Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

New Mac Pro Pricing

If the following is correct:

12 Core CPU (2.7 GHz) with 11 core running flat out = same speed as 10 Core CPU (3.0 GHz) with 10 cores running flat out.

12 Core (2.7 GHz) with 10 or less cores running flat out = 11% slower than 10 Core CPU (3.0 GHz) with same num of cores running flat out.

10 Core CPU = $600 cheaper than 12 Core CPU.

I think Intel $$$ per core pricing makes no sense here (assuming the leaked info is correct).

I hope that Apple pushed for the 'top bin' 12 core chips - that work at 3GHz (even if they burn 150 Watts)

If Intel are making 60% profit on the 10 core 1.7GHz Xeon E5-2650L v2 ($1395) ...
...what sort of profit margin at they making on the 10 core 3.0GHz Xeon E5-2690L v2 ($2355) ... 75%?

I wonder what will happen if in the back end of 2014, a 16 core x 3GHz A57 comes out for less than $500.

Maybe it will be Nvidia in 2015 that break Intel's compute stranglehold on the PC market? (With their Denver V8 Core?).

AJ

UPDATE : Just found there is a Turbo feature that if thermals allow:
1 x core @ 3.5Ghz
2 x core @ 3.4Ghz
3 x core @ 3.3Ghz
4 x core @ 3.2Ghz
5 x core @ 3.1Ghz
6-12 x core @ 3Ghz

Reference: http://www.tomshardware.com/reviews/ivy-bridge-ep-xeon-e5-2697-v2-benchmarks,3585-2.html

If Apple and AMD can resolve the Open CL issues - this sounds like a winning combination ... now the question really is RR-X ... or higer speced out Mac-Pro (or screaming external disc array)
 
Last edited:
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.

As long as the air can move through them properly, and in adequate volume, orientation is irrelevant. I know of a big player testing them in various orientations right now... Just sayin'.

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

Build quality is getting better with some of the newer components. There are locking displayport and mini-displayport connectors in the displayport spec. It is curious that Intel and Apple are not using them.

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'd like one of those too. But still need a stand-alone system.

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.

As with any high-density computing setup, airflow must be carefully managed. There are some distinct advantages to this form factor over a flat 1U form factor or blade arrangement.
 
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.

When I say Hackintosh, I don't mean only i7 platform. I was also considering going the dual Xeon route, and still could switch to that at some point. The problem is it would cost $10,000 compared to the $3,000 I spent for the i7 configuration, as I would have gone with the 2687W (x2) and at least 64GB RAM and would have needed 2 GPU's instead of 1 (no integrated GPU on Xeon) for my multiple displays. Right now (and with E5 v2) the upgrade would cost about $2,000 per CPU, $600 for the motherboard, $1,600 for 128GB RAM, and another $900 for another Titan, so up to $7,000 more to essentially double performance in only applications that can scale to that many cores and that many threads. It just didn't seem worth it right now with new CPU lines coming out, the switch to 4K displays, and the unknown of what we'll really need to handle 6K R3D's. For all we know an i7-based desktop will be enough to handle it.

Bottom line, a dual Xeon hackintosh is entirely possible. They do exist. There aren't as many and they probably aren't as easy to set up, but they are out there and they can be built. Here's links to some....

http://www.tonymacx86.com/retail-cases/43116-ultimate-evga-sr-2-hackintosh-hackinbeast.html
http://www.tonymacx86.com/mountain-...ram-nvidia-quadro-4000-geekbench-31496-a.html
http://www.tonymacx86.com/desktop-c...triple-7970-x-fire-nvidia-quadro-fx-w00t.html
http://www.tonymacx86.com/user-buil...e-w8-ds-2x-xeon-e5-2660-gtx-590-3gb-64gb.html
http://www.tonymacx86.com/user-buil...-project-ga-79x-up5-wifi-intel-proc-e-ep.html
http://www.tonymacx86.com/user-buil...2-thr-z9pe-d8-ws-12tb-geekbench-28-000-a.html
http://www.tonymacx86.com/user-buil...am-30tb-fiber-storage-os-x-10-8-3-server.html

That said, the premium cost for an Apple computer isn't a bad thing necessarily. One thing I learned from building a Hackintosh is there's no substitute for Apple's build-quality. Nothing you can build will compare to it.
 
