Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Sorry, my mistake, I misread his post (Gbps vs GBps). Obviously TB2 can support 2GBps.

The LGA2011 chipset has 40 lanes for 40GBps (with PCIe 3.0). At most, those TB ports will be able to handle 12 GBps, and only divvied up into 2 GBps cables.

The argument often used here is that TB is way better than SATA because it's not "locked in" to something that may get phased out. Fair enough, but meanwhile, TB2 isn't even enough to handle a four port SAS controller--and this is supposed to replace PCIe ? (not saying you made that statement, but others have)

TB2 is not nearly as future-proofed as PCIe (ironic, considering it doesn't even exist in the market yet), though people are making the argument that SATA should be dropped in favor of TB for that reason.


This might be true if I/O devices were always going at their fastest. Truth be told even the super SSDs Apple is using will likely average around 100MB/s over the corse of the day and only reach their highest speeds about 1% of the time (not counting idle times). The trouble is that 1% will be capped on SATA III and SAYA IV and this is why TB2 is superior. It's so superior in fact I kinda wonder if we will even see SATA IV. If I were to guess I'd say no, we won't. I wonder if all we will see is USB3, TB, and TB2 devices from now on?

I also think we will be seeing the end of three and four drive desktop RAID arrays very soon. It's just no longer needed.
 
Explain to me how the new mac pro is a better fit for a professional than something like this. Apple let down a lot of people with this new mac pro. They let people down when they refused to update the machine for over 3 years. That simply doesn't cut it in the pro market. Apple simply decided it was more interested in dicking around with the form factor and making huge compromises than actually serving the pro market. They will use the poor sales of this new mac pro as an excuse to exit the market.

Image

This machine is much smaller than previous mac pro yet still offers expandability and good thermals and noise levels.

HP also offers an even larger z820.


Image
This new Mac Pro will be perfect for the music industry. Most of the equipment are on racks anyway. Most music producers bring their own hard drives to whichever studio they are working out of. An hp in a studio no thank you!
 
This might be true if I/O devices were always going at their fastest. Truth be told even the super SSDs Apple is using will likely average around 100MB/s over the corse of the day and only reach their highest speeds about 1% of the time (not counting idle times). The trouble is that 1% will be capped on SATA III and SAYA IV and this is why TB2 is superior. It's so superior in fact I kinda wonder if we will even see SATA IV. If I were to guess I'd say no, we won't. I wonder if all we will see is USB3, TB, and TB2 devices from now on?

I also think we will be seeing the end of three and four drive desktop RAID arrays very soon. It's just no longer needed.

Any news on sleep cycles to these TB externals? I was, in the past, unable to do anything to keep external disks from sleeping including shell tools. They just did what they wanted and took ages to get moving again. This sours me to an all external disk setup in timeline editing.
Eject them, they stay spinning. Mount them, they go to sleep. Gah!
 
Any news on sleep cycles to these TB externals? I was, in the past, unable to do anything to keep external disks from sleeping including shell tools. They just did what they wanted and took ages to get moving again. This sours me to an all external disk setup in timeline editing.
Eject them, they stay spinning. Mount them, they go to sleep. Gah!

No idea.

But I have a hunch the time frame I talking about excludes rotational only media. I guess HDDs as we know them are real close to the end of their usefulness. When talking about new systems and the I/O speeds supported, HDDs just look way too slow. I guess they will be replaced by SSHDs for mass storage of 2TB to 8TB sizes. And i guess 1TB to 2TB drives for system and project I/O will be PCIe like MacPro comes with internally.

Since both write to banks of NAND type memory sleep spin-down and so forth is a moot topic. In the case of SSHDs spin-down is probably already being done by the controller.

I don't have a good idea on the time frames but I'm betting it will be difficult to find any SATA connected devices by the end of 2015. So this is speaking to the future-proofers out there and also what we're likely to see between now and then in terms of new desktop and workstation offerings. No one knows the future of course but SSHDs take care of the mass storage issue and PCIe or mSATA devices kill the current SSD models - So I guess both HDDs and SSDs as we see them today will be gone very soon.
 
