PDA

View Full Version : Geekbench scores showing up for i5 + i7




derlockere
Jul 21, 2011, 06:25 AM
Here are some Geekbench scores I found:

Model numbers seem like: Macbook Air 4,1 = 11" / Macbook Air 4,2 = 13"

11":
i7:http://browse.geekbench.ca/geekbench2/view/441499 (32-bit)
5767

i5:http://browse.geekbench.ca/geekbench2/view/441114 (64-bit)
5032

13":
i7: http://browse.geekbench.ca/geekbench2/view/441437 (32-bit)
5835

i7: http://browse.geekbench.ca/geekbench2/view/441556 (64-bit)
6316

i7 via user scottlu13: http://forums.macrumors.com/showpost.php?p=12992496&postcount=40 (32-bit)
5814

i5:http://browse.geekbench.ca/geekbench2/view/441093 (64-bit)
5879

I didn't find more results.
Does anybody know what's the difference between the 32-bit and the 64-bit test? edit: It seems about 10% --> that could be a rather steep margin in CPU speed regarding i5 vs. i7 in the 11"



OSMac
Jul 21, 2011, 06:33 AM
The 13" i5 looks interesting so far, hope it runs silent most of the time.

Rankrotten
Jul 21, 2011, 06:34 AM
Build 11A2063 of Lion in the Airs

dixido
Jul 21, 2011, 06:42 AM
another i7: http://browse.geekbench.ca/geekbench2/view/441556



11" or 13"?

jabooth
Jul 21, 2011, 06:54 AM
i7: http://browse.geekbench.ca/geekbench2/view/441556 (64-bit)
6316


have a quick browse on here (http://www.primatelabs.ca/geekbench/mac-benchmarks/#64bit) - the new i7 Air beats the 17" MBP Early 2010 (6273). Maan, might have to stretch to the 13" i7 :P

Oppressed
Jul 21, 2011, 06:57 AM
Why is the i5 on the 11 inch doing to poorly? Not nearly as big of a gap as the i5 and i7 on the 13 inch model.

trondah
Jul 21, 2011, 07:07 AM
The i5 on the 13" turbo boosts to 2.7 GHz while the 11" boosts to 2.3 GHz.

I wouldn't call the score poor though, I seem to remember my previous 2.4 GHz C2D MBP got like 3700 in geekbench.

OSMac
Jul 21, 2011, 07:16 AM
Found this one as well...

11" i5 .... 5388 64bit

http://www.youtube.com/watch?v=Y5ESp42stwE

KnightWRX
Jul 21, 2011, 07:20 AM
Guys, Geekbench scores are worthless. You can get the same information by looking up where the part number fits in Intel's part list.

It's simply a measure of pure CPU+memory performance.

Typswif2fingers
Jul 21, 2011, 07:24 AM
This totally revealed stupid question, has been proudly brought to you, by me.

:)

Sorry guys, but what do these numbers mean?

arctic
Jul 21, 2011, 07:34 AM
Guys, Geekbench scores are worthless. You can get the same information by looking up where the part number fits in Intel's part list.


So true. Geekbench is hogwash. I recall, just a few hours after the 2010 MBA's were out, a front page article also boasting the Airs to equal or better the MBPs. Yet we didnt see all MBA owners boasting their MBAs kicking butts against MBP performances. What I'd love to see though is for the new MBA's to be included in practical tests just like these:

http://www.macworld.com/article/157893/2011/02/2011macbookpro_benchmarks.html#lsrc.mod_rel

KnightWRX
Jul 21, 2011, 07:34 AM
This totally revealed stupid question, has been proudly brought to you, by me.

:)

Sorry guys, but what do these numbers mean?

Stuff with faster CPUs score higher numbers. Really, that's all these numbers mean. Again, a quick glance at Intel's part list could tell you the same thing.

MattZani
Jul 21, 2011, 08:29 AM
Still waiting for a 13" i5 32bit test, so I can see how it really matches up to my MBP!

fotuwe
Jul 21, 2011, 08:35 AM
What about some Xbench results?

MBABuyer
Jul 21, 2011, 09:23 AM
Will I notice the difference in the i5 vs i7?

I am going to get 256gb no matter what, but I am either going to the store to get the 256gb i5...or waiting till next week for 256gb i7.

I am going to use this for email, internet, school work, projects, no heavy gaming, light imovie occasionally, and videos online.

What difference will the i7 make compared to the i5?

