New iMac Benchmarks Show 10-25% Improvement Over Previous Generation

MacPat333 said:

I believe you are refering to the 21.5" iMac with i7, not the maxed out 27" iMac i7 with Fusion Drive and 2GB grahics card. The SSD maxed out iMac should be even faster.

We have to wait for the scores of the maxed out 27" i7!

Click to expand...

Geekbench doesn't differentiate between fusion drives or GPU, the scores should be about the same. If taking into account geekbench scores, which he posted.

----------

HarryKNN21 said:

Speaking of the new 27", both 2.6 and 2.7 MBP and even the new i7 Mac Mini have still beaten it in 64 bit tests.

And as far as I know Geekbench scores don't involve SSD scores.

Click to expand...

The mini has the same CPU as the MBP. Should be the same score. The 2.7 should be slightly faster though.

http://browser.primatelabs.com/geekbench2/1226743

MacPat333 said:

I would like to know the score for the following configuration:

- 27" iMac late 2012
- 3.4GHz Quad-core Intel Core i7, Turbo Boost up to 3.9GHz
- 32GB 1600MHz DDR3 SDRAM - 4x8GB
- 3TB Fusion Drive
- NVIDIA GeForce GTX 680MX 2GB GDDR5

This is what my configuration will look like. However, I will order only with 8GB RAM and upgrade myself later.

What will be the difference between the above with 8GB RAM and 32GB RAM?

Thanks!

Click to expand...

This exactly what I ordered yesterday! Estimated delivery January 8th-9th.

I have determined the Geekbench test scores in these benchmarks are based on the 32-bit test bench, NOT 64-bit tests.

That is why the scores might look "off" to some of you.

You can add 1,000+ points to each score to roughly get the 64-bit equivalent result.

iamthedudeman said:

Geekbench doesn't differentiate between fusion drives or GPU, the scores should be about the same. If taking into account geekbench scores, which he posted.

----------



The mini has the same CPU as the MBP. Should be the same score. The 2.7 should be slightly faster though.

http://browser.primatelabs.com/geekbench2/1226743

Click to expand...

Thus it surprised me, how could a laptop CPU has a higher benchmark score than a desktop CPU.

had a 2.5 i5 2011 imac that benchmarked at 7500. My 2012 15inch 2.3 mbp benchmarks at 12000+

cgdrennan said:

You post in ignorance my friend. They have already done tear downs of the new iMac and it is very user upgradable... Hard drive, ram, and even the fact that the CPU is not soldered. The screen is held on by magnets like the previous model but since the LCD is laminated to the glass it is easier to get in to.

Click to expand...

Is it?

"Well iFixit weren't overly enthused with some of the new design's internal fixings, making a special reference to the use of glue to bind the LCD and glass to the rest of the frame, instead of magnets as found in previous models. "

I can't get over the fact that my retMBP is as fast (well, almost) as the fastest 27 iMac! And much faster than the fastest 21 iMac. I am really confused why Apple configured the new iMacs this way.

What sort of performance hit do people think they take from having a 5400 RPM drive? I've read elsewhere that they use 2.5" laptop drives - that sounds like a step back.

Looking at these Geekbench marks the mid 27" 2011 iMac score seem on the low side. I just checked my Geekbench scores and the last one was an overall score of 12881.

For me the USB3 Ports and if they really reduced the display reflexions by 70% are things I'm sooo happy about.

I'm sorry if the new Imac is such a big disappointment for so many people but I'm looking forward to get mine delivered.

Unfortunately it looks like people think the 10-15% increase is because of the thinness... not because that is the average speed upgrade that Ivy Bridge has provided.

Naimfan said:

What sort of performance hit do people think they take from having a 5400 RPM drive? I've read elsewhere that they use 2.5" laptop drives - that sounds like a step back.

Click to expand...

As compared to what? A 7200 RPM hard drive of the same density? It's about 28% slower. I mention density because a 7200 RPM drive that is 500GB will read slower than one that is 1TB and that slower than a 3TB drive (assuming a single platter is involved and other factors) and thus at some point a high density 5200 RPM drive will be typically as fast as a smaller 7200 RPM drive. Both will be slower than a typical SSD drive. But then there is seek times for smaller files and older SSDs drives had much slower write times than reads. It's not so simple when you get into the details. You could RAID two 5200 RPM drives and get much faster performance than a single 7200 RPM for perhaps a similar price if you have the bays/box for it and so on (e.g. my new Mac Mini has two 1TB 5200 RPM drives in it; I plan on looking to see if I can reconfigure them with RAID for faster performance. The lack of a physical install disc might make it a bit more complicated than it would otherwise be, but I'm sure it could still be done... actually here's the Apple support article on how to do it: http://support.apple.com/kb/HT4886 )

Naimfan said:

So in exchange for more money, we get slightly faster performance which will be invisible to almost EVERY user of an iMac, no optical drive, a non-user upgradable machine, and a new design that values form over function.

