PDA

View Full Version : Benchmarks for Mid-2012 MacBook Pro and MacBook Air




MacRumors
Jun 12, 2012, 02:47 PM
http://images.macrumors.com/im/macrumorsthreadlogo.gif (http://www.macrumors.com/2012/06/12/benchmarks-for-mid-2012-macbook-pro-and-macbook-air/)


With the first Mid-2012 MacBook Pro and MacBook Air models reaching the public, Primate Labs has collated some data (http://www.primatelabs.com/blog/2012/06/macbook-air-pro-benchmarks/) from its Geekbench 2 benchmarking database to assess the raw performance of these systems compared to their predecessors.

For the MacBook Pro, Primate Labs has data on four different models: the new Retina model with 2.3 GHz and 2.6 GHz processors and the non-Retina model with 2.3 GHz and 2.7 GHz processors. The top-of-the-line 2.7 GHz system registers with a Geekbench average score of 12,303, roughly 16% higher than the top-of-the-line 2.5 GHz Sandy Bridge system from the previous generation.

http://images.macrumors.com/article-new/2012/06/geekbench_mid_2012_macbook_pro.jpg


Notably, the leaked MacBook Pro benchmark (http://www.macrumors.com/2012/05/14/unreleased-2012-macbook-pro-and-imac-models-showing-up-in-benchmarks/) from mid-May does appear to have been legitimate, with details corresponding to the new non-Retina 15-inch MacBook Pro. That machine is designated MacBookPro9,1, while the 13-inch model is designated MacBookPro9,2. The Retina MacBook Pro appears as MacBookPro10,1.

On the MacBook Air side, top-of-the-line systems are seeing boosts of over 20% in Geekbench scores over their corresponding predecessors. As with the MacBook Pro, even the low-end systems of the new generation outperform the high end of the previous generation.

http://images.macrumors.com/article-new/2012/06/benchmarks.png


Geekbench testing focuses on processor and memory performance, providing comparisons of raw power between machines but only telling part of the story. But with these machines seeing significant boosts in graphics performance with the addition of Intel HD 4000 integrated graphics and/or the NVIDIA GeForce GT 650M, and systems with solid-state drives using faster drives than in the previous generation, real-world performance should see marked improvement.

Article Link: Benchmarks for Mid-2012 MacBook Pro and MacBook Air (http://www.macrumors.com/2012/06/12/benchmarks-for-mid-2012-macbook-pro-and-macbook-air/)



daneoni
Jun 12, 2012, 02:50 PM
Chess will run blazingly now!

iBug2
Jun 12, 2012, 02:51 PM
Great single core performance on the Pro. The scaling to 8 threads isn't as good as 8 real cores but close enough.

chrmjenkins
Jun 12, 2012, 02:51 PM
These kinds of benchmarks were more relevant in the powerPC days. Now it's pretty much, "Yup, those faster intel processors that we already saw the benchmarks of are in the new MBP."

tkhan456
Jun 12, 2012, 02:52 PM
So has anyone seen one of the new MBP with Retina in the wild at a store? I haven't found a single store in my area with one. I try to not spend $3000 on something without trying it first.

entropy1980
Jun 12, 2012, 02:52 PM
Can't wait for my 13" Pro to get here tomorrow!!! Bummed to not see any benchmarks for it yet though...

DrFu79
Jun 12, 2012, 02:53 PM
Now let's see Gaming Performance. :cool:

sweetbrat
Jun 12, 2012, 02:53 PM
So has anyone seen one of the new MBP with Retina in the wild at a store? I haven't found a single store in my area with one. I try to not spend $3000 on something without trying it first.

If you check in the MBP forum, someone over there found one at the store and bought it. http://forums.macrumors.com/showthread.php?t=1385037

Demosthenes X
Jun 12, 2012, 02:55 PM
So has anyone seen one of the new MBP with Retina in the wild at a store? I haven't found a single store in my area with one. I try to not spend $3000 on something without trying it first.

Checked out the local Apple Store today, only last-gen machines. I thought for sure if they were shipping yesterday they would have been on display today...

djrod
Jun 12, 2012, 02:56 PM
The top of the line Macbook Air is almost as fast as my 2009 i7 iMac :eek: 7007 vs 8500

marcusj0015
Jun 12, 2012, 02:56 PM
I was really expecting it to be a lot faster then that. :/

fel10
Jun 12, 2012, 02:57 PM
Very nice, if my MBP was in bad shape, I would definitely get the retina one, but I still have a few more years with my current MBP.

amien
Jun 12, 2012, 03:01 PM
1.8 and 2.0 isnt hardly worth the extra investment on the air?

nec207
Jun 12, 2012, 03:03 PM
I was really expecting it to be a lot faster then that. :/


Does that mean these new ivy bridge CPU's are 17% faster than less year CPU's ?

coder12
Jun 12, 2012, 03:03 PM
Chess will run blazingly now!

No kidding! I remember back in the DOS days playing Battle Chess (or something like that...) and it took like 30 seconds to 2 minutes for the computer to think. ;)

CylonGlitch
Jun 12, 2012, 03:04 PM
Anyone else notice that the MBA chart doesn't look skewed correctly.