Thanks,
Swayne

ghsNick
Jul 21, 2011, 09:43 AM
I've been reading that the i5 in the 11" is worse than the i5 in the 13"

*Score Wise*

Can anyone confirm or deny this?

thewalkman
Jul 21, 2011, 09:48 AM
The 11'' i5 has 1.6GHz, the 13'' i5 1.7GHz.

MacRumorUser
Jul 21, 2011, 10:15 AM
The 11'' i5 has 1.6GHz, the 13'' i5 1.7GHz.

Alongside the fact that the 1.6 turbo to 2.3 whilst the 1.7 to 2.7ghz

jimboutilier
Jul 21, 2011, 10:37 AM
What about some Xbench results?

I'm with you. Looking forward to XBENCH results.

No benchmark is completely valid for all users but Geekbench is pretty much meaningless for a feel for a systems overall speed. Xbench isn't perfect but a lot better. Enough real user reports in combination with a variety of benchmarks are best.

I know if I bought a 2011 MBA it wouldn't seem twice as fast even though the CPU may be twice as fast. Lots of other components and thee way they interact are just as important. I'll be impressed if XBENCH scores go up 20% over what I'm currently getting (score=140).

rumz
Jul 21, 2011, 11:06 AM
A couple more i7 1.8ghz geekbench scores (64-bit):

http://browse.geekbench.ca/geekbench2/view/441841

http://browse.geekbench.ca/geekbench2/view/441829

I just did a clean install of Snow Leopard on my 15" early 2006 2.16ghz Core Duo MBP... I'll have to see what GeekBench reports. The 32-bit tests in their Mac results browser has that machine at roughly 2600 points. Sounds like my new MBA will punish it soundly ;)

Bob Coxner
Jul 21, 2011, 11:28 AM
What about some Xbench results?

Here's one at 263 http://db.xbench.com/merge.xhtml?doc2=541209

bp1000
Jul 21, 2011, 12:13 PM
I got my new mba earlier today and ran geekbench

13" i5 1.7ghz

32bit
5470

64bit
5853

drewyboy
Jul 21, 2011, 12:25 PM
Looks like on xbench base 13" airs are beating base 13" pro's. That's good news. I can't afford anything more than the base 13". It's just too much to justify going for the 256 for $300 more. $300 buys a lot of external :)

nebulos
Jul 21, 2011, 12:38 PM
So true. Geekbench is hogwash. I recall, just a few hours after the 2010 MBA's were out, a front page article also boasting the Airs to equal or better the MBPs. Yet we didnt see all MBA owners boasting their MBAs kicking butts against MBP performances. What I'd love to see though is for the new MBA's to be included in practical tests just like these:

http://www.macworld.com/article/157893/2011/02/2011macbookpro_benchmarks.html#lsrc.mod_rel


2010 MBA as fast as 2010 MBP? NO.

this is an extremely common mistake, one that was very convenient for 2010 MBA marketing:

the test you're referring to is not geekbench, it's a test 'suite', called speedmark (i believe), which involves several different tests, including file transfer speeds and other operations that are disk-bound, which is why the SSD on the Air helped it score comparably, when averaging over all tests, to the MBP.

geekbench is purely CPU and RAM (has nothing to do with disk speed).

***************

EDIT: Sorry, I guess I misunderstood your post. I don't know what article you first referred to. It reported comparable geekbench scores for the 2010 MBA and MBP? the Macworld article is the one I was referring to, which seemed to me to be the one that planted the idea that "2010 MBA ~ 2010 MBP" for 13" models.

***************

EDIT 2: In fact, I guess the Macworld article does show that for the 2010 13" models, MBA was, overall, or, on average, about as 'fast' as MBP. The crux of the issue is the meaning of 'fast'. Some tasks are disk bound, some are CPU bound. We just have to be careful when reporting/interpreting test results.