In short, the new iMac joins the Macbook Air and the retina Macbook Pro as a disposable computer.

Apple, in my view, needs to get over its obsession with thinness.

And why, given the weight reduction and the elimination of the optical drive, does it cost MORE?

Click to expand...

25% is huge..

WardC said:

Their tests are bunk, because as I have shown, the top-end 3.4GHz 2011 model is actually faster than the top-end 21.5" -- their readings are not accurate, or they ran the 32-bit tests and not the 64-bit tests (this is what I am guessing).

Click to expand...

It's not their tests that are "bunk" but the way they calculate and subsequently report representative scores for each model (and yes it's not clear whether they include 32-bit results). I've consistently managed to get scores significantly higher than the reported values for my system. Their summary scores should be taken with a grain of salt and here's why:

Without knowing how they calculate the representative score, I'm willing to bet they do nothing more than take an average of all the scores submitted for each model. I'm also willing to bet that these scores are not normally distributed but are skewed heavily towards the upper end of the range, meaning the arithmetic average will significantly underrepresent the potential performance of each model. The cause of the skewness likely relates to people that submit scores under non-optimal settings. Perhaps they have other applications open and running or have lots of extensions and add-ins consuming system resources in the background. I always quit everything including the dashboard before running. It's not clear whether Primate Labs bothers to filter outliers either.

I've written to Primate Legs (or their forums?) to request a confirmation of their method but they seem to be unwilling to disclose or discuss the reasons for their summary scores being significantly lower than many of the submissions for each system. How much lower they are depends on the characteristics of the data submitted. If they include 32-bit results in their scores it exacerbates the problem even further.

If you want unbiased CPU performance results, why not just look at Intel's own benchmarks for each processor? I realise GeekBench tests disk speed as well which is another thing that complicates the meaning and usefulness of their combined scores.

spcdust said:

Looking at these Geekbench marks the mid 27" 2011 iMac score seem on the low side. I just checked my Geekbench scores and the last one was an overall score of 12881.

Click to expand...

Yep, they're all too low. The real question is why. See my attempt for an explanation above.

For a company that charges a lot of money for their software and whose results are widely reported it's strange that they seemingly haven't implemented a decent results function.

I'm happy and have a new 27 on the way to replace my current 27. The machine works for me.

MacSince1990 said:

25% is huge..

Click to expand...

No, it isn't. It's not going to let someone surf Facebook or eBay any faster, and it's not going to let you read email or the NY Times online any faster. It's like having a car that can do 160 MPH that's been upgraded so it can now go 200 MPH. Since the overwhelming majority of people rarely, if ever, go over 80 or so MPH, they'll never know the difference.

Phx08 said:

Holy crap...all the people complaining need to really put a cork in it.

If you don't like the product, don't buy it. Simple.

Click to expand...

People complaining and people complaining about the people complaining are a time honored Mac Rumors tradition.

Allow us our vices.

I agree with many of you. I actually use my optical drive from time to time when it comes to picture transfers or when I'm making copies of other disks. To remove that basically defeats the purpose of the iMac, which I use to reduce the footprint on my desktop. If suddenly I have to have all of these plug in peripherals, then it defeats the purpose.

I also agree with Apple's obsession with 'thin'. It came back to bite them with the iPad, and will with other devices. Form over function is not a viable equation. Unless of course you ask my wife with respect to shoes.

dashiel said:

I’m always curious what people who say this do with their computers. I look at people like Louis C.K. who edited the first two seasons his show on a 13" MacBook Pro and then come here and see people moaning about benchmarks … so I’m curious what do you actually do that you need such power?

Click to expand...

I edit tons of photos and some video and my wife does intense scientific epidemiological research. We are talking incredible number crunching and graphic needs. The faster, the better, because all the new cameras and software demand more and more resources.

I am waiting for the Mac Pro. It is so sad that Apple has forgotten about its core of professional clients!!!!!!