Going from the 11-inch Mid 2012 at 5770 to the 13-inch Mid 2012 at 6915 (a delta of 145) is about 1/2 the difference of the 13-inch Mid 2012 at 6915 to the 11-inch Mid 2012 7004 (a delta of 89). Just looks odd.

needfx
Jun 12, 2012, 03:05 PM
they just trashed my maxed out mid11 mba

edit
seems that the 5710 to 7004 bump does not register as significant to a couple of you, specifically 22,67%

swingerofbirch
Jun 12, 2012, 03:07 PM
No one cares about the poor 13-inch MacBook Pro . . .

ivbaseball06
Jun 12, 2012, 03:10 PM
Interesting that the numbers are very similar to the leaked benchmarks from a few months ago, especially considering that there was an iMac benchmark too. Maybe Apple was developing two new products, the Retina MBP and some new iMac. Maybe instead of shadowing each other by releasing them the same day they decided to break their releases up, or perhaps the iMac isn't ready yet. Those benchmarks for the new iMac are the biggest evidence of a potential iMac release in the next month or so, IMO. A man can dream right?

charlieegan3
Jun 12, 2012, 03:10 PM
big jump in the air performance! good to see.

Not such a jump in the Pro though...

auero
Jun 12, 2012, 03:13 PM
I think it should be noted that last generation 17" Macbook Pro's didn't come with 8GB of memory.

LaWally
Jun 12, 2012, 03:14 PM
1.8 and 2.0 isnt hardly worth the extra investment on the air?

Upgrading from 1.8 (i5) to 2.0 (i7) costs $100, which is about 4.7% of the cost of the MBA 13" (i5) fully configured. For that you get a 1.3% increase in gb score.

I don't like that math.

malman89
Jun 12, 2012, 03:14 PM
Now let's see Gaming Performance. :cool:

Could check the 650M on notebookcheck.net - it's been covered extensively well before yesterday.

I was really expecting it to be a lot faster then that. :/

Ivy Bridge has always been known to be a smaller performance upgrade, but 16-20% might matter for some people.

theSeb
Jun 12, 2012, 03:15 PM
So if the Macbook Pro geekbench was legitimate, that would suggest the Ivy Bridge iMac was legitimate too. Can we expect a silent spec bump?

heisetax
Jun 12, 2012, 03:15 PM
No 17" model means that I will be left to only used models. I used to do that. Plus it saves dollars. Apple has done their part in keeping my bank account untouched. They finished my want to update with keeping the Mac Pro the longest computer to not keep an update. But at least they are still selling one, unlike the 17" laptop.

I had a 15" PowerBook, but after purchasing the first of at least 4 17" Ma laptops, I refuse to go back in size.

I'll have to check out the Frankentosh market, that Hackintosh to most. No wonder I do not feel welcome at the local Apple Store. I have my own requirements & will not go with what Apple is selling unless it meets those requirements.

What good is a 1,0000,000 X 1,000,0000 pixel display if it is to small to see or have enough useable space to work with?

I give the new Macs 0 Stars at this point. Let's see if they can change that. To make matters worst I've purchased only Macs since they came out in 1984. I always thought that Apple has always been run by the almighty dollar, now I know it for sure.

diamond.g
Jun 12, 2012, 03:18 PM
Now let's see Gaming Performance. :cool:

The correct question is will it run UE4 (http://www.youtube.com/watch?v=MOvfn1p92_8). :p :D

nec207
Jun 12, 2012, 03:18 PM
Could check the 650M on notebookcheck.net - it's been covered extensively well before yesterday.



Ivy Bridge has always been known to be a smaller performance upgrade, but 16-20% might matter for some people.

Has that been confirmed yet? It was claim the new CPU this year would be 16-20% faster than last year CPU.

DribbleCastle
Jun 12, 2012, 03:22 PM
1.8 and 2.0 isnt hardly worth the extra investment on the air?

Exactly my thoughts. Really surprised that it adds so little in performance.

bugfaceuk
Jun 12, 2012, 03:26 PM
I bet Safari is snappier.

sw1tcher
Jun 12, 2012, 03:26 PM
No previous 15" MBP with 2.2GHz Quad-core i7 in charts but I guess it's safe to say that the new 15" MBP with 2.3GHz quad-core i7 is about 15% faster, right?

Was hoping for better.

Guess I'll pull the trigger on a refurb 15" MBP with 2.2GHz Quad-core i7 for $1359 vs a 15" MBP with 2.3GHz quad-core Intel Core i7 for $1799. Then use money saved to upgrade HDD to a SSD to close the performance gap in half? Seems like a smarter buy.

xDKP
Jun 12, 2012, 03:31 PM
Exactly my thoughts. Really surprised that it adds so little in performance.

Also looked for the CPU-upgrade speeds... Looks like the money is better spend on the memory upgrades?

biohead
Jun 12, 2012, 03:31 PM
Looking at a lot of benchmarks in the database, there are a lot running 10.7.3 - I'd have at least thought they'd have shipped with 10.7.4 as a base!

