True, when I use my computer for word processing, browsing etc, it only uses 1-5% of CPU. But processing music, video or graphics pushes it to the edge, and I am not a patient person. I'm looking forward to this upgrade.
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 iMac Benchmarks Show 10-25% Improvement Over Previous Generation
- Thread starter MacRumors
- Start date
- Sort by reaction score
True, when I use my computer for word processing, browsing etc, it only uses 1-5% of CPU. But processing music, video or graphics pushes it to the edge, and I am not a patient person. I'm looking forward to this upgrade.
The Mac Mini is the better buy, buy far. Even if you add a SSD, 16gb Ram, and a display, or even a Thunderbolt display. The mini is a better buy hands down.
The imac gpu is better on the imacs. But CPU wise the Mini's are as fast.
http://browser.primatelabs.com/geekbench2/1226743
The imac gpu is better on the imacs. But CPU wise the Mini's are as fast.
http://browser.primatelabs.com/geekbench2/1226743
The iMac uses desktop CPUs.Or even an inch thicker.
You know, so that form doesn't cripple function for twice the price.
It's a freakin' desktop. Limp wristed pansies doen't have to carry it around.
What components?
The fastest desktop CPU and GPUs available instead of gimped mobile junk
that gets too hot and throttles down 50% under heavy loads.
2-3X faster and less expensive.
That was easy. Magic.
The iMac has had mobile GPUs for quite some time, so it's unreasonable to expect them to have even fit a desktop GPU in to the previous model.
Do you have personal experience or testing with this model getting hot and throttling down 50%?
Where the hell is everyone getting these percentages from??
I look forward to reading all of the "helpful advice" that we will see in the months to come because of this rather misguided article.
----------
NO, they are not. *beats head into the wall*
It's already started.
Geekbench shows a similar score and to you it means that it's just as fast for all users? I give up.
Not Necessarily Faster
The fastest new 21 inch iMac (2012) is not necessarily faster then the previous fastest iMac from mid 2011.
I have a mid 2011 iMac with the Intel Core 7 (second iMac listed on the graphic) and I just ran the current Geekbench test and received a score of 12800.
To compare machines over time you have to re-test the old machines with the current software. Both OS X Mountain Lion (10.8) and a newer version of Geekbench have come out since the iMac mid 2011 was first released and tested. My results were slower on Lion (10.7) and the older version of Geekbench.
The fastest new 21 inch iMac (2012) is not necessarily faster then the previous fastest iMac from mid 2011.
I have a mid 2011 iMac with the Intel Core 7 (second iMac listed on the graphic) and I just ran the current Geekbench test and received a score of 12800.
To compare machines over time you have to re-test the old machines with the current software. Both OS X Mountain Lion (10.8) and a newer version of Geekbench have come out since the iMac mid 2011 was first released and tested. My results were slower on Lion (10.7) and the older version of Geekbench.
NO, they are not. *beats head into the wall*
It's already started.
Geekbench shows a similar score and to you it means that it's just as fast for all users? I give up.
This reminds me of the people who wonder why ram or an ssd didn't increase their geekbench scores. Someone should explain the exaggerated impact of hyperthreading on benchmarks relative to actual use.
Year to year they don't normally have a huge jump in CPU speed but for someone like me using a 2005 G5 Tower this will be a huge improvement, as it will be for anyone with an iMac more than 3 years old.
Can't wait to see the test for the 27" with i7, Fusion and top video card.
Suddenly my 'pro' G5 is looking so much older.
Can't wait to see the test for the 27" with i7, Fusion and top video card.
Suddenly my 'pro' G5 is looking so much older.
are you saying that you want a 40 lb iMac with desktop components then? cause it would be a lot faster if iMacs weren't using laptop graphics.
I really wish that MR would stop highlighting GB scores as if they are the gospel and the all-encompassing indicator of Mac performance. Even a small disclaimer at the botton of the article would be a start.
That would be nice. I don't think X86 core speed should be the big point of focus here. The imacs now have the option of leveraging CUDA, which wouldn't have been possible with any of the older ones. The display implementation is different. Implementation can make a bigger difference at times than panel technology. You gain usb3. There are interesting points without focusing on drag racing cache loads (ripping off another of deconstruct's lines, because they're funny).
but not as good as it COULD be.
