New Mac Pro Pricing

[Joseph Coleman]

Here is an article I stumbled upon. It should give a pretty good idea of where the Mac Pro pricing will hover around. Its shows that the top tier 12 core CPU alone will retail around $3000. That, plus 2 gpu's and custom boards to fit the new case means this little guy is going to be expensive. I was expecting $5000+ but I wouldn't be surprised if it was in the 6000+ range

http://wccftech.com/intel-ivy-bridge-ep-impresses-xeon-2697-v2-benchmarks-prices-leaked/

 

[Subhadip Sen]

You are being far too optimistic! The bill of materials for the 2 FirePro W9000 GPUs ($3500 each) and CPU (2695 V2 or 2697 V2) alone comes up to over $9000, and that is assuming hefty OEM discounts. Not to mention an expensive PCI-e SSD and a thoroughly unique thermal system and chassis. The configuration shown at the event in June is likely to run up closer to $15k-$20k.

I do think there will be entry level configurations - possibly with a 6 core or 8 core and a single mid-range GPU (FirePro W5000?) - and that could possibly start in the $5000-$6000 range you mention.

 

[Jeff Kilgroe]

Comparing the current Sandy Bridge EP Xeon pricing to the new Ivy Bridge EP prices, we see that the 2687W or 2690 (current models) and the new 2697 flagship 12-core Ivy Bridge model are right at the same price per core. One could say the new Ivy Bridge EP is a better deal -- essentially the same price per core, but increased performance at lower clock rate and much lower power consumption. Also more cores per chip meaning higher density.

That said, I still have to agree with Subhadip on most everything else. I don't want to think that a maxed-out Mac Pro will be $15K or more, but I can't seem to find any way to make the parts add up to less. I suppose I shouldn't complain. I have over $15K invested in my Z820 workstation and that doesn't take into account the Rocket card, monitors or control surface.

I'm betting the entry-level model Mac Pro will be priced around $2K and for sure under $2500 (as in $2499.95). As you upgrade the GPU, add a second GPU, upgrade the CPU, max out the RAM, move from the 128GB (presumed) base internal storage to 1TB+, etc... it's going to add up real quick.

 

[Richard Tran]

Does the limited 4DIMM slot bother you, Jeff?

 

[Gabor D.]

So for someone who'd like to get a good desktop computer for editing in FCPX, but don't have the money to spend 10Ks on a Mac Pro, is better off with a maxed out iMac? I would be curious to know how would a maxed out iMac would compare with an almost entry level Mac Pro in terms of power. (price-wise they almost the same according to current opinions on Mac Pro pricing, maybe the maxed out iMac is little bit more expensive )

 

[Brian Merlen]

why not max out a quad core mini? for starters, nothings soldered... yes the gpu will suck...but the cpu and the ram aint bad...add a fusion drive and its a nice boost relative to optical. i think it makes a decent cheap edit bay. granted it'll never perform like a xeon or something with a dedicated GPU

 

[Mike Halper]

Build a Hackintosh, overclock it, and be done for a while. Mine is running very stable. Color correcting and transcoding in Redcine-X Pro, color correcting and rendering in Resolve, editing, rendering, etc. in FCP7, FCPX, and Avid MC7. All are running with zero issues right now. I had a problem with Premiere Pro CC and Redcine-X Pro running because of the GTX Titan not being fully supported, but I fixed it with kexts/drivers from the 10.9 preview and all is good now. I also ran the Prime95 test and passed while overclocked at 4.4GHz and RAM running at 2400 MHz (i7-3770K). And I also have my Titan running at x16, Rocket at x8, Atto HBA at x8, Decklink at x4, and will be adding an Atto RAID controller that will also run at x8. And they will really run at those bandwidths with the motherboard I have (Asus Maximus V Extreme). That's something that no Mac Pro tower or the new Mac Pro will be able to do. Current Mac Pros don't have the power supply, the # of slots, or the lanes to handle it. With the new Mac Pro, there's not enough bandwidth with all those Thunderbolt ports combined to handle all the cards we're all likely to want to use at their full bandwidth plus the various external storage and monitors that would also be run through them.