No idea.

But I have a hunch the time frame I talking about excludes rotational only media. I guess HDDs as we know them are real close to the end of their usefulness. When talking about new systems and the I/O speeds supported, HDDs just look way too slow.

Good points. If we see all NAND or cached NAND to platter I can see how the latencies would probably disappear. Making the issue moot. Storage working like Toyota Prius. Hopefully with a little more under the hood though. ;)
 
This might be true if I/O devices were always going at their fastest. Truth be told even the super SSDs Apple is using will likely average around 100MB/s over the corse of the day and only reach their highest speeds about 1% of the time (not counting idle times). The trouble is that 1% will be capped on SATA III and SAYA IV and this is why TB2 is superior. It's so superior in fact I kinda wonder if we will even see SATA IV. If I were to guess I'd say no, we won't. I wonder if all we will see is USB3, TB, and TB2 devices from now on?

I also think we will be seeing the end of three and four drive desktop RAID arrays very soon. It's just no longer needed.

You're just searching for the wrong term. Bam!
 
IDK Thinking of switching.

This is a very interesting thread. From my perspective the jury is still out. I consider myself to be a "pro" customer but maybe I am wrong. I currently use a Mac Pro 4,1 2x2.93 with 24GB RAM and the Mac 680 GTX. I regularly use products like Cinema 4d, Modo, After Effects and Premier for work as well as most of the remaining Adobe Product Line. I am also a hobbyist musician (meaning no one pays me) and a PC gamer. I "THINK" I am the target market for Mac Pro. Obviously no one from Apple has told me that, just my guess.

That being said, I am not particularly concerned about the external storage, or the RAM as much as I am the video card. This part really troubles me. Primarily because one way that I have extended the life of my 2009 Mac Pro has been to continually upgrade the video card. Not being able to do that means that this machine will have a shorter shelf life for me.

To be honest this preview has made me seriously consider a PC. Being able to upgrade my video card whenever I need to, as well as being able to buy a an affordable video card that will be great for work AND gaming, not to mention being able to get rid of dual booting would be SUPER convenient for me. From that perspective a WIN PC looks very attractive.

Video Card issues are not new to Macs obviously and to some extent the idea of a PC has always been attractive for me. I went through this same feeling when Apple switched to Intel. However, at that time Adobe did not offer the ability to change the software your OS required for free like they do now. But as it is today most software companies offer installers for Win and OSX for free. It's still not 100% but enough of them do that the cost to switch is greatly reduced.

That being said, I am really excited about Mavericks (OMG FINDER TABS!) and I am in a completely mac ecosystem, ipad, iphone, appletv, timecapsule, macbook pro and a MacPro at work. So the idea of adding a PC into the mix is a little bit annoying. The question is are the annoyances enough to dissuade me from switching?

Right now I am taking a wait and see approach, following forums threads like this one, reading news sites. On release I'll be reading reviews and looking at what actually happened, deciding what I think that means for the future and then I will decide. I have been planning to upgrade my computer since before the 2012 MacPro was released but held off to wait for this one. However, as it stands now, I would say that I am 50/50 on switching. Which I find shockingly high given my history of using Macs since college.

Obviously take that for what its worth, I am just one user and so it's clearly anecdotal at best, but I cant help but feeling there must be plenty of others like me (Narcissism?), and that its not a good thing for Apple.
 
SATA Express enables SATA over PCIe, it's not really a follow up to SATA 3. It may actually help storage solutions over Thunderbolt.

I wasn't suggesting it would be detrimental to thunderbolt. I was under the impression that it intended to allow for higher bandwidth, and yeah I know it's SATA over PCIe.
 
No idea.

