Fastest Core i7 iMac model rated best buy
Performance site Bare Feats has compared the performance of Apple's new Core i3, i5, and i7 27 inch iMacs and determined the high end model offers the best performance per dollar.
The report contrasts the performance benchmarks rankings of Cinebench 11 and Geekbench 2 to compare the the $1699 entry level dual core 3.2GHz Core i3 against the midrange $1999 quad core 2.8GHz Core i5 and the high end $2199 quad core 2.93GHz Core i7.
Overall, the performance scores indicate that the Core i5 offers a performance boost up to 41 percent at just 18 percent more cost, while the high end Core i7 is up to 45 percent faster than the Core i5 while being just 10 percent more. It's also 104 percent faster than the Core i3 model while being just 29 percent more expensive.
"Going by the Cinebench rating, you pay $400 for each rating point on the Core i7, you pay $526 for each rating point on the Core i5, and you pay $629 for each rating point on the Core i3," the site noted, concluding, "the top iMac model is the best buy."
Factors influencing performance
The test results reflect numbers published in our review of the new iMac models, which show a marked increase in performance between the Core i3, i5, and i7 processors (below).
The entry level Core i3 CPU design does not support Intel's "Turbo Boost," a feature that enables individual CPU cores to dynamically speed up when only one is being actively used. It does support Hyper Threading however, which enables the dual core chip to act like a virtual quad core when running software that is optimized for multiple cores.
The Quad Core Core i5 and i7 also pack twice as much L3 cache; 8MB, compared to 4MB on the dual core models. All of the machines use a faster new 1333MHz bus for RAM, although in testing, the site reported that using faster RAM only appears to offer a slight advantage up to 3 percent on the new machines.
However, despite the newest 2.93GHz Core i7 model only being clocked 5 percent faster than last fall's Core i7 iMac, the site noted that "the new Mid-2010 iMac Core i7 was 3 to 23 percent faster than the Late 2009 iMac Core i7 in our various CPU intensive tests."
Compared to the mobile Core i7 used in the MacBook Pro, "the new iMac Core i7 is 43 to 120 percent faster running our 'multi-core aware' suite," the site reported.
Additional considerations when upgrading
The site also points out that while some users may decide they have light uses that don't demand top performance, there are good reasons to pay extra to "future proof" for new apps and uses that will end up making the machine useful longer.
It also recommends considering future resale value, where the additional residual value of a higher end model may recover enough to warrant the additional initial cost.
The report contrasts the performance benchmarks rankings of Cinebench 11 and Geekbench 2 to compare the the $1699 entry level dual core 3.2GHz Core i3 against the midrange $1999 quad core 2.8GHz Core i5 and the high end $2199 quad core 2.93GHz Core i7.
Overall, the performance scores indicate that the Core i5 offers a performance boost up to 41 percent at just 18 percent more cost, while the high end Core i7 is up to 45 percent faster than the Core i5 while being just 10 percent more. It's also 104 percent faster than the Core i3 model while being just 29 percent more expensive.
"Going by the Cinebench rating, you pay $400 for each rating point on the Core i7, you pay $526 for each rating point on the Core i5, and you pay $629 for each rating point on the Core i3," the site noted, concluding, "the top iMac model is the best buy."
Factors influencing performance
The test results reflect numbers published in our review of the new iMac models, which show a marked increase in performance between the Core i3, i5, and i7 processors (below).
The entry level Core i3 CPU design does not support Intel's "Turbo Boost," a feature that enables individual CPU cores to dynamically speed up when only one is being actively used. It does support Hyper Threading however, which enables the dual core chip to act like a virtual quad core when running software that is optimized for multiple cores.
The Quad Core Core i5 and i7 also pack twice as much L3 cache; 8MB, compared to 4MB on the dual core models. All of the machines use a faster new 1333MHz bus for RAM, although in testing, the site reported that using faster RAM only appears to offer a slight advantage up to 3 percent on the new machines.
However, despite the newest 2.93GHz Core i7 model only being clocked 5 percent faster than last fall's Core i7 iMac, the site noted that "the new Mid-2010 iMac Core i7 was 3 to 23 percent faster than the Late 2009 iMac Core i7 in our various CPU intensive tests."
Compared to the mobile Core i7 used in the MacBook Pro, "the new iMac Core i7 is 43 to 120 percent faster running our 'multi-core aware' suite," the site reported.
Additional considerations when upgrading
The site also points out that while some users may decide they have light uses that don't demand top performance, there are good reasons to pay extra to "future proof" for new apps and uses that will end up making the machine useful longer.
It also recommends considering future resale value, where the additional residual value of a higher end model may recover enough to warrant the additional initial cost.
Comments
And, the 2.8 Quad is an i5 not an i7... the chart shows i7 2.8Ghz with 8GB ram..
the i5 isn't 2.66Ghz either, as noted in first reply. Is the 2.66Ghz is a typo? Are you sure which i5 is being tested? the Dual core 3.6Ghz with 4MB L3 or the Quad core with 8MB L3?
Says Quad, but as far as typos' go 3.60 would be closer since 2.8 doesn't have a 6 anywhere in the spec.
I don't think any of this chart is legit.
$1699 dual-core 3.2GHz Core i3
$1899 dual-core 3.6GHz Core i5
$1999 quad-core 2.8GHz Core i5
$2199 quad-core 2.93GHz Core i7
Since the dual-core Core i5 is the one that fits in my budget (with 8 GB RAM), I would be curious to see how that one stacks up.
Software only grows more complex and hungrier. Files ever larger. My first machine has a whopping 48 K of RAM and no drives at all!
Whoa, Flashback!