and the unknown of what we'll really need to handle 6K R3D's. For all we know an i7-based desktop will be enough to handle it.

according to mark toia he played 6k dragon files on his mbp in rcx at 1/4 and he also mentioned that he shot at 17:1 which looked like 8:1 (compared to epic) ....so it seems like the cleaner sensor makes compression and with that the handling of the 6k raw files actually easier then the 5k files of epic.....i don't cliam to understand why, but if half of what he is saying is correct and we can expect even better performance after the hardware and rcx are tweaked and ready for the world, i am very optimistic about maybe even getting speed increases with our existing systems.....
when the mac pro was announced i thought i will definitely get one....now i could imagine that a new mbp (maybe even 17"?) with the latest chips, 32ram, mavericks and new versions of rcx and fcpx might double the performance of my existing top of the line mbp (when comparing epic 5k now and dragon 6k workflows) and that will probably put the mac pro on the back burner for me again....
i looked into hackintosh for the bang for the buck....but the time/hassle just isn't worth it to me.....and apple always shines when it comes to tweaking software to run best on their hardware....
 
I didn't see Mark's post about that. I mainly looked at the video and was satisfied. That's really good news.

I really think my i7 Hackintosh will be enough. This thing just screams. Resolve runs so much smoother on it and I've gotten renders as high as 100fps with it, but usually around 60-70fps, when working with non-R3D material (such as ProRes4444). I haven't tested R3D's yet, but might do that later today. I know I get real-time transcoding from my rocket with 5K to ProRes4444. FCPX starts up in 2 seconds. Premiere Pro CC in about 5 seconds. Avid and FCP7 also in about 5 seconds. My Cinebench scores are Open GL 49.27fps and 8.89 pts CPU. Geekbench 32-bit has been between 15,500 and 16,700 (I should just buy the 64bit version)

An 8-core Dual Xeon will get only 10-12 pts on Cinebench and I think around 21,000-22,000 tops (and I usually see around 17,000 for even the 2687W) on Geekbench, so I'm getting at least 2/3 the performance at a fraction of the cost. I'd rather build another one of these and farm them together if I really needed more processing power for rendering. Anyway, this Hackintosh turned out better than I thought it would and I couldn't be happier with it.

Intel should really consider changing the i7 platform into a multiprocessor variation. Imagine all the people that aren't willing to spend so much on a dual Xeon configuration, but would be willing to throw down $200-$400 more on a 2nd CPU. And they could even have an i7 Pro line (or change the Extreme line to dual processor) and charge more for them, somewhere between i7 and Xeon, no ECC RAM requirement, basically a stripped down Xeon. I think most of the people that feel they need the Xeon will still buy the Xeon, so they'd only be increasing their revenue with more dual CPU systems. If it was only maybe $2000 more it would be no question, but $7,000 more to go dual Xeon is a huge leap.
 
When I say Hackintosh, I don't mean only i7 platform. I was also considering going the dual Xeon route, and still could switch to that at some point. The problem is it would cost $10,000 compared to the $3,000 I spent for the i7 configuration

Exactly. That was part of my point.

Bottom line, a dual Xeon hackintosh is entirely possible. They do exist. There aren't as many and they probably aren't as easy to set up, but they are out there and they can be built. Here's links to some....

Yep, I've built a couple myself.

That said, the premium cost for an Apple computer isn't a bad thing necessarily. One thing I learned from building a Hackintosh is there's no substitute for Apple's build-quality. Nothing you can build will compare to it.

Indeed. And I have also found that they are typically priced right in line with similar offerings from HP or other PC vendors and still built nicer. They also hold their resale value a lot better than the PC's do and have the best warranty coverage in the biz.
 
An 8-core Dual Xeon will get only 10-12 pts on Cinebench

Cinebench is a worthless benchmark, IMO. It is a poor test of the latest E5-26xx Xeons because it's such a quick and dirty test and it actually doesn't throttle-up the CPUs fully in the render testing, so you really don't know what you're getting. The OpenGL test is also a couple generations behind so some older cards actually test out better than some newer ones which outperform them in the real world.

Sorry, I'm just picking nits. On an E5-2687W system or similar configuration, you can actually get better scores in Cinebench by running the system under load to throttle the CPUs (or by throttling them all the way up by default gets you most of the way there) and then pull the plug on whatever else you were doing right as you're jumping into the Cinebench test. Stupid trick, but it works and shows the inadequacies of Cinebench as a useful benchmark.