But I have a hunch the time frame I talking about excludes rotational only media. I guess HDDs as we know them are real close to the end of their usefulness. When talking about new systems and the I/O speeds supported, HDDs just look way too slow. I guess they will be replaced by SSHDs for mass storage of 2TB to 8TB sizes. And i guess 1TB to 2TB drives for system and project I/O will be PCIe like MacPro comes with internally.

Since both write to banks of NAND type memory sleep spin-down and so forth is a moot topic. In the case of SSHDs spin-down is probably already being done by the controller.

I don't have a good idea on the time frames but I'm betting it will be difficult to find any SATA connected devices by the end of 2015. So this is speaking to the future-proofers out there and also what we're likely to see between now and then in terms of new desktop and workstation offerings. No one knows the future of course but SSHDs take care of the mass storage issue and PCIe or mSATA devices kill the current SSD models - So I guess both HDDs and SSDs as we see them today will be gone very soon.

I think you're probably correct about all this, and I think Apple's new PCIe SSD is a possible new standard. So what is to stop some company in the future from creating a PCIe card which can hold, say, 8 of those things and run at PCIe 3.0 8x for the whole card?

This is another issue I have with the New Mac Pro form-factor: PCIe is far more future-proofed as it is higher bandwidth than Thunderbolt2. If these solutions appear, say, within the next 3 years, the Mac Pro will be left having external thunderbolt drive hubs at many multiples the cost as a simple PCIe card drawing its power from the motherboard.

So yes, that workstation by HP (was it this thread? I forgot) did have a bunch of SATA bays and ports that may be obsolete in a few years, but it will STILL be more future proof than the New Mac Pro because of its PCIe slots. (I didn't look at the PCIe config, but there are LGA2011 boards with Four PCIe 16x slots and a PCIe 8x).
 
I think you're probably correct about all this, and I think Apple's new PCIe SSD is a possible new standard. So what is to stop some company in the future from creating a PCIe card which can hold, say, 8 of those things and run at PCIe 3.0 8x for the whole card?

This is another issue I have with the New Mac Pro form-factor: PCIe is far more future-proofed as it is higher bandwidth than Thunderbolt2. If these solutions appear, say, within the next 3 years, the Mac Pro will be left having external thunderbolt drive hubs at many multiples the cost as a simple PCIe card drawing its power from the motherboard.

So yes, that workstation by HP (was it this thread? I forgot) did have a bunch of SATA bays and ports that may be obsolete in a few years, but it will STILL be more future proof than the New Mac Pro because of its PCIe slots. (I didn't look at the PCIe config, but there are LGA2011 boards with Four PCIe 16x slots and a PCIe 8x).

Well first I'd like to recommend that you visit the User CP for your account here and change the number of posts per page to 100 instead of the default 25. I wish 100 were the default here actually - then people wouldn't end up saying the same things over and over again in these discussions and we could have a more linear progression of ideas and information. As it is everyone forgets what the thread is about or what has been said, every 25 posts. And it's not about mobile devices because my Android phone loads a page of 100 posts in only a fraction of a second longer than it does for 25.

Anyway, OT suggestions aside, you seem to be asking the right questions - at least from where I sit. But it's not settled out enough for my brain to make intelligent comments. I mean just in this year alone we have been hit with the inundation of SATA III capable devices, TB and TB2 almost back to back, Bluetooth 4.0, USB3.0, mSATA and several mPCIe (card edge) standards, PCIe v3, and the popularizing of (relatively) low cost PCIe direct storage cards - it also looks like we're getting pretty close to seeing PCIe v4 too.

