Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Obviously Apply believes this to be a major upgrade in Pro Technology. I don't see why everyone is arguing over this first benchmark test. Until these are released, I wouldn't put much faith in the first benchmarks.

unfortunately, so far, most of the previous benchmarks (ie rMBP, etc) that leaked prior to release were spot-on in performance. granted they are using OS 10.9 (still don't understand why people use "OS X 10.x") and there may be bugs with the OS, i'd think performance won't be much further than what was posted.
 
At least Apple finally decided to equip Mac Pros with workstation GPUs. That's the only positive thing I have to say about this generation.

The actual choice of GPU is terrible. AMD FirePro only bests the Nvidia Quadro for CAD/CAE. For all other industry-standard applications, Nvidia Quadro reigns in performance. The worst part is that even if I want to connect an Nvidia Quadro, Nvidia Tesla, or Xeon Phi, I have to do it through a Thunderbolt 2.0 Chassis, which is bottlenecked to a measly 4x PCIe 2.0.

Apple is just parading the spec without thinking of the customer, and claiming it's designed for a wide-array of industry apps. The only upside to this is that there are two of the AMD GPUs, but all this really allows for is more, larger screens (inter-GPU communication has never been good). In reality, this is just designed for small- to medium-scale 4k media creators/editors. It reminds me of when Apple paraded the rMBP for CAD despite it not having a mobile workstation GPU.

With two workstation GPUs soldered into an in-house chassis, the pricing is going to be sky-high. At least having only one CPU should bring down costs, but 1866 MT/s ECC RAM isn't cheap, and Apple already overprices RAM as it is. Not to mention the worst component of all – the PCIe SSD. eMLC NAND costs a small fortune, and if it isn't eMLC, every customer with a heavy workload will avoid it like the plague.

The cost would have been excusable if the performance was decent, but with 1.25 GBps Sequential Read and 1.0 GBps Sequential Write, it sits at the low-end of the PCIe SSD market. Even my MLC SSD RAID 0 array pulls around 1.8 GBps for both Sequential Read and Sequential Write. At least Apple provides Thunderbolt 2.0 ports, that's the saving grace for storage (provided they implement the controller correctly for heavy load).

And finally, there's the "innovative" cooling solution. With vents placed all around the circumference of the chassis, how is one supposed to position this machine without obstructing airflow? Does the single large fan on the top of the chassis not scream "SINGLE POINT OF FAILURE" to anyone? (Even high-end fans have a lifespan of 50,000 hrs.) What happens to temps when both GPUs and CPU are under heavy load?

It's ultimately a niche product for an ever-dwindling niche market of media professionals.
 
I'm not as pessimistic as you. Dual Xeon and it would be neatly perfect.

Costs: Probably 6-12k. Competive compared to HP and Dell. WORKSTATIONS are not cheap. This is not a home personal computer.
 
Why do they build this machine without dual CPU option?

The most likely scenario is that dual processor Mac Pro sales were so small that it wasn't worth doing.

All of this is related to sales numbers and a lack of growth in the workstation market. Mac Pro sales were small - tiny even compared to Apple's other lines. Tiny sales on systems that last 5-6 years from a demanding customer base. Not worth really focusing on for a company now in Apple's position. We heard rumours that there were internal discussions they wanted to cancel it, the 2012 "update" shows it was on the fence before that.

So now we have a solution that to me looks like it focuses on the majority of their old market, aims to capture new users with its differences/focus, and is likely a lot cheaper to produce. Losing 30,000 sales a year doesn't much matter if you make more per system and manage to grow the market for it beyond what you lose.
 
The most likely scenario is that dual processor Mac Pro sales were so small that it wasn't worth doing.

Mac Pro sales were small - tiny even compared to Apple's other lines. Tiny sales on systems that last 5-6 years from a demanding customer base. Not worth really focusing on for a company now in Apple's position. We heard rumours that there were internal discussions they wanted to cancel it, the 2012 "update" shows it was on the fence before that.

