Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
New 12-Core Mac Pro Once Again Shows Up in Benchmarks
- Thread starter MacRumors
- Start date
- Sort by reaction score
Yeah...I don't really get this site...
It's called Macrumors but there certainly are a lot of shills and haters on here!
Yeah...I don't really get this site...
It's called Macrumors but there certainly are a lot of shills and haters on here!
Yeah, it's almost like the site is made up of a diverse compilation of nerds, professionals, and people who think they are nerds or professionals, all who have varying levels of education, experience, and maturity.
You know, unlike any of those other websites.
Looks like a Dalek
I like it, looks great and powerful. Just the price tag makes it unlikely for me with 12 cores. 4 or 6 might be justifiable as hobby system.
I like it, looks great and powerful. Just the price tag makes it unlikely for me with 12 cores. 4 or 6 might be justifiable as hobby system.
Ridiculously powerful desktop.
Audio engineers only need the base line models. Video the mid model. Animation and number crunching the high end model.
Audio engineers only need the base line models. Video the mid model. Animation and number crunching the high end model.
Usually people complain that that intended audience is smaller than it should be. You know, people who want a fast yet sleek Mac for professional use but don't care about PCI slots, upgradeable GPUs and CPUs, DVD capability, and extra hard drive slots.
A really really cool machine, but I'm worried that pros will have to switch to Hackintosh or worse. Actually I wouldn't care if they switched to Hackintosh, but I'm sick of Windows being so popular.
----------
Umm no it's not, ironically. Or else "FTP stands for future teleportation protocol" would be hate.
----------
So the bottom line is, even with all this spiffy new hardware, the new 12-core model is not even 18% faster (29721 vs. 25208) than what is an essentially an upgraded Mid-2010 Mac Pro (a.k.a. "Mid-2012") model?
I don't know what else a new computer can be. There hasn't been any recent leap in CPU technology.
----------
It would be interesting to see. The best bang for the buck overall would be a low-end desktop Dell, but I'd like to look at the high-end stuff too. The problem is that GeekBench tests only CPU and RAM, not disk speed or GPU.
----------
Yeah...I don't really get this site...
It's called Macrumors but there certainly are a lot of shills and haters on here!
You're calling a bunch of long-time Mac users shills. What would you call lilo777 once you see him?
That's my problem with it too. The results are essentially meaningless except as a relative measure between models.
What I'd like to see is a test built around free/open source Apps, something like Handbrake as far as video is concerned or the equivalents in photo/audio processing. Something like Barefeats does. This way a 'Pro' machine can be compared to a consumer unit as it's doing an identical tangible task.
A Geekbench score of 20000 tells me that a unit with a score of 25000 is likely to be faster but means nothing to me regarding the Apps I might use.
That's exactly what a benchmark is
I have a feeling that this latest Mac Pro will be HEAVILY tested by every hardware website around. You'll see the results soon enough... across a variety of tasks.
That's exactly what a benchmark is
Good point but some benchmarks can be useless, which is exactly opposite to their Raison d'etre no?.
I agree... some benchmarks can be useless.
Is GeekBench one of them though?
You're right about one thing... comparing between models. I see nothing wrong with using a particular benchmark to compare the old Mac Pro with the new Mac Pro.
That should offer some insight.
Actually, the Geekbench scores are pretty close to real world results. For example, I had a Mac Pro 1,1 @ 2.66ghz Quad core which scored about 5,000. I sold it and bought a 2011 MBP @ 2.3ghz Quad core which scores roughly 10,000.
I did some comparisons in Handbrake compressions, photoshop benchmarks and they weren't far off from the Geekbench scores, all tests we're pretty close to twice the power difference.
That will be the final word on performance.
A benchmark can tell you particular details, the complete number being a combination of various abilities in an indiscriminate total, but a breakdown of speed during specific functions will help define the true user-end performance.
If the hardware and OS are seriously redesigned to aid efficacy, the benchmarks between new and old Macs won't be accurate as a buying guide.
Agreed.
I would hope that people use both benchmarks and real-world tests to make the final decision
As a software developer I'm a little bit disappointed about the new Mac Pro. It seems to be aimed only at people who do heavy graphical stuff and not so much at other professionals who need a fast workstation.
For me the Mac Mini seems to be the best choice, but I would love to see something in between the Mac Mini and Mac Pro that isn't an all-in-one solution like the iMac. I guess I'm not the only one who needs/wants this...
For me the Mac Mini seems to be the best choice, but I would love to see something in between the Mac Mini and Mac Pro that isn't an all-in-one solution like the iMac. I guess I'm not the only one who needs/wants this...