Like allot of apple products, as great as this is just one more step and this would be a AWSOME product. The same generation i7 comes with a 6 core version. As a video /3d /post user i would have payed more money to get those 2 extra cores. I would have payed even MORE if i had those 6 cores in my new MBPr. But again apple has CHOSEN to be a step behind the PC market, by leaning more CONSUMER then PRO.
That said I am very happy with this new generation of Ivy Bridge products from Apple.
Like allot of apple products, as great as this is just one more step and this would be a AWSOME product. The same generation i7 comes with a 6 core version. As a video /3d /post user i would have payed more money to get those 2 extra cores. I would have payed even MORE if i had those 6 cores in my new MBPr. But again apple has CHOSEN to be a step behind the PC market, by leaning more CONSUMER then PRO.
That said I am very happy with this new generation of Ivy Bridge products from Apple.
Okay so I'm kind of haunting the thread to a degree, but it interests me. I've explained the reason we wouldn't see a 6 core imac before. They do not use the same chipsets or socket type. The 3930k is a Sandy Bridge E variant. It pulls up to 130W and uses the LGA2011 socket type. It's also likely not compatible with Apple's thunderbolt implementation due to the lack of integrated graphics which they currently use. The 3770 pulls up to 77W and uses the LGA1155 socket. As far as I'm concerned, the Sandy Bridge E variant was never even a contender. Next year the Ivy Bridge E5s won't change anything in this regard. They'll still be using that socket, and they still lack integrated graphics. I don't see Apple making an additional design to accommodate that single cpu. A better solution would be to offer a more competitive base configuration in the mac pro.
I should add that I understand your concerns, and I know how resource hungry some of those programs can be.
quad can easily beat octo in lots of situations
Come on, Renzy, you should know better than this.
The fundamental reason that a quad could beat an octo is that much software cannot (or doesn't) scale past a few cores. Sometimes this can be addressed by rewriting the software - but often the algorithms (that is, the problems) have serial dependencies that inherently limit the amount of parallelism that is possible.
So, if you have one of the common problems that doesn't scale past four real cores (ignore hyperthreading, it's often useless or worse), you may note that:
So a 3.4 GHz quad Imac could trounce a 2.4 GHz duodeco Mac Pro on many applications.
Come on, Renzy, you should know better than this.
The fundamental reason that a quad could beat an octo is that much software cannot (or doesn't) scale past a few cores. Sometimes this can be addressed by rewriting the software - but often the algorithms (that is, the problems) have serial dependencies that inherently limit the amount of parallelism that is possible.
So, if you have one of the common problems that doesn't scale past four real cores (ignore hyperthreading, it's often useless or worse), you may note that:
- single-socket systems often have a higher clock rate the multi-socket systems
- octo-core currently must be multi-socket (no octo Core CPUs)
- multi-socket systems have NUMA, so performance can suffer
So a 3.4 GHz quad Imac could trounce a 2.4 GHz duodeco Mac Pro on many applications.
Or even an inch thicker.
You know, so that form doesn't cripple function for twice the price.
It's a freakin' desktop. Limp wristed pansies doen't have to carry it around.
What components?
The fastest desktop CPU and GPUs available instead of gimped mobile junk
that gets too hot and throttles down 50% under heavy loads.
2-3X faster and less expensive.
That was easy. Magic.
You haven't a clue what you are talking about.
Apple could have made it 2-3x faster AND less expensive, simply by making it fatter? How? What CPU could they have used? Like..actually name the model please. And GPU? Good luck ever fitting a full sized desktop GPU into an All-in-One. The 680MX option in the 27" by all accounts is pretty damn good. Even so, a GTX 680 (full sized desktop version) isn't going to be 2-3 times faster, and certainly not LESS expensive.
All this complaining about the new iMac is SO dumb.
They took out the optical drive, a move anyone with half a brain saw coming MONTHS ago. 10-15% performance gain in the CPU compared to last year? Well...yes, that was known and expected pretty much since Ivy Bridge was announced. News Flash! ALL 2012 COMPUTERS RUNNING IVY BRIDGE CPU'S SAW A 10-15% IMPROVEMENT OVER LAST YEAR!