Ve3tro
Jun 12, 2012, 03:31 PM
Looking forward to my 15" Pro. :D

theSeb
Jun 12, 2012, 03:34 PM
My 2011 13" Ultimate gets 5813 in 32-bit Geekbench. Those are impressive gains for the top of the range 2012.

themcfly
Jun 12, 2012, 03:39 PM
So if the Macbook Pro geekbench was legitimate, that would suggest the Ivy Bridge iMac was legitimate too. Can we expect a silent spec bump?
You can also see that the motherboard number in the leaked MacBookPro9,1 is exactly the same of the motherboard number in the benchmarks coming out. BIOS slightly different. Now i think they were indeed real on the MBP side, let's hope for the iMac too.

grooveattack
Jun 12, 2012, 03:40 PM
My macpro (see sig) just got a 12,600.

Looking forward to getting a fully specced retina MBP and seeing the same speeds! Awesome! Probably why they are perhaps killing the pro?

OrangeSVTguy
Jun 12, 2012, 03:49 PM
Funny how the retina and same non-retina MBP scored with the retina being just slightly higher. Pushing 4x the pixels would think it would be less?

jakebrosy
Jun 12, 2012, 03:50 PM
My mid-2011 15" 2.3GHz i7 comes in at 11440...so not a huge jump for the portable though it still doesn't compare to my MacPro's 21730 but it's WAY more portable. Sadly I'm mid-cycle so I'll hold for next year...but will be jonesing hard for that retina display when the time comes.

And for the naysayers: video processing needs raw processing power and the extra 4 cores are routinely maxed out when outputting from FCX. It makes a HUGE difference over the i5 or the dual-core i7 - and narrows the gap significantly with the MacPro.

If you edit a few photos, send email and surf the web you probably don't need anywhere NEAR the processing power of an MBP. But, don't narf about the folks that do.

amien
Jun 12, 2012, 03:53 PM
Upgrading from 1.8 (i5) to 2.0 (i7) costs $100, which is about 4.7% of the cost of the MBA 13" (i5) fully configured. For that you get a 1.3% increase in gb score.

I don't like that math.

In the netherlands its even 150 euros. Against 1549 euro for laptop. About 10%
Dont know what the effect of the battery duration is

dona83
Jun 12, 2012, 03:57 PM
These kinds of benchmarks were more relevant in the powerPC days. Now it's pretty much, "Yup, those faster intel processors that we already saw the benchmarks of are in the new MBP."

If you look at cpubenchmark.net, the new Core i5 3317U has been tested to be slower than the old Core i5 2467M, and the Core i7-3667U isn't available at all, so this test is still very much relevant.

ipacmm
Jun 12, 2012, 04:02 PM
Here are some unboxed pics, forget the performance the retina display is incredible!

Unboxed MacBook Pro with Retina Display (http://blog.soundearphones.com/odds-ends/macbook-pro-retina-display-unboxing/)

theSeb
Jun 12, 2012, 04:03 PM
Funny how the retina and same non-retina MBP scored with the retina being just slightly higher. Pushing 4x the pixels would think it would be less?
Geekbench is a CPU test so what the GPU is doing would not impact it in any way.

The one plausible explanation, if the difference was at least a couple of hundred points, might have been the improved cooling system in the RMBP, which would allow Turbo Boost to stay on for longer.

However, 44 points is an insignificant difference in this benchmark. I can run it over and over again and get such small variances on each run. This is due to the fact that it is run by users in an uncontrolled manner. Therefore the computer could be doing various things whilst the test is running.

agentjomo
Jun 12, 2012, 04:05 PM
No 17" model means that I will be left to only used models. I used to do that. Plus it saves dollars. Apple has done their part in keeping my bank account untouched. They finished my want to update with keeping the Mac Pro the longest computer to not keep an update. But at least they are still selling one, unlike the 17" laptop.

I had a 15" PowerBook, but after purchasing the first of at least 4 17" Ma laptops, I refuse to go back in size.

I'll have to check out the Frankentosh market, that Hackintosh to most. No wonder I do not feel welcome at the local Apple Store. I have my own requirements & will not go with what Apple is selling unless it meets those requirements.

What good is a 1,0000,000 X 1,000,0000 pixel display if it is to small to see or have enough useable space to work with? :confused:

I give the new Macs 0 Stars at this point. Let's see if they can change that. To make matters worst I've purchased only Macs since they came out in 1984. I always thought that Apple has always been run by the almighty dollar, now I know it for sure.

Yeah - Ideal time for the Hackintosh shops to make a comeback and make some sales to support the old skewl pro users! I would love to upgrade my Mac Pro - but new Mac Pro updates are weak sauce! Maybe Power Computing can be resurrected???

The Mercurian
Jun 12, 2012, 04:12 PM
My late 2011 2.5Ghz MBP with 16Gb Crucial RAM scores just over 11,000.

And I will use that for what I do.

Honestly ? I'm happy with my machine and that I didn't wait. The performance boost isn't enough to have justified the wait when I wanted to buy.

