[-hh]

Admin ... I'm not quite sure where best to start a quote on this ongoing conversation..."Last" is as good as anything else.

...

Yes, bandwidth wise. I think it more likely for people to have a handful of high-bandwidth devices+controllers than 36 different low bandwidth devices. Maybe I'm wrong.

It is also starting to smell like a contradiction: we were previously told that internal HDD bays could be eliminated '...because 80% didn't use them', so if we apply the same logic here, then there should have been some evidence that many MP customers were using tons of low performance add-ons. Hmmm...wonder how many of these were external USB2 HDDs?

In any case, what I think I see that's starting to be teased out is that there was a trade-off that Apple had to make regarding the allocations of PCI lanes, and that this trade-off was made more confusing by their simultaneous change from PCIe to a proprietary PCIe slot for the two slots that the onboard GPUs now reside in. See next...

I'm not sure why you are basically comparing future TB2 products (which do not exist) to current PCIe offerings. Eventually, most all PCIe cards will be 3.0 and higher. If you can say we will forever be doomed to 4x PCIe2.0 cards for drive controllers and such, I can say the new Mac Pro will forever be doomed to overpriced Thunderbolt 1 controllers. You're using theoretical examples such as low-cost thunderbolt2 solutions (which I grant will someday exist), I don't see why I don't get to do the same with PCIe 3.0 solutions (which I think are much more likely to come about).

...and while pedantically one can make all sorts of goodness statements, the pragmatic reality is that it is impossible for customer costs to not go up: if nothing else, the replacement of the 'open' PCIe slot with a proprietary one means that unless a large swath of the WinTel PC market suddenly adopts it, there's now zero leveraging opportunities for all GPU manufacturers to make one card for both the PC & Mac markets. Even if the Mac card requires a custom firmware flash, that's dirt cheap in comparison to a second dedicated hardware production line.

But they do. Maybe someday that will change, but currently Thunderbolt ->SAS is a total ripoff of nearly biblical proportions. Even if they do come down in price and become TB2 compatible (which they probably will), they will still require a ratio of roughly one TB2 port per SAS port. So if I want 2 ports, I need two controllers...

Shades of Firewire .. and SCSI .. again: the market will charge what the market will bear, and since the removal of standard PCIe slots creates a market fragmentation who then only has one (TB) path to a capability solution, the fixed costs become a much larger percentage and the subsequent cost will be dear.

The current Mac Pro is ancient. As far as what markets the current Mac Pro is in, maybe if Apple had kept up with the technology instead of farting around for the last 3 years, it would be in other markets. This is not a valid comparison.

Well said: it is a rigged comparison. Sure, we can point to some of the supposedly extenuating factors, but the legacy product compares poorly because it had not been maintained to contemporary 'state of the shelf', let alone what's emerging in late 2013.

The function is still there! *with use of expensive external PCIe Chassis throttled to 2GBps!

It is also possible to take a Ford Mustang and make it into a lawnmower...but that doesn't make it a good value for the customer.

In the end, perhaps the Tube will be a good value for some segment of Mac customers ... and perhaps these customers are also the ones who currently buy the Mac Pro. Or maybe not. Time Will Tell.


-hh

 

[VirtualRain]

My 2-cents on some of this...

2.) What would be the maximum bandwidth available per port in that scenario?
[/INDENT]

Actually I assuming there were 12 channels of Thunderbolt on the new mac pro (for 12GBps)--bandwidth which is easily trounced by two PCIe 8x 3.0 slots. You're saying that the three TB2 controllers are only capable of a combined 6GBps? That's even less than I thought! What a waste of the LGA2011 platform!

If I understand TB correctly... Each TB controller has access to 4 PCIe 2.0 lanes (2GB/s bandwidth). This is not a limitation of the 2011 platform, Apple's implementation, or anything else. It's the TB spec.

There are three TB controllers in the new Mac Pro for a total of 6GB/s bandwidth available across those 6 TB ports. So each pair of TB ports shares (switches) 2GB/s bandwidth.

So if you had an bunch of SSDs capable of 1GB/s each and you plugged one into each of the 6 TB ports, your RAID0 array would theoretically perform at 6GB/s.

However if you had a bunch of SSDs capable of 2GB/s each, you could plug one into each of 3 TB ports (so they are all on different TB controllers) and your RAID0 array would also perform at 6GB/s.

Lastly, if you had a bunch of SSDs capable of 2GB/s each and you plugged one into each of the 6 TB ports, your RAID0 array would still only perform at 6GB/s (not 12GB/s) as it's now being throttled by the capacity of the PCIe lanes feeding the TB controllers.