AppleInsider, thank you for this interesting article. However, you really need to clarify the exact model as defined by "Late 2009 iMac" in your performance comparison bar graph. I assume that late 2009 model was NOT an i7 as the bars show the 2010 i7 to be almost 200% faster, which doesn't make sense. So what CPU and Clock is it?
The CPU is cited in the chart directly above the graph.
Late 2009 iMac - Core 2 Duo 3.06 Ghz, 4GB RAM.
It's the previous year's model.
As far as "future proofing" goes, with these new iMacs, that concept is rather silly. In all likelihood, these will be the last iMacs without USB 3.0. This will also adversely affect their resale value.
Frankly, with Apple's current line up, unless you are a professional and can justify the expense, I think most users would be best served by waiting until another round of hardware updates before buying new machines if at all possible.
-kpluck
Ignoring the minor details in the other posts, I agree with the conclusion. The i7 would be my choice.
Minor details that funny. Also there is no such thing as future proofing your system. Also an AIO will never be a best buy because what you get is always what you are going to have, you can't upgrade anything.
Also on paper this is a gaming system with no real gaming library. Unless Apple still considers Quake 4 a valid game.
C'mon guys, these look like (or similar to) the same graphs and specs that were in yesterday's report, and there are still faux paux's! There is no mid-2010 2.66 Ghz i5!!!!! Get it right, will ya?
If you're going to use a term like "faux pas," it would be a good idea to learn how to spell it first.
The CPU is cited in the chart directly above the graph.
Late 2009 iMac - Core 2 Duo 3.06 Ghz, 4GB RAM.
Thank you, Samurai! My feeble eyes overlooked that part.
Being an owner of the Oct. 2009 iMac i7 2.8GHz (8GB RAM), it would be interesting to see that older i7 iMac model added to the same performance graph. The 0.13GHz faster CPU alone probably wouldn't make the 2010 i7 model "feel" noticeably faster than the 2009 model, but I am curious about the impact of the 1066 RAM vs. 1333 RAM.
Anyway, thank you.
The 3.6GHz i5 gets about 8000 in Geekbench (same as the quad 45nm i5) vs 10000 for the older quad i7. It also reaches that with a 73W TDP vs 95W TDP so it should run cooler.
The i5 has a spare IGP so power consumption can be lower too if Apple uses it, it supports VT-d unlike the i3 as well as Turbo Boost so Parallels/VMWare gaming may become far more viable.
The quad core models do have GPUs with double precision support and 27" displays but if you don't need double precision, the 5750 is only 30-50% faster than the 5670 and the 27" display is overkill for most people.
I'd say the best value is determined after purchasing a machine and then weighing up how much you regret your purchase based on usage and for the vast majority of people, spending the extra $500 to get from the dual 3.6GHz i5 to the quad i7 won't be worth it.
But strictly speaking, there are a lot us us digging our heels in with our current gear till we're forced to upgrade. Dead Power supply. Blown Motherboard. Or... getting that Adobe Illustrator, Photoshop, or InDesign file from a client or vendor that we CAN'T OPEN and have to pull the trigger on a software upgrade that requires new gear.
Adobe CS5 requires minimum Intel Core 2 Duo... good bye to all those aging Power PC machines in Art Departments and Design Studios out there.
Be nice, I resemble that remark
It's been my experience over the years, that if you do any kind of professional work at all on a computer, BUY THE HEAVIEST IRON YOU CAN POSSIBLY AFFORD that gets the job done. Stuff it with as much RAM and HD space as you can pry out of your budget. Buying low end will make a machine, however capable, be obsolete twice as fast. The higher end machine will give you a much longer service life.
Agreed, in part. RAM and a larger HD can always be purchased later, and will likely be cheaper in the future. Processor speed, graphics cards, & bus speed, are all things that you would be stuck with (are not likely upgradeable).
So, I generally go with the Mac that has the best non-user replaceable components I can afford.
A side note... I was at the Apple Store the other day, helping a friend purchase a MacBook Pro. She wanted the 15" model. The salesman tried to talk her down to the 13". Then when she insisted on the 15", he tried to talk her down to the entry level model, when she wanted the mid-range one.
I was first surprised, then kinda annoyed... But I guess the salesman was just trying to save her some money.
ATI Radeon HD 5750 w/ 1GB GDDR5 VRAM over the ATI Radeon HD 5670 w/ 512MB GDDR3 VRAM
-OR-
Quad Core i7 2.93GHz (turbo up to 3.6GHz) over the Quad Core i5 2.8GHz (turbo up to 3.33GHz)???
Both upgrades are $200 each.
If you had to choose upgrading the CPU -OR- the GPU on the lower-end 27" model, would you choose the
ATI Radeon HD 5750 w/ 1GB GDDR5 VRAM over the ATI Radeon HD 5670 w/ 512MB GDDR3 VRAM
-OR-
Quad Core i7 2.93GHz (turbo up to 3.6GHz) over the Quad Core i5 2.8GHz (turbo up to 3.33GHz)???
Both upgrades are $200 each.
I'd say it depends on your intended use...
Future proofing is a bad goal to set for most people. Look at the jumps in technology within a year. I don't see future proofing as a motivator for up buying.
I disagree. I bought my late 2006 Macbook Pro with the top processor for this very reason. In the past, I waited for the "deals" on clearing models when new ones were introduced. While I paid less, I also got a much less useful life out of the machine.
My MBP is still very usable and it's almost 4 years old. My strategy in the future will be to buy the fastest model; I think it costs me less in the long run in terms of upgrades.
Perhaps if you're a geek and buy the latest and greatest every year, you would be correct.
Tommy
My MBP is still very usable and it's almost 4 years old.
Yes, but you could have got the enrty level model, and had enough left over to have bought a brand new MBP last year that currently beats your model in performance.