I don't mind being end of the line for this moddle = its a well established machine. I'll let all the early adopters work the kinks out of this retina meachine and maybe think about one next release when they drop the old MPB, drop the price for the retina models, have 32GB RAM, 1TB SSD and 3.5GhZ Ivy bridge and scoring 15,000 :P

Eidorian
Jun 12, 2012, 04:14 PM
It is all about that delicious GK107 based GT 650M.

Torrijos
Jun 12, 2012, 04:16 PM
Does that mean these new ivy bridge CPU's are 17% faster than less year CPU's ?

Pretty much...

Intel has such an advance in CPU performances over AMD that they focused this generation on the graphics (that still lag AMD integrated graphics but are much better than previous Intel offerings), and the implementation of a new foundry process (3D transistor to improve efficiency of the CPU) reducing consumption.

The final blow for AMD dominance in the integrated GPU market (for cheap laptop or small form-factor computers) is supposed to be Intel's next generation Haswell, with further improved graphics.

Whether we like it or not CPU single-threaded performances per GHz (most apps aren't well optimized) haven't evolved a lot in the last couple of years and GPUs have become more important with the introduction of OpenCL.

Ivy Bridge is a big leap in that regard since it's Intel's first Integrated GPU supporting it (OpenCL Programs ran on the CPU on Sandy Bridge defeating the purpose of OpenCL), personally that's one of the reasons I've been waiting for Ivy Bridge before making a purchase decision... And I'm still waiting since I was looking more for an iMac or mac mini even ^^.

hashtlt
Jun 12, 2012, 04:18 PM
When do we get a 27" Retina thunderbolt display
how much? and how soon?

Eidorian
Jun 12, 2012, 04:25 PM
Pretty much...

Intel has such an advance in CPU performances over AMD that they focused this generation on the graphics (that still lag AMD integrated graphics but are much better than previous Intel offerings), and the implementation of a new foundry process (3D transistor to improve efficiency of the CPU) reducing consumption.

The final blow for AMD dominance in the integrated GPU market (for cheap laptop or small form-factor computers) is supposed to be Intel's next generation Haswell, with further improved graphics.

Whether we like it or not CPU single-threaded performances per GHz (most apps aren't well optimized) haven't evolved a lot in the last couple of years and GPUs have become more important with the introduction of OpenCL.

Ivy Bridge is a big leap in that regard since it's Intel's first Integrated GPU supporting it (OpenCL Programs ran on the CPU on Sandy Bridge defeating the purpose of OpenCL), personally that's one of the reasons I've been waiting for Ivy Bridge before making a purchase decision... And I'm still waiting since I was looking more for an iMac or mac mini even ^^.On-die eDRAM appears to be next year's magic bullet for the IGP arena. Both Intel and AMD are working on getting a super shader packed IGP that is not crippled by sharing the DDR3 memory bus next year.

Haswell will have improved Ivy Bridge shaders but you are looking at a jump from 16 to a whopping 40 on GT3 and the inclusion of eDRAM. Hueg dies guaranteed.

Torrijos
Jun 12, 2012, 04:28 PM
My macpro (see sig) just got a 12,600.

Looking forward to getting a fully specced retina MBP and seeing the same speeds! Awesome! Probably why they are perhaps killing the pro?

Those scores aren't the whole picture though...

Xeon processors (workstation CPUs used in Mac Pros and concurrent workstations) have more memory channel and better memory throughput, ECC.
So if you work on RAM intensive tasks, Mac Pros, probably, still have an edge. Another important factor is the apparition of PCIe SSD cards that offer WAY better data-rates and latencies, again for specific tasks (heavy DB duties).

There is no way to cheat some of the benefits of Xeon CPUs, or a true workstation, but the question is aren't most DB task run out of Linux servers, the same goes for heavy calculations, and most CAD-like apps are only available on Windows.

Most of the people really annoyed by the stagnant state of the MPro are probably heavy duty Audio or video professionals, they might even be happy with a Xeon iMacPro (or something like that, as long as it offers a lot of RAM and one or two PCIe).

nilk
Jun 12, 2012, 04:53 PM
My macpro (see sig) just got a 12,600.

Looking forward to getting a fully specced retina MBP and seeing the same speeds! Awesome! Probably why they are perhaps killing the pro?

The top-end 12-core Mid-2010 Mac Pro gets 21653, and that is 2 year old hardware. The new top-end Mac Pro naturally will do better than that. I don't think they are killing the pro (especially with the recent rumors of a more significant update in 2013), but it certainly isn't their main focus.

It's amazing that the latest MacBook Air benches 170% faster than my Mac Pro (granted I have the slowest Mac Pro they ever made, the 2006 2.0 GHz 4-core model).