---------------------

...and while pedantically one can make all sorts of goodness statements, the pragmatic reality is that it is impossible for customer costs to not go up: if nothing else, the replacement of the 'open' PCIe slot with a proprietary one means that unless a large swath of the WinTel PC market suddenly adopts it, there's now zero leveraging opportunities for all GPU manufacturers to make one card for both the PC & Mac markets. Even if the Mac card requires a custom firmware flash, that's dirt cheap in comparison to a second dedicated hardware production line.

-hh

I made this point earlier (I feel like this thread is going in circles), but your sort of implying we live in some kind of PCIe expansion card nirvana now. The market for Mac Pro PCIe cards is too small to get significant vendor attention. It's not dirt cheap to write drivers, do QA and offer support or we would be awash in plentiful cheap PCIe cards for the Mac Pro.

The market for TB perhipherals is already vastly bigger than Mac Pro PCIe cards (all MacBook and iMacs for 2 generations now). Moving to an interconnect that is supported in both the vast mobile computing market and the tiny pro workstation market is really a great thing. It's still early days for TB but it is the best path forward for vendors, consumers and pros. Economies of scale are possible with TB that simply can never exist with workstation PCIe cards.

See My Previous Rant in Post 164... https://forums.macrumors.com/posts/17524608/

 

[dollystereo]

Remeber this post..
This machine is gonna sell like hot cakes. They are gonna be seriously out of stock.
For the CPU crunching guys (like me, I do math) this is an excellent solution.
4 internal drives is not gonna make any difference.

 

[subsonix]

My 2-cents on some of this...

If I understand TB correctly... Each TB controller has access to 4 PCIe 2.0 lanes (2GB/s bandwidth). This is not a limitation of the 2011 platform, Apple's implementation, or anything else. It's the TB spec.

There are three TB controllers in the new Mac Pro for a total of 6GB/s bandwidth available across those 6 TB ports. So each pair of TB ports shares (switches) 2GB/s bandwidth.

So if you had an bunch of SSDs capable of 1GB/s each and you plugged one into each of the 6 TB ports, your RAID0 array would theoretically perform at 6GB/s.

However if you had a bunch of SSDs capable of 2GB/s each, you could plug one into each of 3 TB ports (so they are all on different TB controllers) and your RAID0 array would also perform at 6GB/s.

Lastly, if you had a bunch of SSDs capable of 2GB/s each and you plugged one into each of the 6 TB ports, your RAID0 array would still only perform at 6GB/s (not 12GB/s) as it's now being throttled by the capacity of the PCIe lanes feeding the TB controllers.

If we leave out TB2, that's exactly my point! You probably have to go back a couple of pages (lol) to find how this started.

The discussion of TB2 is really a tangent to the topic of creating a RAID from all ports.

 

[VirtualRain]

If we leave out TB2, that's exactly my point! You probably have to go back a couple of pages (lol) to find how this started.

The discussion of TB2 is really a tangent to the topic of creating a RAID from all ports.

Yeah ok, but TB2 changes nothing in this regard. All it does is combine the two channels into one to enable Display Port 1.2 and 4K displays. The 4 PCIe lanes and the associated data bandwidth for TB peripherals doesn't change at all with TB2. So TB1 or TB2... no difference... both exactly the same as far as data peripherals are concerned. IMHO, they really should have called this TB1.1 as a result. There's not enough difference to warrant a "2".

 

[mabhatter]

Yeah your right, 7 Teraflops, this is just a new Mac mini.

External expansion is much better than internal. With my old pro pulling it out on to the desk and basically taking the whole thing apart to swap out one drive was a pain. Now I can store the mac pro out of sight and just have my hard drive bay on desk. win.

Also may I direct you to the news post: https://www.macrumors.com/2013/06/1...ase-mac-pro-stashed-inside-a-giant-metal-box/

Mac Pro internal drives are only SATA. Come back when they are SAS. "Desktops" ave not used "professional" deives in ages, even from other companies. To feed today's 12-core processors, anything less that 6-8 SSD in raid 5 is underpowered. Or just use a 10Gb FC connection to a SAN. and at that point the "computer" is the cheap part of the setup.

 

[knowhate]

.

 

[-hh]

?...

I made this point earlier (I feel like this thread is going in circles), but your sort of implying we live in some kind of PCIe expansion card nirvana now. The market for Mac Pro PCIe cards is too small to get significant vendor attention. It's not dirt cheap to write drivers, do QA and offer support or we would be awash in plentiful cheap PCIe cards for the Mac Pro.