My system only cost $3K for the motherboard, CPU, 32GB 2400MHz RAM, GTX Titan Superclocked, case, power supply, 500GB SSD, optical drive, in-bay memory card reader, Firewire card, additional fans, and some cables. And if you know the prices of those components, you know a huge chunk of that was the Titan, so you could get great performance for $2K or even less). I score 15,500+ on Geekbench 32-bit (would likely be 17,000+ at 64-bit), and I got a 16,700 score overclocked at 4.6GHz (just 32-bit, so maybe 18,500 at 64bit?). Sure not as high as a 12-core fully loaded Mac Pro or the new Mac Pro, but it performs extremely well for 4-cores and it's at a much lower cost. You could build a small farm of them if you wanted and still be less than the new Mac Pro will be.

The only reason I see for me to upgrade at this point is 1) get 4K display support from integrated graphics with the new Haswell Intel HD Graphics 2) more CPU power for 6k R3Ds. So that would mean either upgrading to Haswell motherboard and CPU for #1, or going to a dual Xeon for #2. Most likely I'd go the Haswell route first as I suspect I'll be able to handle 6K fine since to preview it I should only need 1/2 debayer with 4k displays and 1/4 debayer for 1080p displays and I'll have the Rocket-X, but either upgrade could be a ways off still.

As much as I've been standing behind Apple for such a long time, the new Mac Pro is a huge disappointment to me. If there was no need to add any expansion to it at all, that would be one thing. But as soon as you want to expand it (like we all do in our field) you run into issues of limited bandwidth through the current Thunderbolt standard and dealing with a lot of cables. Not just Thunderbolt cables, but also power cables. And if you're like me, working on set as a DIT very frequently, the last thing you want to do is start off your day dealing with wiring a mess of cables just to get set up and going, and then undoing all that at the end of the day, every day. It becomes a real pain in the ass. I think Apple is completely missing the boat on what people really want and need. Thunderbolt may be the answer a few years from now when the optical version is available and the bandwidth of one port exceeds or meets the needs of an x16 Gen 3 card (or whatever is current at the time), but right now it's just not enough.

If I want a laptop, I'm definitely going the MacBook Pro route. For a workstation, I'm going the Hackintosh route. Just copy someone else's build that's stable and that meets your needs. If enough people start building Hackintoshes maybe Apple will wake up and bring back some kind of tower with internal expansion. Unlikely at this point, but you never know.

Just some thoughts. Hackintosh isn't for everyone, but I think it's worthy of serious consideration.

 

[Brian Merlen]

Even if TB had the bandwidth we needed until Apple fixes the "interpolarity" issues as far as external GPU usage goes, we still are screwed. If they fix it, then I'd just use a Mac Air hooked to a GPU and Rocket...That is why apple won't do it, it'll eliminate our need to use workstations entirely... (and the expensive laptops)

 

[David Bourla]

I'm sorry, the new MacPro is beautiful but lame. What part of this machine is Pro? ATI cards and one CPU? Hello??? That's never going to be an expandable powerhouse for video.

 

[Meryem Ersoz]

Lots of power and lots of throughput and a small, portable footprint is good enough for me, if they can keep the pricetag at or under $5K. I was initially bummed about the lack of expansion, but I think the pros outweigh the cons. I think a Hackintosh is great for a desktop unit, but I think for mobility this will rock my world....like everyone, I kinda scoffed at the trash can design but once you consider how intelligently it works for cooling a powerful system in such a compact design, well, I just want to get my hands on one sooner than later...

 

[Subhadip Sen]

I'm sorry, the new MacPro is beautiful but lame. What part of this machine is Pro? ATI cards and one CPU? Hello??? That's never going to be an expandable powerhouse for video.