So now we have a solution that to me looks like it focuses on the majority of their old market, aims to capture new users with its differences/focus, and is likely a lot cheaper to produce. Losing 30,000 sales a year doesn't much matter if you make more per system and manage to grow the market for it beyond what you lose.

Pretty much nailed it on the head with all the cynical posts here.
 
Imagine if they told the truth.....,

"We at Apple have gotten filthy, stinking rich selling phones. Hundred dollar bills fall on us like rain in a rain forest. There is nothing we abhor more than a product that can be changed by our customers. This is especially true when this meddlesome fiddling results in greater value that lengthens a product's useful life. Thus we set out to turn our flagship product into something more fitting for our purposes. At first we just made it into a tiny box but Phil said he'd feel foolish walking on stage with it. So we needed a gimmick. We needed to make removing all those capabilities into an asset. Phil wadded up the plans for the tiny box and threw them into the trash. And we all stared. At this great moment Jony picked up a small fan on a nearby desk and set it on top of the trash can. "I think we've found our gimmick gentlemen !!!"

And thus the decontented Pro was born. We were so pleased we immediately voted to give each other 1 million more shares of stock and sent the can & fan to engineering to flesh out. Then we moved on to planning the next company picnic"

Classic.

Well put, and probably much closer to the truth than you'd think.
 
Last edited:
The 3.33 GHz could be mine. It got the same score. I upgraded it from an 8-core 2010 Mac and placed 96 GB of ram on 6-slots to get proper three-channel performance. The graphics card is a GTX580 with 3GB ram and DisplayPort. Samsung 512 GB as system disk and 8TB internal raid-0. It's a good finishing and editing station. When the Red Rocket X arrive, we might by a Cyclone PCI-expansion and put it in there with two additional GPU's. Unfortunately, the new MacPro will probably not be able to compete with this set-up...
 
Geekbench is just such a useless benchmark.

Says nothing about x264 performance, 7-zip performance, OpenGL performance, OpenCL performance, LLVM performance and so on.

True,Until there are some tests running actual programs its not worth getting in a tizzy about , when i see if Photoshop for example has a dramatic performance jump then i will start to take notice.
 
Geekbench is just such a useless benchmark.

It's comparing raw CPU scores, which isn't entirely useless, just rather generic and vague.

Like if your current computer scores considerably better in Geekbench than your previous one, you can expect CPU intensive applications to perform better on it across the board. You just won't know how much better, exactly.
 
Well if you look at the existing prices for the Sandy Bridge EP 8 Core processors (v1), the prices range from $1000 to $2100 before the typical 33% Apple processor premium. You would have to expect that the v2 versions would be roughly the same. So you are looking at $1333 to $2793 for the processor alone. Apple currently charges $300 for a 256GB upgrade in the Macbook Air. This is a lower performing version of the drive so expect another 33% on top of that upgrade price ($400). 2GB FirePro W5000 cards run about $450 each ($600 with the Apple markup). 32GB of 1866 Registered ECC DDR3 runs about $400 at crucial (Apple will easily charge $800 for this).

So with to bottom 8-core processor, I would estimate the system would cost:
Processor - $1333
256GB M.2 SSD - $300
2xFirePro W5000 - $1200
32GB RAM - $800
Other components - $750 (case/power supply/MB/TB controllers/fan/etc)

Total ~$4400

Unless Apple is going to sacrifice its existing margins, 8-core machines are going to be expensive as well. Get your RAM someplace else and bring it down to an even $4k.

GL

THE PAIN!!!! :(

True,Until there are some tests running actual programs its not worth getting in a tizzy about , when i see if Photoshop for example has a dramatic performance jump then i will start to take notice.

Each program is different, and handles tasks differently. For instance an NLE won't tax the CPU until you want to render our a sequence or playback more than one layer of uncompressed footage.

Photoshop is a rather light program, and the biggest boon to the application will be the amount of RAM in the system.
 