Sorry for your confusion - you're right that its not been good for Mac cards for .. decades. My point really was more that Apple is moving to proprietary - AGAIN - just as it was appearing that they were finally going to get their non standardization sorted out, which will increase customer costs. True, the software development for a driver isn't cheap, but it is in comparison to the setup of the new hardware manufacturing line, which is where fragmantation nails us.

The market for TB perhipherals is already vastly bigger than Mac Pro PCIe cards (all MacBook and iMacs for 2 generations now).

Sure, but if that factor really was powerful enough to have the suggested influence, shouldn't we already be awash in a plethora of cheap TB peripherals by now?

Moving to an interconnect that is supported in both the vast mobile computing market and the tiny pro workstation market is really a great thing. It's still early days for TB but it is the best path forward for vendors, consumers and pros. Economies of scale are possible with TB that simply can never exist with workstation PCIe cards.

Sure....in theory for the loooong term (I'm more likely to retire before it happens in he real world). In the meantime, even though there's all of these Mac laptops with TB, the WinTel PC *desktop-only* market still outnumbers all Macs combined by 2:1 (laptops & desktops combined).

Hence, the PCIe slot still is quite relevant...and will be - it probably will be threatened more by the rise of the iPad than Thunderbolt.

Remeber this post..
This machine is gonna sell like hot cakes. They are gonna be seriously out of stock.
For the CPU crunching guys (like me, I do math) this is an excellent solution.
4 internal drives is not gonna make any difference.

We are all different. While you personally might love the 2013 Tube, the real question for you will be if you still love the 2019 Tube that still has literally the same GPUs because Of classical Apple neglect, while the rest of the PC market has moved on and has 4x more performance.

Anyone know if the old Mac Pro towers (4,1 - 5,1) will be able to have their Airport cards upgraded with the new wifi AC thats in the new Mac Pro? I wonder what type of connector the new Mac Pro wifi will have- maybe the same card as in the new Macbook Air?

The answer is found in the Apple OEM USB3 card sold for these systems today...

...in other words, not a chance in Hades.

-hh

 

[VirtualRain]

Sure, but if that factor really was powerful enough to have the suggested influence, shouldn't we already be awash in a plethora of cheap TB peripherals by now?

-hh

Touché

 

[Tesselator]

I guess they'll come at about the same rate we saw USB2.0 devices releasing once USB2.0 was finalized.

But keep in mind there's a whole bunch of devices which can operate within the USB3 specification just fine and have no need to sport TB or TB2 connectivity.

 

[subsonix]

Yeah ok, but TB2 changes nothing in this regard. All it does is combine the two channels into one to enable Display Port 1.2 and 4K displays. The 4 PCIe lanes and the associated data bandwidth for TB peripherals doesn't change at all with TB2. So TB1 or TB2... no difference... both exactly the same as far as data peripherals are concerned. IMHO, they really should have called this TB1.1 as a result. There's not enough difference to warrant a "2".

It's a different discussion and perhaps you are right, but intel's preview of TB2 showed a 1100MB/s for storage. So there seems to be some uncertainties still about that, from Anandtech:

Since there's 20Gbps of bandwidth per channel, you can now do 4K video over Thunderbolt. You can also expect to see higher max transfer rates for storage. Whereas most Thunderbolt storage devices top out at 800 - 900MB/s, Thunderbolt 2 should raise that to around 1500MB/s (overhead and PCIe limits will stop you from getting anywhere near the max spec).

 

[VirtualRain]

It's a different discussion and perhaps you are right, but intel's preview of TB2 showed a 1100MB/s for storage. So there seems to be some uncertainties still about that, from Anandtech:

Ok, sorry, I see what you're saying. Thanks for your patience as I catch up here.

Yeah, I read something similar about some limitations on TB1. The fact is both TB1 and TB2 pass the same x4 data stream, so I don't know what's been holding TB1 back. I guess it might possibly be the dual channels in the TB1 physical connection vs a single channel in TB2. But you would think that a TB2 channel (if not TB1) could do a full 2GB/s, not just 1100MB/s or whatever.

 

[-hh]

I guess they'll come at about the same rate we saw USB2.0 devices releasing once USB2.0 was finalized.

Frankly, I'm not that optimistic because the market conditions are different.

For example, looking at the history of Firewire, it didn't really become 'cheap' until eSATA came along ... at a lower cost ... to be a competitor.