Neither is it intended to be a powerhouse. It's clearly targeted at app developers and content creators and will certainly be "fast enough" for a vast majority of it's target customer base, in a very small footprint. Those AMD FirePro GPUs are plenty powerful (more importantly - power efficient, which is a prime concern for this product) and we have seen the move from CUDA to OpenCL well underway this year. Most relevantly, Adobe CC and Resolve 10 will be ready with OpenCL for AMD's GPUs. Single CPU, the law of diminishing returns does start to kick in above 12 cores / 24 threads for a lot of applications used by the target customers. In fact, I am concerned the 12 core is clocked too low, perhaps a higher clocked 8 or 10 core would be the better choice for post production apps.

Having said all of that, I am definitely going to stick with the configurable and expandable powerhouse. For our usage scenario (finishing long form at 4K) this thing is ill suited.

 

[Michael Jarvis]

I'm sorry, the new MacPro is beautiful but lame. What part of this machine is Pro? ATI cards and one CPU? Hello??? That's never going to be an expandable powerhouse for video.

Hey Jeff, whats your take on this. I was definitely going with the new Mac Pro...but a little concerned. I am not a computer tech. I know my way around software and the hardware essentials but not a tech guru like some of the people on here...especially Jeff.

Anyone think there is a way this new system could be expandable and handle 6K? ...also work with the new Rocket...or even need a Rocket. I was hoping the next system would not need a Rocket card, at least at 5K.

 

[Jeff Kilgroe]

Does the limited 4DIMM slot bother you, Jeff?

As long as I can work within 128GB RAM for up to 12 cores, I'm good with it. Actually we'll see 256GB RAM available for this thing before the next generation of Xeon CPUs arrive...

So for someone who'd like to get a good desktop computer for editing in FCPX, but don't have the money to spend 10Ks on a Mac Pro, is better off with a maxed out iMac? I would be curious to know how would a maxed out iMac would compare with an almost entry level Mac Pro in terms of power. (price-wise they almost the same according to current opinions on Mac Pro pricing, maybe the maxed out iMac is little bit more expensive )

Totally different machine with different target purposes. An iMac may fit your needs alright, but it lacks the CPU horsepower, memory throughput and connectivity of the Mac Pro. It's 1/4 of the system the Mac Pro is, but does have it's advantages for some things, other than cost. Like the all-in-one form factor if that fits your needs.

why not max out a quad core mini? for starters, nothings soldered... yes the gpu will suck...but the cpu and the ram aint bad...add a fusion drive and its a nice boost relative to optical. i think it makes a decent cheap edit bay. granted it'll never perform like a xeon or something with a dedicated GPU

Once again, nowhere near the same class of machine. Everything in the mini is soldered except for RAM and drives. Although it's easier to service a mini than it is to service an iMac. Mini is not as powerful as the top model of iMac. Think of it as a Macbook Pro without the battery, discrete GPU or screen.

My system only cost $3K for the motherboard, CPU, 32GB 2400MHz RAM, GTX Titan Superclocked, case, power supply, 500GB SSD, optical drive, in-bay memory card reader, Firewire card, additional fans, and some cables. And if you know the prices of those components, you know a huge chunk of that was the Titan, so you could get great performance for $2K or even less). I score 15,500+ on Geekbench 32-bit (would likely be 17,000+ at 64-bit), and I got a 16,700 score overclocked at 4.6GHz (just 32-bit, so maybe 18,500 at 64bit?). Sure not as high as a 12-core fully loaded Mac Pro or the new Mac Pro, but it performs extremely well for 4-cores and it's at a much lower cost. You could build a small farm of them if you wanted and still be less than the new Mac Pro will be.