There is still no MacBook that can do 32GB of RAM. I believe the iMac can (unofficially), though. There are non-Apple notebooks that can do 32GB, though, so Apple is a bit behind in this regard.

awesomator
Jun 12, 2012, 05:30 PM
I wouldn't focus on this benchmark too much. I feel at this point the slight variation of processor speed doesn't make a noticeable difference. The noticeable difference with the new retina MacBook pro retina screen will definitely be the blazing hard drive and the improved gpu

Joesg
Jun 12, 2012, 05:47 PM
No benchmark for the Retina with i7-3820QM? I assume it would be same as the regular Macbook Pro or slightly better, similar to the i7-3720QM.

powerslave12r
Jun 12, 2012, 05:51 PM
Somewhat relevant, Intel's tick-tock cycle (http://www.anandtech.com/show/5626/ivy-bridge-preview-core-i7-3770k).

rot@ti.org
Jun 12, 2012, 05:54 PM
The Macbook Air chart is mislabeled. The number for the 1.8 ghz 13-inch model says 6,915. But the original chart from PrimateLabs says 6,197. The relative size of the bar is closer to 6,197 than 6,915.

drewyboy
Jun 12, 2012, 06:08 PM
I'm curious as to what the base 13" MBP would bench w/ 8GB ram. That's what I have in mind is save $300 (Edit: compared to the upper 13") and use $100 for 16GB ram and the remainder for a 256GB Samsung 830... just need to wait for one to go on sale. I figure that'll give a good few thousand increase in the benchmark. Anyone w/ experience of doing this to last years late version?

Shacklebolt
Jun 12, 2012, 06:27 PM
Kind of worried heat is going to be a major issue with the RBPs.

Fast processors notwithstanding, FCPX render times increase substantially at high CPU temperatures (ditto for any heavy-duty CS work). I appreciate that the new assymetric fans "make it seem quieter", but if you want this thing to work optimally for heavy-duty work, you're not going to be able to have it on your lap, couch, or anything else that'll trap heat.

So for any Pro work, unless this thing's cooling system is miraculous (which it may well be -- I certainly hope so), there's a real danger that you'll only be able to use it when it's sitting on a...

...desk...top.

iSayuSay
Jun 12, 2012, 06:33 PM
Intel is more and more advancing on mobile chips. If 2012 iMac benchmark appearing a few time ago turns ou to be accurate, you can see that CPU wise i7 3730 on top end iMac earns no higher than one on MBP.
Means their peak performance is equal.

Still .. GPU is a weak sauce when it comes to laptop. Somethin like 7970M or GTX 690M need much thicker case than any MBP.

cpnotebook80
Jun 12, 2012, 06:44 PM
Just checked my geekbench 6b bit score and its 11108
its a 2008 MAC PRO.. coupled with the 27" LED display. it all looks great to me.
but still, to have power of mac pro in portable size is a unicorn esp one with geekbench score of 21000!! imagine that..maybe one day

gturban
Jun 12, 2012, 06:50 PM
Anyone done a xbench on the new retina laptops? Would be interesting to know what speed the SSD setup will do..

ezekielrage_99
Jun 12, 2012, 07:03 PM
Great performance from the MBA's they'll be awesome for people wanting to get the most out of Facebook and Twitter ;)

Dankex
Jun 12, 2012, 07:43 PM
Isn't it weird that the the 2.3Ghz 15-inch has only 4GB RAM when the previous 13-inch has 4GB?

I have a November 2007 MacBook and im going now to upgrade for the 15-inch (was hoping they would increase the resolution on the 13-inch, like the Air, but, as they did not, i must now go to the 15-inch), but the 4GB RAM is a pain... I think i will buy 2x4GB elsewhere (is that the maximum i can put in?) as it is not that expensive. I think 4GB for Lion is not enough...

Val-kyrie
Jun 12, 2012, 08:03 PM
No 17" model means that I will be left to only used models. I used to do that. Plus it saves dollars. Apple has done their part in keeping my bank account untouched. They finished my want to update with keeping the Mac Pro the longest computer to not keep an update. But at least they are still selling one, unlike the 17" laptop.

I had a 15" PowerBook, but after purchasing the first of at least 4 17" Ma laptops, I refuse to go back in size.

I'll have to check out the Frankentosh market, that Hackintosh to most. No wonder I do not feel welcome at the local Apple Store. I have my own requirements & will not go with what Apple is selling unless it meets those requirements.

What good is a 1,0000,000 X 1,000,0000 pixel display if it is to small to see or have enough useable space to work with?

I give the new Macs 0 Stars at this point. Let's see if they can change that. To make matters worst I've purchased only Macs since they came out in 1984. I always thought that Apple has always been run by the almighty dollar, now I know it for sure.

Is it the size or the resolution that you like most about the 17" MBP? I realize the 15" MBPs all lack an expresscard slot, but the new retina display has piqued my interest with its ability to handle multiple resolutions, including 1920x1200, and a putative reduction in glare. There is an article about the hi-res display over at anandtech.com. I will however reserve my judgment until I am able to see one in person. The proof is in the pudding, as they say.

Edit: I understand your concern about the size of the text, etc., but I am still interested to see how the screen handles different resolutions.

Second post deleted--wrong info.

Digital Skunk
Jun 12, 2012, 09:26 PM
Here are some unboxed pics, forget the performance the retina display is incredible!

Unboxed MacBook Pro with Retina Display (http://blog.soundearphones.com/odds-ends/macbook-pro-retina-display-unboxing/)

That screen is MEGA crisp, even from looking at those compressed images on my iMac I can see how sharp that thing is . . . GEEZE!