With all this hitting at relatively the same time it's hard to say anything about what will shake out. When comparing MP6,1 TB2 to boxes with PCIe v3 x16 card slots I'd keep in mind a few things:
  • There are six full speed TB2 ports capable of about two gigabytes per second each, That's faster than the burst speeds of any three (and most combinations of four) of the fastest SSD drives you can buy in RAID0 - again per TB2 port.
  • We're talking about speeds of 2GB/s and per port or not that's more than most cards actually use. Even SAS and SATA RAID cards don't. Sure benchmarking applications do but that's about it. Even most video editors don't use that much bandwidth. 4K to 1080p/60 at 4:4:4 has no call to go over even half that.
  • Around 99% of the speed increases people are enjoying from solid-state devices happens at the extreme low end of the bandwidth and is due to severely decreased latency. 99.9% of all I/O happens at under 200MB/s and that's 1/10th the speed TB2 claims.
  • As I understand the solid-state storage market prices will not drop like they do with rotational media. Someone here made the argument that R&D is still on-going at a relatively furious pace and this requires large budgets - and so prices won't drop. And this sounds about right to me. Some of these things would seem to suggest that we won't be seeing too many internal 8-drive PCIe devices.
  • Internal PCIe (card edge) devices need custom drivers for each OS the company sells them for whereas TB and TB2 needs only one driver and all connected devices just work - much like USB. I think this is much more desirable for OEMs and VARs.
What's going to become popular and take the lead for Desktop and Workstation grade systems? I dunno but another thing to consider is that the MP6,1 is certainly not designed to sit as a file-server. To me it seems feasible that TB2 for the faster devices and USB3 for the more common will become the standards in a year or two.
 
I can't understand why Apple is calling a non-expandable, no disc drive machine a "Pro" machine.

"Pro's", aka Professionals, do NOT want external expansion boxes cluttering their desk just because Apple deems it necessary to not allow people to expand it's internal components. ("expansion is external")

Professional users also want to burn media (videos, photos, etc) to DVDs and Blu Rays. Now, on top of a ~$3000+ machine, they have to purchase an additional external drive just to do that. Apple is not gonna undercut the price of the high end ridiculously priced 15" rMBP. ($2799)

Don't get me started on the internal storage. WHY can Pros NOT expand the internal storage? Yes, there are external drives, but when it comes down to it, it's yet ANOTHER external expansion box, cluttering an already cluttered desk.

Have you honestly looked at the desk of any computer professional? It's a mad scientist lab. Anyone complaining about aesthetics, really isn't working hard enough.
 
This is another issue I have with the New Mac Pro form-factor: PCIe is far more future-proofed as it is higher bandwidth than Thunderbolt2. If these solutions appear, say, within the next 3 years, the Mac Pro will be left having external thunderbolt drive hubs at many multiples the cost as a simple PCIe card drawing its power from the motherboard.

To add to what Tesselator said about bandwidth...

I think PCIe was anything but future proof on the Mac the way things were going... and here's why I think the future is brighter with TB...

The Mac Pro market is very niche, so the end result has been a rather limited choice of PCIe cards for the Mac Pro that are truly designed and certified for it. And those that were, often demanded a premium price. Beyond those limited choices, there are some PC cards that will work in a Mac Pro with hacked kernal extensions or flashing and usually come with some compromise (eg. vendor support, boot OS X, will it continue to work with the next version of OS X, etc.).

On the other hand, the overall Mac market is orders of magnitude larger than the market for the Mac Pro, so while TB is still in it's early days, the addressable market for any vendor building a TB peripheral is much much larger than what they might have been looking at by offering a PCIe card instead.

A vendor who sells a good TB peripheral has the last two generations of nearly every MacBook and iMac owner, as well as new Mac Pro owners as their addressable market. That's going to create a lot more products with economies of scale making them more affordable.

To add to this, as PC laptop vendors (another market much more vast than the Mac Pro market) get on board with TB this will continue to add momentum.
 
To add to what Tesselator said about bandwidth...

I think PCIe was anything but future proof on the Mac the way things were going... and here's why I think the future is brighter with TB...

The Mac Pro market is very niche, so the end result has been a rather limited choice of PCIe cards for the Mac Pro that are truly designed and certified for it. And those that were, often demanded a premium price. Beyond those limited choices, there are some PC cards that will work in a Mac Pro with hacked kernal extensions or flashing and usually come with some compromise (eg. vendor support, boot OS X, will it continue to work with the next version of OS X, etc.).