This reasoning always confuses me. Not the Hackintosh part, but the hardware people promote as an alternative. Once again, this, and most systems people tend to build as Hackintoshes, are not the same calibre of machine. I see a continuous stream of Hackintosh configurations based on the desktop i7 platform. Gee, that's fine and great, it may fit your needs. And truth be told, Apple has a gaping hole in their product offerings by not shipping a competent desktop class i7 system (although that may be changing). What I'm saying is that while this is an alternative that some may find useful, it doesn't solve anything for people who need the 8, 10 or 12+ core Xeon hardware. If you need a reasonably configured 12-core Xeon system, even building it yourself, it's going to cost you nearly double what that i7 desktop system costs.

Neither is it intended to be a powerhouse. It's clearly targeted at app developers and content creators and will certainly be "fast enough" for a vast majority of it's target customer base, in a very small footprint. Those AMD FirePro GPUs are plenty powerful (more importantly - power efficient, which is a prime concern for this product) and we have seen the move from CUDA to OpenCL well underway this year. Most relevantly, Adobe CC and Resolve 10 will be ready with OpenCL for AMD's GPUs. Single CPU, the law of diminishing returns does start to kick in above 12 cores / 24 threads for a lot of applications used by the target customers. In fact, I am concerned the 12 core is clocked too low, perhaps a higher clocked 8 or 10 core would be the better choice for post production apps.

Having said all of that, I am definitely going to stick with the configurable and expandable powerhouse. For our usage scenario (finishing long form at 4K) this thing is ill suited.

It will be interesting to see which CPU options Apple will offer. I also think that many will be better served with faster clocked CPUs having fewer cores. A 10-core 3.1GHz or 8-core 3.7GHz will be more advantageous to many, depending on the software being used. Intel has faster 12-core CPUs coming, this is just the first wave of Ivy Bridge EP Xeons. However, the cost will increase with speed as it usually does. Premiere CS6/ CC seem to crap out (diminishing returns) between 16 and 20 threads for most things, even working with R3Ds. So unless you know for a fact that you need the extra cores -- doing lots of 3D rendering or other such tasks that scale to large numbers of CPUs -- I think an 8 or 10 core system would be more suitable.

One area that I know many are really salivating over the new Mac Pro is for high-density computational work. Multiple Thunderbolt 2's means there's 20Gbps interconnect fabric right out of the box and if stacked on their side and staggered up like soda cans, even though they're only one CPU per system, it's still a very efficient design and offers better density than current dual-Xeon 1U offerings. Apple is marketing this as a pro workstation, but they may have unwittingly created a truly awesome compute "appliance" that may offer a real advantage over the "blade" paradigm.

 

[Jeff Kilgroe]

Hey Jeff, whats your take on this. I was definitely going with the new Mac Pro...but a little concerned. I am not a computer tech. I know my way around software and the hardware essentials but not a tech guru like some of the people on here...especially Jeff.

Anyone think there is a way this new system could be expandable and handle 6K? ...also work with the new Rocket...or even need a Rocket. I was hoping the next system would not need a Rocket card, at least at 5K.

I think it's going to perform a lot better than people realize. Where it's going to fall short initially is in the expandability department, especially for those who want to replace a current Mac Pro and bring peripherals and storage over to the new box. And the shift in connectivity/ expandability is going to impact a lot of people in terms of overall cost as well as peripheral availability for a while. Peripheral makers are going to need to re-think their approach to their products. The future is not PCIe cards for most devices, but rather devices that can conveniently plug into a simple port (Thunderbolt or similar). Kludgy PCIe to Thunderbolt expander boxes are the middle-man at this point and will eventually need to go away. This is going to take time, but it's something that will ultimately happen. This isn't entirely an Apple concept here, Intel has just as much to do with it and their vision of future computing.

 

[Dustin Cross]

stacked on their side and staggered up like soda cans

Jeff,

I don't think these things are going to be very happy laying on their side. What I have seen so far, they are designed to stand upright and the cooling system won't work correctly laying them on their side.

The biggest problem with Thunderbolt right now is the fragile cable and non-locking connector.

I really hope Apple rolls the new Mac Pro flat, sticks it on the back of a 4K monitor and offers us an iMac Pro. That would be more exciting to me than this cone design.