That makes a 17" Retina even more drool worthy.

dlbmacfan
Jun 12, 2012, 10:18 PM
Not sure which system they're benchmarking for the 13" Air. I've got the ultimate version (2.0Ghz + 8Gb) and I got a Geekbench score of 7705, which is 10% better than what they're posting.

dr Dunkel
Jun 13, 2012, 12:40 AM
I'd love some sort of full benchmark for Mac, not just the one-eyed Geekbench.

macduke
Jun 13, 2012, 01:12 AM
I'm curious as to what the base 13" MBP would bench w/ 8GB ram. That's what I have in mind is save $300 (Edit: compared to the upper 13") and use $100 for 16GB ram and the remainder for a 256GB Samsung 830... just need to wait for one to go on sale. I figure that'll give a good few thousand increase in the benchmark. Anyone w/ experience of doing this to last years late version?

It won't affect the benchmark in any noticeable way. The Geekbench is really good at testing the processor and memory speeds while ignoring other variables. I've seen machines with SSD vs spinny disks, 4GB vs 16GB ram but have the same CPU and ram speed and the Geekbench hardly discerns between them. It's really effective at isolating the CPU. Now would you notice a big difference in actual use with more ram and a fast SSD? Definitely. But that's for another benchmark. The MacBook Pro R-Type apparently uses the Samsung 830, so for me coming from an early 2011 model with 220MB/s SSD, the difference is pretty huge. I'm also doubling my ram from 8GB to 16GB, so it will hardly even have to page that ridiculously fast 512GB drive. CS6 is going to fly and I can't wait to try it out!

Has anyone seen boot times for these new machines? Bet it's crazy.

donx
Jun 13, 2012, 02:22 AM
I really want to see benchmark 13 inch MacBook Pro Mid 2012 vs 13 inch MacBook Air Mid 2012.

Zulithe
Jun 13, 2012, 08:40 AM
Agh those retina MBPs are extremely pretty... I must stay strong and wait for next year's haswell (http://en.wikipedia.org/wiki/Haswell_%28microarchitecture%29) MBP retinas which will be blazing :|

DribbleCastle
Jun 13, 2012, 09:11 AM
The Macbook Air chart is mislabeled. The number for the 1.8 ghz 13-inch model says 6,915. But the original chart from PrimateLabs says 6,197. The relative size of the bar is closer to 6,197 than 6,915.

Yeah, looks like they fixed it. Makes way more sense now.

BigMcGuire
Jun 13, 2012, 11:12 AM
No kidding! I remember back in the DOS days playing Battle Chess (or something like that...) and it took like 30 seconds to 2 minutes for the computer to think. ;)

LOL, I thought I was the only one who played that with my dad as a kid. :P

macbook123
Jun 13, 2012, 01:27 PM
Can anybody help me understand why the Macbook Air fares so well in comparison to the new Pros? I mean, a score of 7k for a 2-core 2.0 GHz versus a score of 10.8k for a 4-core 2.3 GHz. Wouldn't we expect the latter to be at least twice as fast as the former? What about the scaling prevents that?

I'm seriously reconsidering my Retina MBP order. The only advantages are a bigger and better display and 50% faster CPUs, but at the cost of significant weight, bulk, and price. Hmm...

mariusaz
Jun 13, 2012, 03:11 PM
Now run 3DMark on it at full resolution and post the results.

louden
Jun 13, 2012, 06:13 PM
they just trashed my maxed out mid11 mba

edit
seems that the 5710 to 7004 bump does not register as significant to a couple of you, specifically 22,67%

I'm with you - 22% is a big chunk

----------

I was wrong to think this was a minor spec bump over the 2011MBA. The processors and HD speeds are pretty nice on the new airs.

Upgrading from a 2010MBA, I definitely think it's better to go for the i7 and just use Apple's SSD.

Makes the huge refurb discount on top of the line 2011 airs not so spectacular.

KnightWRX
Jun 13, 2012, 06:17 PM
No kidding! I remember back in the DOS days playing Battle Chess (or something like that...) and it took like 30 seconds to 2 minutes for the computer to think. ;)

The computer still takes that long to think, except today, it has analyzed, classified and created paths for every possible outcome until the end of the game, written it out to a relational database, indexed all the data and created lookup table queries to always select the best path where the outcome is a better chance for it to win.

The DOS computer in that same time had figured out about 2 moves ahead for 3 key pieces.

Moore's law, ain't it grand ? :D

macbook123
Jun 13, 2012, 06:28 PM
Why do the quad-cores sale so poorly compared to the Air? They're only about 1.5 x faster when they should be 2.3 x and 2.6 x faster?

KnightWRX
Jun 13, 2012, 06:49 PM
Why do the quad-cores sale so poorly compared to the Air? They're only about 1.5 x faster when they should be 2.3 x and 2.6 x faster?

How do you figure ?

macbook123
Jun 13, 2012, 06:57 PM
How do you figure ?