As for the i7 as a multiprocessor platform, that's effectively what the E5 series Xeons are. Yes, they're hugely more expensive, but they also offer a lot of enhancements -- more cores, onboard PCIe host and support for more lanes, more robust memory controller with greater throughput. I guess if you want to step back from the expensive E5 options, there are some lower cost solutions and some of the E3 chips are multiprocessor capable. Not a huge market for that, though. I think when it comes to multiprocessor systems and packing a lot of cores into one box, those who truly need it are willing to spend the money. Otherwise they fall back to a 6 core i7 desktop.

It's really too bad that Apple doesn't offer an i7 desktop system. IMO, they should keep the full size tower with dual Xeon option alongside the new Mac Pro. They should also launch this new Mac Pro with the ability to not only install the Xeon CPUs, but also an i7 desktop CPU. For as modular as this thing is internally, it wouldn't be a huge deal. Some people out there are speculating that this may happen. I think it would be a welcome option and would provide at least a config or two around $1000 or under $1500.

Although, I don't want to get anyones' hopes up. I'm skeptical of such reports. Apple would have a tough time with such offerings and managing to keep a big enough market for the "pro" users. Half or more of the people who will buy the new Mac Pro could do just fine with an i7 desktop configuration.
 
Last edited:
The point worth considering is this. With higher 17:1 compression with Dragon the question is it easier or more difficult to decompress these higher compressed images, than say 8:1 images? Yes, the files may get smaller, but it doesn't automatically mean better real life performance.
 
I was not aware of the limitations of Cinebench. I'm still happy with the performance I get on my system. I just ran the 64-bit Geekbench and got a 17,067 score with a 4.4GHz. I might change the overclock back to 4.6GHz at some point to test that again. I've already saved so much money on this vs buying a current Mac Pro (and I have much more capability), and it will pay for itself many times over, that I'm not worried about resale value.

Right now I would rather have a higher clock than a bunch of cores. I don't think any of the normal apps we use in this business really take advantage of all the cores, but they all do benefit from higher clock speeds. The only time I ever see all my cores getting used is when I render in Compressor with a quickcluster. Other than that I rarely see all the cores getting used at their potential.

I would be very surprised if Apple threw an i7 in the new Mac Pro as an option. Even if they did I still don't think I'd be interested in being forced to use external expansion. I'd need at least 4 expansion chassis for my cards (Rocket, Decklink or just go Ultrastudio, Atto H680, and Atto R680), plus the Thunderbolt cables, and additional UPS or mess of AC cord splitters. Also, there's no Thunderbolt 2 expansion chassis out right now, so if you want to use it right away you're using slower TB1 expansion chassis now and will need to upgrade to TB2 chassis later. There's no guarantee current models can be upgraded to TB2. Sonnet won't even upgrade the 1st version of their Echo Express Pro with the extra power that the 2nd version had, so I don't see them offering an upgrade program for any of theirs. And then you have TB2 cables as well.

So you're looking at having a Mac Pro cylinder, 4 expansion chassis, a bunch of external hard drives, Thunderbolt cables, power cords, eSATA and/or SAS cables, mini DisplayPort cables for monitors, all going all over the place. Add any USB cables you have for wired keyboard and mouse, control surfaces, webcam, portable hard drives, or whatever else you might want to connect, and it's going to be a complete mess for anyone who wants to have a fully loaded system.

The only thing I think that could alleviate that issue is a single chassis with multiple slots and Thunderbolt ports. However, you'd still have the issue of limited lanes per TB cable. We don't know for sure how scalable the TB ports on the Mac Pro will be and if they can really be combined in some way to increase bandwidth for PCIe cards.

There's also the limitation of the GPU being offered and possible inability of upgrading the GPU at a later date.

It just seems poorly thought out as a replacement for the current Mac Pro. I think Apple is aiming for a different customer than the current Mac Pro is targeted at. That's fine, but there's still a lot of people that need something like the current Mac Pro.
 
The current Mac Pro has been replaced by the current full spec'd Imac .

I know because I was forced into it by the death of my Mac Pro ( dammit! )........ cuts 5k footage fine.