So they took out the optical drive which apparently some professionals need for work. You can't adapt to new technologies? Or shock...buy an external drive to go along with your thousands of dollars in equipment? Why should everyone else be lumped with a component most of us don't use anymore?
So let's summarise:
2012 Model has:
- Faster CPU's
- USB 3 vs USB 2
- Much better GPU's
- Option of Fusion Drive storage
- By some accounts MORE user upgradeable than previous iMac's
- Thinner and lighter
and apparently Apple has failed? Riiiiiight.
It's a freakin' desktop. Limp wristed pansies doen't have to carry it around.
That's an offensive comment that isn't funny.
Do I need to send a gay NFL starter to carry you around to make the point?
Yup. I kinda semi clarified myself a couple posts afterwards.
Any usual application that's limited to a certain amount of cores, the current gen iMac would stomp on the Pro. In that situation, individual core performance is far more important.
But any application that can leverage (I've been using this word a lot today) multiple CPUs, such as rendering or video editing, the slower, older architecture of the Pro can still beat the iMac.
My argument was that the Pro is still better than the iMac at the things the Pro is generally used for.
Accepted without question.
Except for the people who buy a Mac Pro to do stuff that a 2007 MacBook could easily handle, because they want to show off.
One of my Thanksgiving recipes called for "semi-clarified butter"....
Last edited:
Two things I've noticed about the new iMac.
1. If you want a gigabyte of graphics you have to get the high end 27" iMac and if you want the 2GB graphics you have to get the high end model.
2. the 21" come with 5400rpm HDs whereas the 27" comes with 7200rpm HDs. Its seems that one the 21" I can't upgrade to a HD with 7200rpm. Again I'd have to go with the 27"
So basically if I wanted a 21" I have to settle for less then what I want in terms of specs. Is this cause the iMac is so thin now or is Apple being cheap in trying to force people to get the higher end models??
1. If you want a gigabyte of graphics you have to get the high end 27" iMac and if you want the 2GB graphics you have to get the high end model.
2. the 21" come with 5400rpm HDs whereas the 27" comes with 7200rpm HDs. Its seems that one the 21" I can't upgrade to a HD with 7200rpm. Again I'd have to go with the 27"
So basically if I wanted a 21" I have to settle for less then what I want in terms of specs. Is this cause the iMac is so thin now or is Apple being cheap in trying to force people to get the higher end models??
I look forward to reading all of the "helpful advice" that we will see in the months to come because of this rather misguided article.
----------
NO, they are not. *beats head into the wall*
It's already started.
Geekbench shows a similar score and to you it means that it's just as fast for all users? I give up.
You can beat your head against the wall all you want. Never said it's as fast for all users. Stop putting words in my mouth. I never said that you did.
For most tasks, the processor in the Mini is on par with their desktop counterparts. Geekbench aside, that is a fact. I said the mini is the better buy, when the performance is almost the same save for the GPU. And it's cheaper. You can order from Apple a 2.6 i7 mini, with a SSD and add your 16GB ram for $1200 out the door. Add a monitor and it's the better buy in my opinion. You're getting a machine that can perform on par with the high end imac for alot less. Even add a thunderbolt display and it's still less.
There isn't as much distinction between desktop and laptop CPU's as much as years past. Thermal constraints have blurred the lines for performance. Process is the same, thermal designs are only separated by as little as 20W for mainstream desktop parts and laptop parts. Higher end desktop parts(xeon) no there isn't a equal laptop part. As they require higher wattage.
Desktop CPU thermal and wattage is more in line with laptop parts than high end Xeon parts. Next year the line between laptop and desktop performance will be almost non-existent for mainstream CPU's for Haswell.
37, 47, 57W thermal design power (TDP) mobile processors.
35, 45, 55, 65, 77, and ~100W+ (high-end) TDP desktop processors
Now take a aspirin.
So is ‘turtlenecked overlord’. But that doesn’t stop you from using it. What comes around goes around I guess?