Now you know... I'm surprised you have to ask... Ignorance is not bliss, nor are condescending questions like, "...what do you actually do that you need so much power?" Not everyone sits around on Facebook all day, or plays video games. We are out in the field working very hard and expect our tools, that we pay a lot of money for, to be up to the job. Really surprised anyone even has to ask that question

Well, I'm happy with the results because...

My newly built editing Hackintosh is getting 14049 in Geekbench 64bit.
32GB of RAM, 2x 6gbps SATA HD RAID, SSD, 2GB GFX Card, two large IPS screens - £2000 all in. Glad I finally bit the bullet and built it, no looking back now.

Naimfan said:

No, it isn't. It's not going to let someone surf Facebook or eBay any faster, and it's not going to let you read email or the NY Times online any faster. It's like having a car that can do 160 MPH that's been upgraded so it can now go 200 MPH. Since the overwhelming majority of people rarely, if ever, go over 80 or so MPH, they'll never know the difference.

Click to expand...

Keep on telling yourself that. If you don't know the difference between a car than can do 160 and one that can do 200, you don't deserve such a car and surely could never afford one. Unlike cars, you CAN run your computer at full speed with no real danger and the mere fact that you appear to believe that the ONLY thing people do with computers is freaking Facebook and e-mail tells me that you would have been better off with a $300 netbook or an iPad because you don't actually make use of your computer worth a darn.

To me, a 25% increase in speed means 25% more DSP effects I can use in Logic Pro without red-lining the engine (yes, it happened on several tracks I've worked on with my 2008 Core2Duo and my quad-core i7 is over 300% faster, so it'll actually be a bit more than the 25% difference mentioned here, but that's another story). To me, a 25% increase in speed means a Handbrake encode that would have taken 4 hours will now only take 3.2 hours (multiply that times the number of jobs to encode and the savings really start adding up).

As for cars, if you're racing at the track, a 25% increase in power will make a very noticeable difference. If you're just going to drive in the city to get milk, you probably won't notice it. I gather you just drive to get milk and that's why you think it doesn't matter. Well, it's not designed for you, so don't worry about it.

Jeez.... thinner makes it faster. Duuuh.

MagnusVonMagnum said:

...and the mere fact that you appear to believe that the ONLY thing people do with computers is freaking Facebook and e-mail tells me that you would have been better off with a $300 netbook or an iPad because you don't actually make use of your computer worth a darn.

To me, a 25% increase in speed means 25% more DSP effects I can use in Logic Pro without red-lining the engine...

...If you're just going to drive in the city to get milk, you probably won't notice it. I gather you just drive to get milk and that's why you think it doesn't matter. Well, it's not designed for you, so don't worry about it.

Click to expand...

The point that the guy was making is that the vast majority of computer usage is put to low-processor intensity jobs like Facebook, or word processing, or surfing porn. The fact of the matter is that even owning V34 SUV's, capable of towing a small town across the grand canyon, most people will still drive it to go get groceries and pick the kids up from school. They just want to look cool doing it. The iMac looks cool.

Just to compare dicks, I don't know what plugins you're running, but I frequently write 20-or-so track ditties in Live 8.0 on a 2-core i5 without maxing out my processor too much. REGARDLESS... there will ALWAYS be more that the computer can do than the processor will allow. Thus, why upgradable computers (like the Mac Pro... if only they'd update the base specs a bit more...) are probably a better investment to a power user who wants to keep current or eek the most performance out of a 5+ year old system.

Remember kiddies: more thinner=more faster. its like racing stripes for a computer!

Phx08 said:

Holy crap...all the people complaining need to really put a cork in it.

If you don't like the product, don't buy it. Simple.

Click to expand...

I would normally agree, but this isn't the comment section on an Engadget article that spawns fanboy wars. Most people here are longtime Mac OS customers, and unfortunately, short of going hackintosh, you have to use Apple computers.

So, unfortunately, you are kind of stuck with what you get when a new iMac comes out. If you want a little power over a mac mini, but don't need to step up to a MacPro, the iMac is sort of what you have to buy. People just want the best iMac they can get.

I'm not 100% happy with this design almost exclusively because the HDD isn't user replaceable. I can deal with BTO RAM, but HDD just fail too often to have to ditch a computer because of that. However, my 21.5 arrives today because I HAVE to upgrade (G5 is becoming more worthless by the day) and this is the only computer that really fits my needs with Mac OS.

So, I say complain all you want.