Which could be a problem for the new Mac Pro if some of the prices been kicked around on this thread are real.
 
The point worth considering is this. With higher 17:1 compression with Dragon the question is it easier or more difficult to decompress these higher compressed images, than say 8:1 images? Yes, the files may get smaller, but it doesn't automatically mean better real life performance.

In general, higher compression = higher load on CPU.

I bet we'll see a lot less 17:1 footage than you think, though. I constantly have to argue people away from shooting 3:1. "are you suuuuuuuure that 8:1 is good enough for this?" "Yes. Hobbit was shot at 6:1. And released in 4K. This is going to 1080p on the web and should probably be 10:1 but we'll compromise." "I dunno . . . it could really impact the quality . . . "
 
In general, higher compression = higher load on CPU.

"Yes. Hobbit was shot at 6:1. And released in 4K.

Ah...never thought of it that way.

On another note, I was under the impression that Hobbit was finished in 2k only.
 
I purchased a Mac pro in 2008. It was a 2.8ghz. I thought about upgrading the CPUs to the 3.2 ghz version. It would have cost more to flat out buy the CPUS than buy the 3.2Ghz Mac Pro all together. The cost of these computers will cheaper than buying the components yourself at retail prices.
 
I purchased a Mac pro in 2008. It was a 2.8ghz. I thought about upgrading the CPUs to the 3.2 ghz version. It would have cost more to flat out buy the CPUS than buy the 3.2Ghz Mac Pro all together. The cost of these computers will cheaper than buying the components yourself at retail prices.

We shall see and there's a good bet you're right about that. Apple occasionally works some real magic with component pricing when they standardize their systems. Also related to the 2008 Mac Pro, they were the first to offer the 3 and 3.2 GHz Xeons -- they had a 45 day exclusive on the chips ahead of HP and everyone else. Not only that, but the 3.2GHz Mac Pro when released in '08 with 16GB RAM, 500GB HDD and NVidia 7800GT 512MB video card was priced $1600 less than what HP was selling their equivalent configuration for. Even with their famous 20% coupons, the HP was still more than $1K more than the Mac. On the flip side, the price of the HP dropped over the coming months and Apple held their price.

I'm expecting the entry-level Mac Pro to be cheaper than the entry-level pricing of the current unit. As we climb up the configuration ladder, however, it's going to get expensive rather quick as these systems always do. I think a lot of the pricing speculation is just that at this point. All we can do is wait and see what happens, Intel is about to start shipping out these Xeon chips. We're probably only 40 to 60 days away from that new Mac Pro. Or at least I hope that's all... Assuming all the other components are ready for mass production.
 
Yeah, I think the Mac Pro that Apple showed was the top of the line model to wow everyone. The entry level probably won't be nearly as powerful.

I suspect the entry level Mac Pro will be 6 core, 16GB of RAM, 128GB SSD, Dual Radeons and be $2999.99

The one they showed is probably more like $4999.99 - $5999.99
 
I paid $10 -12K for my 2008 Mac Pro V2 new.

I suspect with current Apple pricing strategies & how it will sit within their existing product range it will be cheaper than my original 2008 purchase....alot cheaper. So I'm with Jarek.

Of course I bow to the great tech. minds posting here. But at the end of the day given what I'm doing now on an Imac, do I need much more ? If dataloads are shrinking because of a cleaner sensor etc, etc (or maybe not according to Jake ) than maybe I'm better sitting with a fully spec'd Imac & upgrading it every year.

Either way it is good to see Apple back in the game of producing quality workstations & lappys for professional editing/vision/audio/vfx applications. Which given where we were at the beginning of the year with the belief by some that "Apple is abandoning the pro market" is great news !
 
Last edited:
If dataloads are shrinking because of a cleaner sensor etc, etc (or maybe not according to Jake ) than maybe I'm better sitting with a fully spec'd Imac & upgrading it every year.
If data loads indeed shrinks with Dragon, then why do we need more powerful RR? Clearly, data loads are going up:-)
 
I was so sure I wanted to wait for the new Mac Pros until today. I Just got offered a Xeon 16 Cores 2.4ghz w/ Asus Mobo for $1000. Now if the base model mac pro exceeds $2500 with a hexacore and lower tiered graphics, Then the choice would be obvious. What would you guys do since this offer won't last for me?
 
Back
Top