I am sure performance of the new Mac Pro Cone will be great. My concern is with ergonomics. I don't think it will be happy laying on it's side.



Dusty

 

[Nick Pasquariello]

I think it's going to perform a lot better than people realize. Where it's going to fall short initially is in the expandability department, especially for those who want to replace a current Mac Pro and bring peripherals and storage over to the new box. And the shift in connectivity/ expandability is going to impact a lot of people in terms of overall cost as well as peripheral availability for a while. Peripheral makers are going to need to re-think their approach to their products. The future is not PCIe cards for most devices, but rather devices that can conveniently plug into a simple port (Thunderbolt or similar). Kludgy PCIe to Thunderbolt expander boxes are the middle-man at this point and will eventually need to go away. This is going to take time, but it's something that will ultimately happen. This isn't entirely an Apple concept here, Intel has just as much to do with it and their vision of future computing.

Jeff, I'm really not well versed in the internal protocols of such things: would such Thunderbolt devices easily be swappable to USB3.0? Or similar? Is there a functionality to Thunderbolt that gives it an advantage here other than bandwidth?

 

[Thomas Koch]

What does the current 12 core go for? It will be less than that.

 

[roryhinds]

The new Mac Pro's single CPU, no Nvidia and Thunderbolt 2 bottleneck... all make for someone else bring out a workstation machine for professionals who need heavy iron.

HP are you listening?

 

[paul schefz]

Just came across this http://www.macrumors.com/2013/08/14...ew-mac-pro-posts-impressive-benchmark-scores/
The last Mac Pro I bought was almost 10 years ago, iMacs and Mbps are great but his new Mac Pro might finally be the one to really give me the extra power compared to all other apple offerings....whi is why I think apple has had a Mac Pro in a while.....
I don't need room for built in expansion....I don't even want it....I want a crazy fast CPU with top of the line gpu, and most important I want a version of fcpx tailored to this system.....I expect the new Mac Pro (even a lower end one) with a new fcpx to be about 2-3 times faster then my topoftheline Mbp or iMac....otherwise apple would not bother....
i also think that mavericks will help quite a bit....

 

[Björn Benckert]

I do not know much about difference CPU's and such even though I spent more cash on computers than most. What I learned from that is that pretty much all apple computers can be sold after 3 or even 5 years of use. Normally for about 20% or more of the price paid when new. For PC's that number is pretty much zero. Even the PC's we paid more than 20k USD for is a really tough sale 5 years later if they even held up for that long before having some major fault. Basically the apple machines are much longer in demand and they are really well built, especially the pro boxes.

And even though I think these trashcans looks a bit like toys and does no seem to have much more of a better build than the really early Imac's I still think they will hold up quite well.

Also even more important, I think since they will sell in high volumes I think we will see quite a change into how computers is beefed out. The thunderbolt connectors will be used in ways we can even not imagine today. I have no doubt that there will be big external render engines built in huge boxes, rackmounts for multiple trashcans etc, much more but in a similar manor that we see for the mac mini's today. The size, the silent and low power consumption / low heat is really also something that will make these units great for render farms.

So personally I'm holding back a bit on the computer investments until these things hit the market. Just got my Mlogic TB red rocket encloser today, must say it's really bad as to have the rocket in a little box that allows us to use it on pretty much any computer in the office without dissembling two machines to swap the card... just moving over the TB cable from one can to another in the machine room will be a much more pleasant experience.

I'm also piss tiered on all dual link DVI extension cables that we need today. Apparently there is quite low cost / functional TB to fiber to TB cables coming. For us that's huge as all the longer DVI bridges we tried have random issues when used on long distance together with the large apple cinema displays, dropout pixels and other oddities are more of a standard, so if we can get a nice 4k tb cinema display ready to be ordered at the same time as the trashcans accompanied by high quality low cost TB extension cables then there is not much to complain about.