For USB, it was expensive in the early iMac days ... but concurrent with USB2 development, its controller chips were pushed down to be a cheap commodity item, which encouraged broad adoption by the manstream WinTel PC market, as there were now a hundred vendors who could crank out new products with minimal in-house development costs. Thus, it was a combination of reduced barrier to marketplace entry and a broadening of the market to sell them in.

But keep in mind there's a whole bunch of devices which can operate within the USB3 specification just fine and have no need to sport TB or TB2 connectivity.

True, but Intel owns both and USB3 was effectively to market 'first', and has Intel much more strongly pushing it, much like USB2 before it (and also has the advantage of backwards compatibility). The net result has been vendors putting it everywhere, even where it isn't required - - there's even now USB3 mice being sold by Dell, HP, Logitech and others.

What this really reveals is that just because two similar products exist doesn't mean that things are really going to be in a 'competitor' environment to drive down costs: USB3 is already cheaper and more ubiquitous than TB, so the introduction of TB as a new competitor is not going to have any influence on USB3 prices...

...and since the WinTel PC market is functionally a commodity market with minimized profit margins, there's simply not the budget reserve to add TB as a superfluous feature until such time that TB can clearly do something that USB3 can't which is also a classical 'Killer App' in market desirability so as to afford them a degree of product differentiation to improve profit margins.


-hh

 

[slughead]

On a side note, I think the repetition of this thread is actually healthy. Most threads with multiple pages don't get read (because they're mostly FOS). This allows facts to get buried. Repeating your arguments makes them stronger and more directed as well as allows information to be presented to different people. Because of the nature of forums and information, facts need to be presented periodically to keep them fresh and accurate.

Understanding of these things as well as the tech itself can change over time, therefore it's good to say "yes, this is still the case" or "it is not this way anymore" or "that guy is totally wrong"

I think while technically inefficient, repetition is a good thing.

 

[slughead]

...and since the WinTel PC market is functionally a commodity market with minimized profit margins, there's simply not the budget reserve to add TB as a superfluous feature until such time that TB can clearly do something that USB3 can't which is also a classical 'Killer App' in market desirability so as to afford them a degree of product differentiation to improve profit margins.

I agree, it is highly unlikely TB will be widely adopted in the Wintel desktop world. I could see, however, a hole in the laptop market that might be filled by TB. High performance external storage/accessories while simultaneously accepting fewer ports is not something foreign to laptop users. TB could make it possible for people who would normally have a desktop for productivity to switch to a laptop (being as how choices are made on the margin). TB really is a viable solution for external storage up to 1100MB/s (as others have pointed out)--this could turn a moderate-performance PC laptop into a decent video editing platform.

Desktop users, however, have PCIe which is clearly superior to TB in terms of bandwidth, product availability, and price. They also have USB 3 for low-cost, medium/low-bandwidth uses that don't require anything like thunderbolt. TB is extraneous and pointless for desktops.

 

[devilcm3]

more CPU choices, if anybody is interested

 

[drsox]

Image

more CPU choices, if anybody is interested

deleted

 

[barmann]

Frankly, I'm not that optimistic because the market conditions are different.

For example, looking at the history of Firewire, it didn't really become 'cheap' until eSATA came along ... at a lower cost ... to be a competitor.

......

...and since the WinTel PC market is functionally a commodity market with minimized profit margins, there's simply not the budget reserve to add TB as a superfluous feature until such time that TB can clearly do something that USB3 can't which is also a classical 'Killer App' in market desirability so as to afford them a degree of product differentiation to improve profit margins.


-hh

Well put .

Also, it could be argued that USB, FW and even eSATA, simply kept up with growing demands, and co-existed somewhat peacefully .
I don't think eSATA had much impact on FW pricing, too niche, but USB kept the FW market honest .

Which seems odd to me; why not eSATA speed, FW flexibility, and USB prices ?
It could be called Lightning or something .

I agree, it is highly unlikely TB will be widely adopted in the Wintel desktop world. I could see, however, a hole in the laptop market that might be filled by TB. High performance external storage/accessories while simultaneously accepting fewer ports is not something foreign to laptop users.

.....

Desktop users, however, have PCIe which is clearly superior to TB in terms of bandwidth, product availability, and price. They also have USB 3 for low-cost, medium/low-bandwidth uses that don't require anything like thunderbolt. TB is extraneous and pointless for desktops.

But how can that possibly work ?

Laptop and desktop peripherals need to be identical, else it creates a gap between the two akin to OSX vs. iOS devices .