On the other hand, the overall Mac market is orders of magnitude larger than the market for the Mac Pro, so while TB is still in it's early days, the addressable market for any vendor building a TB peripheral is much much larger than what they might have been looking at by offering a PCIe card instead.

A vendor who sells a good TB peripheral has the last two generations of nearly every MacBook and iMac owner, as well as new Mac Pro owners as their addressable market. That's going to create a lot more products with economies of scale making them more affordable.

To add to this, as PC laptop vendors (another market much more vast than the Mac Pro market) get on board with TB this will continue to add momentum.

Those are very excellent points that I find highly depressing (defintely seem like they could be true though). Simply due to the tiny marketshare of the Mac Pro, nobody would make drivers for cool PCIe storage solutions. TB makes it more likely Mac Pro will see expansion.

While PCIe is more future proof in theory (and definitely on other platforms), on Mac it may be better to adopt solutions more universal to all the products on the platform.
 
Those are very excellent points that I find highly depressing (defintely seem like they could be true though). Simply due to the tiny marketshare of the Mac Pro, nobody would make drivers for cool PCIe storage solutions. TB makes it more likely Mac Pro will see expansion.

While PCIe is more future proof in theory (and definitely on other platforms), on Mac it may be better to adopt solutions more universal to all the products on the platform.

The same is true outside the Mac market as well. Have you seen how the desktop vs laptop breakdown is developing? It all points to the fact that boxes with internal card expansion will be increasingly a niche product over time. The availability of affordable computers relies on economy of scale.
 
Well first I'd like to recommend that you visit the User CP for your account here and change the number of posts per page to 100 instead of the default 25. I wish 100 were the default here actually - then people wouldn't end up saying the same things over and over again in these discussions and we could have a more linear progression of ideas and information. As it is everyone forgets what the thread is about or what has been said, every 25 posts. And it's not about mobile devices because my Android phone loads a page of 100 posts in only a fraction of a second longer than it does for 25.

Anyway, OT suggestions aside, you seem to be asking the right questions - at least from where I sit. But it's not settled out enough for my brain to make intelligent comments. I mean just in this year alone we have been hit with the inundation of SATA III capable devices, TB and TB2 almost back to back, Bluetooth 4.0, USB3.0, mSATA and several mPCIe (card edge) standards, PCIe v3, and the popularizing of (relatively) low cost PCIe direct storage cards - it also looks like we're getting pretty close to seeing PCIe v4 too.

With all this hitting at relatively the same time it's hard to say anything about what will shake out. When comparing MP6,1 TB2 to boxes with PCIe v3 x16 card slots I'd keep in mind a few things:
  • There are six full speed TB2 ports capable of about two gigabytes per second each, That's faster than the burst speeds of any three (and most combinations of four) of the fastest SSD drives you can buy in RAID0 - again per TB2 port.
  • We're talking about speeds of 2GB/s and per port or not that's more than most cards actually use. Even SAS and SATA RAID cards don't. Sure benchmarking applications do but that's about it. Even most video editors don't use that much bandwidth. 4K to 1080p/60 at 4:4:4 has no call to go over even half that.
  • Around 99% of the speed increases people are enjoying from solid-state devices happens at the extreme low end of the bandwidth and is due to severely decreased latency. 99.9% of all I/O happens at under 200MB/s and that's 1/10th the speed TB2 claims.
  • As I understand the solid-state storage market prices will not drop like they do with rotational media. Someone here made the argument that R&D is still on-going at a relatively furious pace and this requires large budgets - and so prices won't drop. And this sounds about right to me. Some of these things would seem to suggest that we won't be seeing too many internal 8-drive PCIe devices.
  • Internal PCIe (card edge) devices need custom drivers for each OS the company sells them for whereas TB and TB2 needs only one driver and all connected devices just work - much like USB. I think this is much more desirable for OEMs and VARs.
What's going to become popular and take the lead for Desktop and Workstation grade systems? I dunno but another thing to consider is that the MP6,1 is certainly not designed to sit as a file-server. To me it seems feasible that TB2 for the faster devices and USB3 for the more common will become the standards in a year or two.