(4*2.3 GHz)/(2*2.0 GHz)=2.3

(4*2.6 GHz)/(2*2.0 GHz)=2.6

Gomff
Jun 13, 2012, 07:17 PM
Happy with my late 2011 2.5Ghz Macbook pro with hi res non glossy screen.

I think the new design is beautiful, if thinness is your thing but not practical for me. I travel around a fair bit and use the optical drive often, and I like the bigger screen real estate of the high res screen, despite the lower resolution.

I'm not keen on the soldered on parts and the price.....I think this design Macbook Pro will go through a few nominal iterations which will undo things like glued in batteries etc. Also hopefully the price will come down and the resolution will improve to allow for more screen real estate.

KnightWRX
Jun 13, 2012, 07:41 PM
(4*2.3 GHz)/(2*2.0 GHz)=2.3

(4*2.6 GHz)/(2*2.0 GHz)=2.6

Not exactly how computing works unfortunately. Especially parallel processing...

macbook123
Jun 13, 2012, 09:12 PM
Not exactly how computing works unfortunately. Especially parallel processing...

It is actually how it works in my experience. Hence the question.

nec207
Jun 14, 2012, 03:53 AM
Great performance from the MBA's they'll be awesome for people wanting to get the most out of Facebook and Twitter ;)

Can you elaborate not sure what you trying to say.

The MBA's got big speed bump fron less years MBA's .

libmanj
Jun 14, 2012, 11:33 AM
Does anyone know what configurations the Apple store is stocking of the new Macbook Airs?

Digital Skunk
Jun 14, 2012, 12:00 PM
Does anyone know what configurations the Apple store is stocking of the new Macbook Airs?

Always the base models with no CTOs.

nec207
Jun 14, 2012, 12:19 PM
(4*2.3 GHz)/(2*2.0 GHz)=2.3

(4*2.6 GHz)/(2*2.0 GHz)=2.6

Not sure what you are trying to say. But intel was saying ivy bridge was 20% to 30% fater than sandy bridge.

Like others here doing benchmarks confirm the new CPU's are 20% faster than last year CPU's .

In the Apple keynotes not the CPU but the video part of it Apple say the video card is 60% faster than last year video card.Not sure if people have done benchmarks to confirm this.


It is myth every 18 month CPU double in speed. It every 18 month CPU transistors double thus getting smaller.

It more like CPU double every 3 years than the 18 month myth.

libmanj
Jun 14, 2012, 03:47 PM
Always the base models with no CTOs.

My local Apple store also offered a maxed out version of the Macbook Air 13 as well as the two base configurations. Not sure if they will do something like that this time though.

macbook123
Jun 14, 2012, 11:50 PM
I'm not sure why this is hard to understand. Both the new Air and the new Pro have Ivy Bridge (so your first statement below makes no sense). Given the clock rate of the processors in the Pro, together with the fact that they are quad core, the computer should be 2.3-2.6 times faster than the Air in applications that use all the cores (see trivial math above). However in reality the Pros are a measly 1.5 times faster. So I'm simply asking why the quad cores scale so extremely poorly, making the Pro only slightly faster than the much lighter and slimmer Air.

Not sure what you are trying to say. But intel was saying ivy bridge was 20% to 30% fater than sandy bridge.

Like others here doing benchmarks confirm the new CPU's are 20% faster than last year CPU's .

In the Apple keynotes not the CPU but the video part of it Apple say the video card is 60% faster than last year video card.Not sure if people have done benchmarks to confirm this.


It is myth every 18 month CPU double in speed. It every 18 month CPU transistors double thus getting smaller.

It more like CPU double every 3 years than the 18 month myth.

BayouTiger
Jun 15, 2012, 06:42 AM
Always the base models with no CTOs.

Actually, they usually have a couple CTO configs in stock. I picked up a 13" 2.0/512/8gb last night. Thought I wanted a retina book, but after fondl ing it a bit, it is still too big and heavy. I have done the 15&17 fastest mbp thing in the past, but after having used the 13" air for the past months, I could never go back to a 15". I got the new air for the extra ram, storage, and USB3. I can do the really heavy stuff on my 3.4 iMac. The Air is "fast enough" on the road.

Oh, Moore's law started to fall apart a few years ago. Processor improvements have move to incremental updates rather than revolutionary changes. Biggest development goals are reducing power draw these days.

thermodynamic
Jun 15, 2012, 12:11 PM
The static benchmarks look great, but will a 1~2 FPS improvement matter? Or 4 seconds off a render? Real-life benchmarks will always mean more; static benchmarks are usually better for marketing purposes... not wrongly so...

I wouldn't use static benchmark results as the main selling point (think 'Retina' as being the awesome deal-maker here) as I've still yet to find temperature benchmarks and if the computer throttles over time due to temperature concerns or due to voltage being insufficient (notebookcheck.com found the 2011 models throttled under either condition, but to be fair Ivy Bridge is made with a smaller process and likely requires less electricity to run at, while producing less heat...)

twoodcc
Jun 15, 2012, 03:32 PM
cool. just picked up my 11" Air today! so far so good!

coder12
Jun 15, 2012, 07:36 PM
The computer still takes that long to think, except today, it has analyzed, classified and created paths for every possible outcome until the end of the game, written it out to a relational database, indexed all the data and created lookup table queries to always select the best path where the outcome is a better chance for it to win.