Ok, I'll use a real world example from my field - this actually happens regularly. I'm in the standby power industry. So one of my jobs during a day may be to go to a customer site and impedance test his battery. His UPS may have one string of 32, 100Ah batteries and lets say that this battery provides him with 45 minutes of backup when the utility fails.
So.....I look at the battery manufacturers data sheet and, (this next part of text is deliberate to illustrate my point to people in the real world as ooposed to those who know), it tells me that the spec for that battery is as above 45 mins down to 1.6vpc and that the nominal impedance is 4mΩ.
I test his batteries and find that 90% of them range between 3.8mΩ and 4.2mΩ and are deemed to be in very good condition. 8% of them range between 5mΩ and 6mΩ and are deemed to be Ok, but to be monitored.
The other 2% of them are between 6mΩ and 7mΩ and are deemed to be ones that need replacement. There are no definite open circuit batteries.
All this goes on a glossy report that I tell him will be emailed to him, but before I leaves site he says to me, 'Ok so how long will my battery last and which blocks will fail?'
I don't know and cannot tell from the figures. I can guess from experience that ones in the 7mΩ range may fail first.
He says 'Oh. Well how can I tell then?'
The answer to which is a real world test of dropping the mains and letting run on battery whilst monitoring all battery blocks. Moral of the story as I'm sure you've already guessed is that the benchmark has meaning but almost none to the person who is actually using the equipment.
All of this looked like Greek to me, lol. I didn't understand much.
The benchmarks arent 100% accurate. Just take a look at some of the results and you'll be able to see individuals getting different results with the exact same configuration. It varies.. but its a good way to get a general idea of what to expect.
You've just proven my point beautifully. The benchmark in this case is the manufacturers 4mΩ figure for impedance. (It's actually more simple than you think). But it means nothing to the end user in his day to day actvities although he needs it to.
The Geekbench figures tell me nothing about how long it might take to process a 6Gb video file which is something a lot of people do regularly and would be more able to understand.
I'm not proving your point because I'm 'wrong' on my argument, its because I've no clue about the technical jargon you used.
Again, I disagree.The Mac Pro I had, received about half the score my current MacBook Pro has which is inline with Handbrake encoding; files are now encoded twice as fast (roughly). Nobody is going to take Geekbench as gospel (I would hope, anyway) but its still a great way to ROUGHLY measure processing speeds.
That describes me. So after happily using Mac Pros for many years, I've asked my employer for a workstation running linux. Dual 8-core Intel® Xeon® Processors E5-2650 v2 and a single AMD FirePro V3900. About $4800.
The problem is that in a lot of cases the user won't notice a lot of difference in speed due to the fact that a lot of operations in programs are single threaded.
So the speed of an iMac is what people will get most of the times.
Sure if for example you want to render a 3d animation it will render faster than the iMac due to more cores but what happens until you reach that final rendering is another story.
It would be nicer if the processor had a higher clock speed. On top of the multiple cores I mean. But since there's no room to move this much heat around we're stuck with lower clock rates.
I can always render the final sequence on a render farm. How fast the machine is in getting in to the final render is more important to me.
It's an ok machine but there are so many compromises and in the end all the negatives are starting to stack up.
So the speed of an iMac is what people will get most of the times.
Sure if for example you want to render a 3d animation it will render faster than the iMac due to more cores but what happens until you reach that final rendering is another story.
It would be nicer if the processor had a higher clock speed. On top of the multiple cores I mean. But since there's no room to move this much heat around we're stuck with lower clock rates.
I can always render the final sequence on a render farm. How fast the machine is in getting in to the final render is more important to me.
It's an ok machine but there are so many compromises and in the end all the negatives are starting to stack up.
I'd take a system like that any day. Except that price is... out of my range. I value CPU speed far more than GPU. Why? Because Apple has made it perfectly clear they dont give a crap about proper GPUs. The only GPU need I've ever had was for gaming, and because of the limitations for so many years, I've pretty much abandoned all gaming on my computer.. which has now left me not giving a damn about the GPU. My work relies on Photoshop, Illustrator, Indesign and some After Effects. Not much of a need for two FirePro video cards or a FirePro GPU of any kind. An AMD 270X for roughly $200-$250 would be perfectly fine. Less heat too from a single card.
It should be mentioned that the FirePro D300, D500 and D700 all equate to roughly the W7000, W8000 and W9000 which, on Newegg.com each cost $700, $1,400 and $3,400 as a SINGLE card. Apple puts in two of each. Needless overkill.
Last edited:
Keyboard and mouse sold separately.