My summary of most of this: TB2 is adequate for storage for now and the near future, even if PCIe is technically capable of many times more.

While I do seriously think Thunderbolt is a cool technology (even if NOT a replacement for PCIe), I don't know if it will become a standard on the PC. Even if thunderbolt hard drives start being made, PCIe drives are perfectly adequate and most desktop PCs have extra slots. What will be the standard? I don't know. I do know that PCIe just plain makes more sense for desktop: slot counts are decent, power and space are ample within the average PC case. Why pay for another box with another PSU to get no advantage over PCIe?
 
My summary of most of this: TB2 is adequate for storage for now and the near future, even if PCIe is technically capable of many times more.

While I do seriously think Thunderbolt is a cool technology (even if NOT a replacement for PCIe), I don't know if it will become a standard on the PC. Even if thunderbolt hard drives start being made, PCIe drives are perfectly adequate and most desktop PCs have extra slots. What will be the standard? I don't know. I do know that PCIe just plain makes more sense for desktop: slot counts are decent, power and space are ample within the average PC case. Why pay for another box with another PSU to get no advantage over PCIe?

Ummm, I don't think I agree. First, if indeed the new basic config includes two video cards then that's it for like 90% or the motherboard currently available. When two GPU cards are installed there are no more slots available. Second, this increases system heat which elevates noise levels and increases dust - which decreases life-spans. Also these cards are dependent on custom drivers which TB/TB2 devices are not. With every OS upgrade card users must fear for the functionality or obsolescence of their installed cards - this has been true since IBM came out with the PC XT models and it's still true today on the MP5,1. Every OS release adds and removes compatibility for various specific PCIe cards.

On the issue of performance my guess is that TB2 devices will be the same across the board with no noticeable decreases. The only place I see a problem is if the device is a GPU and ONLY then if it's used for display and not as a compute device. A display GPU connected via TB (or maybe TB2) will suffer a little when used for Game or heavy video. This of course will not affect the MP6,1 which has the display GPUs built in internally already.

So all I see is a potential for expansion and growth. I don't really see it as being "good enough for now but not up to snuff compared to internal PCIe connections" as you suggest.

Oh well, I guess we'll find out soon enough because the basic design has already been determined. By this time next year we will know for sure without having to speculate and guess. :)
 
This is a new and interesting conversation that has not been repeated ad nauseum on this sub-forum. Why not just copy and paste all your previous posts since the beginning of June. It will have the same result. :p :D
 
Gotta repeat it every 25 posts.

There should be a sequel to "White Men Can't Jump" called "Mac Users Can't Click". :D
 
Gotta repeat it every 25 posts.

There should be a sequel to "White Men Can't Jump" called "Mac Users Can't Click". :D
:D

I wish we could at least keep it to one thread. Seems like the whole first page is full of the same discussion and the same points with the same people. I think it's fair to say by now that one side is not going to convince the other? Or have we had any successful conversions? :)
 
:D

I wish we could at least keep it to one thread. Seems like the whole first page is full of the same discussion and the same points with the same people. I think it's fair to say by now that one side is not going to convince the other? Or have we had any successful conversions? :)

Conversions? Hehehe, I guess that's one way to look at it. I would hope it's not like that tho. I would hope we're all openminded enough to admit we don't know everything and consider what others have to say - the collective mind and all that... Especially when talking about all this future stuff.