The DOS computer in that same time had figured out about 2 moves ahead for 3 key pieces.

Moore's law, ain't it grand ? :D

One of my Comp Sci teachers actually went over Moore's law this year, I've never actually thought of how it would work for chess though :)

On a side note, it's nice to be able to come to a forum where everyone has so much knowledge to share. I have yet to be let down by the amount of cool topics people bring up!

ysilver
Jun 27, 2012, 11:36 PM
Has anybody noticed how inconsistent the scale is in the bar graph? The 2.0GHz Air and 1.8GHz Air appear to have very different performance based on the bar in the graph but in fact the difference is less than 100 pts.

I don't know about you all, but this kind of mistake makes me seriously question the integrity of these benchmarks. Is the bar wrong or is the number wrong? I would assume the bar is wrong but then why does the 2.0 processor perform within the margin of error of the chip that costs 90 dollars less!?!

ysilver
Jun 27, 2012, 11:50 PM
macbook123, you are completely right to ask this question. Check out the benchmark of the CPU (as opposed to the system benchmark) here: http://www.cpubenchmark.net/high_end_cpus.html

Based on CPUBenchmark, you might expect as much as a 4x difference in performance. Alas, the gains are only marginal for most common computing jobs. This is because most jobs require synchronous execution. This means that a super hefty job can't just be divided up into smaller jobs and allocated to different cores/threads. They have to be executed in order.

Note: it is extremely hard to write good code to take advantage of all the cores/threads on a CPU.

But let me just anticipate your next question: why would one buy the super-powered MBPs? Let's say you are encoding video or running a compiler. These jobs can be effectively divided into asynchronous jobs and take advantage of the multiprocessing environment (as many as 8 concurrent jobs!!!).

Hope this helps. I'm personally struggling with whether I can justify the MBP over Air and I'm a graphics designer/hacker!

I'm not sure why this is hard to understand. Both the new Air and the new Pro have Ivy Bridge (so your first statement below makes no sense). Given the clock rate of the processors in the Pro, together with the fact that they are quad core, the computer should be 2.3-2.6 times faster than the Air in applications that use all the cores (see trivial math above). However in reality the Pros are a measly 1.5 times faster. So I'm simply asking why the quad cores scale so extremely poorly, making the Pro only slightly faster than the much lighter and slimmer Air.

Pipper99
Jun 30, 2012, 02:43 PM
My geek bench 32 bit score is around 6000 consistently when it should be around 10000 to 12000 per the geekbench benchmarks overall. I have a 2.6/8/512. I've run Apple Hardware Test, ClamXAV, and BlackMagic, and all of those tests come out normal: no malware or hardware problems. BlackMagic shows my SSD speed to be around 400 to 450 which seems to be what most retina users are getting. I've turned off graphic switching.

Any ideas?

ysilver
Jul 3, 2012, 08:44 AM
Try turning down the resolution and see if that affects it. I read a report that UI rendering with the MBP Retina on complex pages can throttle performance.

Report back!

My geek bench 32 bit score is around 6000 consistently when it should be around 10000 to 12000 per the geekbench benchmarks overall. I have a 2.6/8/512. I've run Apple Hardware Test, ClamXAV, and BlackMagic, and all of those tests come out normal: no malware or hardware problems. BlackMagic shows my SSD speed to be around 400 to 450 which seems to be what most retina users are getting. I've turned off graphic switching.

Any ideas?

Pipper99
Jul 3, 2012, 09:06 AM
Try turning down the resolution and see if that affects it. I read a report that UI rendering with the MBP Retina on complex pages can throttle performance.

Report back!

Thanks for the info! Apparently, my initial download of the update didn't get installed because when I tried Software Updates again, there was the retina Macbook Pro update. I'm now getting around 11000 on my Geekbench score. :)

Thanks again for the suggestion!

ysilver
Jul 3, 2012, 09:32 AM
Nice!

What did you update?

Thanks for the info! Apparently, my initial download of the update didn't get installed because when I tried Software Updates again, there was the retina Macbook Pro update. I'm now getting around 11000 on my Geekbench score. :)

Thanks again for the suggestion!

Pipper99
Jul 3, 2012, 09:54 AM
Nice!

What did you update?

There's an update specifically for the retina Macbook Pros that you get when you choose Software Update in the Apple Menu. I thought that I had installed it previously, but I either didn't, or I interrupted the installation. It's all good, now! :)

JazzMasta
Jul 30, 2012, 04:05 AM
Upgrading from 1.8 (i5) to 2.0 (i7) costs $100, which is about 4.7% of the cost of the MBA 13" (i5) fully configured. For that you get a 1.3% increase in gb score.

I don't like that math.


Actually, it's a 13% speed increase. For a 6.7% price increment.

Like the math better now?

brianvoip
Aug 20, 2012, 12:50 AM
From a virtualization perspective, 16GB RAM and the 2.7GHz Quad-core Intel Core i7 with 8MB shared L3 cache is worth the extra $$$... to me. :)