Congrats on that 6-core that runs fast as hell-fire, but is as cool as ice, and looks fantastic.

Thanks. I don't know if I got the record for that 6 Core, but I'm sure you found a way to top it. Again, thanks... :)

----------

That is pretty sweet and from what reading the link it took the guy months to build it. I can only imagine how much more time he took to assess which parts to buy and the time spent going to get the parts. I get it, all part of the 'FUN' into building and continually tweak and make it better or maintain it.

And of course it is going to be cheaper. I build PCs for personal use when I was younger but what professional have time to pick parts, build the computer and support it? It is fair to say from reading this thread Pro-Sumer thinks they are Professional and making a lot of noise. Further more, compared to the old MP top bench mark, I think CPU spec wise it is a fair increase.

You don't have to build what I have (as what I built took many hours and weeks), but you can just go onto TonyMacX86.com and they have basic system setups that would take a few hours to build. Here's the link and it's pretty much spelled out for you there:

http://www.tonymacx86.com/365-building-customac-buyer-s-guide-june-2013.html

Lastly, you can just look at my builds and use them (the homework is done for you). Plus, you don't even need to water cool them; just get a decent fan cooler and just OC it a bit and you'll be just as fast.

Enjoy your next hackintosh build, later... :cool:
 
Last edited:
Thanks. I don't know if I got the record for that 6 Core, but I'm sure you found a way to top it. Again, thanks... :)

My 6-core Geekbench 2 score still tops out at 23,317. You're the 6-core champ. You've earned that record. Moreover, thank you for re-igniting my interest in Nvidia video cards, helping to spur my interest in learning about CUDA and CUDA programming, and for showing me that a self-build can not only be fast, but also look magnificent.
 
really?

When i need to render fluids or particle animations in Cinema 4d. Pure CPU power tops GPU. That means the improvements aren't really there.

Dude, Cinema 4D doesn't have fluids. There is only one plugin to add that, but only gaseous fluids and seriously underpowered... If you however really do fluids simulations, then you have Real Flow and Houdini FX for that. Don't know about RealFlow lately, but HoudiniFX can use OpenCL for PyroFX and FLIP fluids simulations. So it looks to me like you are more an amateur not a real pro since you don't use big boys guns.... And you don't know what is the trend in computing, witch is to use the graphics computational power for heavy duty tasks. New MacPro looks like a champ to me in this areas. And if you are really, really, really a pro then you have some form of render farm to your disposal. ;)
 
My 6-core Geekbench 2 score still tops out at 23,317. You're the 6-core champ. You've earned that record. Moreover, thank you for re-igniting my interest in Nvidia video cards, helping to spur my interest in learning about CUDA and CUDA programming, and for showing me that a self-build can not only be fast, but also look magnificent.

I've never had the cash for a full workstation upgrade to my home studio since the PowerMac G5. Things are however picking up for me and I should be looking to replace a well received iMac in another year or two. When the time comes, I want to drop about $7k on a full workstation.

I think my main two options are going to be an HP/Dell dual socket system, or a custom built PC. Maybe I'll throw in the third, iMac option, but even that machine went down in terms of using it as a workstation IMO.
 
Really....?....

I was hoping for a 30K benchmark, at minimum. Not so impressive. But again, the new Pro is not having a much better CPU that the actual one. Better CPU architecture, yes. Maybe mixed with GPU and Thunderbolt peripherals or in real world usage, the benchmarks showing right now becomes not significant....:D


:):apple:
 
I didn't think you had a point :) People were saying that based on this early benchmark, the brand-new, super-duper, kick-ass Mac Pro was only 10% more powerful that the model that came out 3(?) years ago. You just seemed to reiterate that it was only 1 CPU and not 2, as though that was relevant.

If you were alluding to something deeper, then you didn't mention it.

You are comparing 1 CPU which still beats 2 previous gen CPU's. Also I'm hearing the new intel chipset may not support dual cpu's, at least initially. I'd say that's relevant.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.