I hear ya about the scattering tho. There must be 15 threads which have all devolved into the same contentious points - and you can bet someone is about to start another one anytime now. I guess we shouldn't be surprised tho. We saw the same exact thing happen regarding Apple's intentions pre-WWDC - 20 threads all with the same arguments over and over and over. :p
 
Last edited:
Internal PCIe (card edge) devices need custom drivers for each OS the company sells them for whereas TB and TB2 needs only one driver and all connected devices just work - much like USB. I think this is much more desirable for OEMs and VARs.

This is totally not true. TB devices still need drivers. Because TB really is just PCIe, Thunderbolt devices probably need the exact same PCIe drivers.
 
Excellent Point

:D

I wish we could at least keep it to one thread. Seems like the whole first page is full of the same discussion and the same points with the same people. I think it's fair to say by now that one side is not going to convince the other? Or have we had any successful conversions? :)

Being new here I'm trying to learn as much as I can. Great information scattered among many threads. I did send some questions to the blog staff asking about some new sub-categories (complaints, hackintosh, upgrade problems, new Mac Pro, etc.) to make topics easier to find and follow but they seen unwilling to change anything. Professional use categories would also be nice. The only alternative I see is a private forum which has a fee. Anybody else have ideas on this?
 
Nothing Beats PCIe For Speed Connect. The More - The Merrier. Tyan wins.

... the new basic config includes two video cards then that's it for like 90% or the motherboard currently available. When two GPU cards are installed there are no more slots available.
WARNING: Stop here if you don't yearn for system expansion and growth.
Here's where a Dual CPU Nehalem/Westmere LGA 1366 (PCIe 2: ~$3,700 w/o CPUs, ram and storage) [ TYAN FT72-B7015 - http://www.superbiiz.com/detail.php...3c789f87a210&gclid=CIyixZePlLgCFbNj7AodzxUARA ] or Sandy Bridge/Ivy Bridge 2011 (PCIe 3: ~$5,200 w/o CPUs, ram and storage) [ TYAN B7059F77AV6R - http://www.tyan.com/product_SKU_spec.aspx?ProductType=BB&pid=512&SKU=600000346 ] bare bones server has distinct advantages. It's my xMac, xLinux and xWindows platform for system expansion and growth. For instance, the LGA 1366 has ten PCIe 2 x16 slots (two of which have x4 signals and the other eight can cradle eight double wide PCIe cards at x16 speed), two PCI-Ex1 slots and one PCI 32 bit slot.
So all I see is a potential for expansion and growth. I don't really see it as being "good enough for now but not up to snuff compared to internal PCIe connections" as you suggest.
The Tyan servers aren't that portable, but they beat the hell out of buying a $2,500 4-slot headless chassis or the $7,000+ 8-slot headless chassis. Eight Titans reside in my LGA 1366 Tyan. There's no speedier interconnect, except for the 2011 SB/IB CPU version. I just needed a compute system (and thus didn't need PCIe 3 slots). The $1500 savings (under the 2011 version) paid for the CPUs. Tyan server boards are where I now see the greatest potential for expansion and growth. Here, with three double wide video cards installed in the three middle PCIe 2 x16 slots (thus covering the two PCI-Ex1 slots), you won't be crying that there're no more slots because you'll be left with eight slots: seven (7) PCIe x16 slots and a single PCI 32 bit slot for expansion and growth on the fastest peripheral interconnect there is (save the PCIe 3 Tyan model). Then you could populate four of those seven remaining x16 slots with PCIe storage cards. You might configure three of them for Raid 0 for awesome speed and have the fourth as a backup). But that would leave you with three empty slots. Nature abhors a vacuum and "Pros" abhor empty slots - so can you say hello to a Red Rocket, Infiniband, etc. Moreover, you would have to learn to live without having a snake's nest of interconnect and power supply cables and the financial drain of external peripherals and containers for them. With a Tyan barebones server, the hardest thing to figure out might just be how to fill it up.
 

Attachments

  • AlphaCanisLupusIntEmpSide1.jpg
    AlphaCanisLupusIntEmpSide1.jpg
    1,010.7 KB · Views: 88
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.