If we perform a new test on Macs where we take each computer and toss it out the window of a moving van, I think the first gen MBA may easily be 'faster' than the upcoming Mac Pros.

nebulos
Jul 21, 2011, 01:20 PM
somehow i was smart enough to miss/forget that this is the first post in the thread! ... i got caught up looking up scores and started organizing them, then i looked back up and struck myself on the forehead. anyways, for what it's worth:


13" Samsung Series 9

1.4GHz i5-2537M, Hyperthreading, Turboboost 2.3GHz

32-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=2537M+32-bit&commit=Search): ~3800

64-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=2537M+64-bit&commit=Search): 4500 (only one score)


base 11" 2011 MBA

1.6GHz i5-2467M, Hyperthreading, Turboboost 2.3GHz

32-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=2467M+32-bit&commit=Search): ~4600

64-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=2467M+64-bit&commit=Search): ~5000


11" 2011 MBA with CPU upgrade

1.8GHz, i7-2677M, Hyperthreading, Turboboost 2.9GHz

32-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=macbookair4%2C1+1.8+32-bit&commit=Search): ~5800

64-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=macbookair4%2C1+1.8+64-bit&commit=Search): 6200 (only one score)


base 13" 2011 MBA

1.7GHz i5-2557M, Hyperthreading, Turboboost 2.7GHz

32-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=2557M+32-bit&commit=Search): ~ 5400

64-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=2557M+64-bit&commit=Search): ~5900


13" 2011 MBA with CPU upgrade

1.8GHz, i7-2677M, Hyperthreading, Turboboost 2.9GHz

32-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=macbookair4%2C2+1.8+32-bit&commit=Search): ~5800

64-bit Geekbench (http://browse.geekbench.ca/geekbench2/search?q=macbookair4%2C2+1.8+64-bit&commit=Search): ~6300


... some outlier scores were ignored. we'll see how the averages settle over time.

stylinexpat
Jul 21, 2011, 01:53 PM
My 2010 MBP 13" with the 2.66 processor came in at 3739 but I upgraded my stock HDD to a Vertex 2 SSD so performance seems quite good. This new 13" MBA has some really good numbers coming out.

http://www.primatelabs.ca/blog/2010/04/macbookpro-benchmarks/

mark28
Jul 21, 2011, 02:27 PM
Geekbench = crap.

MBP scores higher than iMac's at Geekbench, yet in real world performance even the base $1200 iMac beats the high end $2500 MBP.

dagamer34
Jul 21, 2011, 03:33 PM
Guys, do something like this: http://att.macrumors.com/showthread.php?t=1111126

OSMac
Jul 21, 2011, 03:49 PM
Guys, do something like this: http://att.macrumors.com/showthread.php?t=1111126

Too bad it takes 2hrs to download?

jimboutilier
Jul 21, 2011, 03:57 PM
Here's one at 263 http://db.xbench.com/merge.xhtml?doc2=541209

Sadly, not a full run of all parts. Looks like the user only did the CPU test not all the other parts of the system. So this is no more useful than a geek bench run.

nebulos
Jul 21, 2011, 04:11 PM
Geekbench = crap.

MBP scores higher than iMac's at Geekbench, yet in real world performance even the base $1200 iMac beats the high end $2500 MBP.

some MBPs score higher than some iMacs. can you show us what real world performance you're referring to where the base iMac beats the high end MBP?

mabovb
Jul 21, 2011, 04:21 PM
I know the different between 32 and 64bit, but are the two different models for both the 11" and 13"?

For instance a 32bit 11", 64bit 11", 32bit 13", 64bit 13"?

How do we tell at the store which one is which?

derlockere
Jul 21, 2011, 04:33 PM
I know the different between 32 and 64bit, but are the two different models for both the 11" and 13"?

For instance a 32bit 11", 64bit 11", 32bit 13", 64bit 13"?

How do we tell at the store which one is which?

The 'bits' refer only to the mode the geekbench test was executed. the 32-bit geekbench test results in lower scores. In fact there is no difference in the Macbook Air models

nebulos
Jul 21, 2011, 04:45 PM
TB allows for short bursts of CPU power. what about steady CPU power?

my concern is that, while turboboosting seems to allow these CPUs to score very high on geekbench, the test only lasts a few seconds.

i don't know much about TB, but what i'm wondering is if we can expect steady CPU power when needed. for example, running audio software that taxes the CPU, will the CPU be able to keep TB up or does the heat eventually (if not quickly) force the CPU to clock back down?

... it's something i can look up, i know, and i will, but if you know and feel like answering, it might help the thread as well.

HiRez
Jul 21, 2011, 05:29 PM
You guys are killing me, I was going to wait and order later, but I really think I have to go down to the Apple Store after work now and pick one up. :mad:

tirerim
Jul 21, 2011, 05:38 PM
TB allows for short bursts of CPU power. what about steady CPU power?

my concern is that, while turboboosting seems to allow these CPUs to score very high on geekbench, the test only lasts a few seconds.

i don't know much about TB, but what i'm wondering is if we can expect steady CPU power when needed. for example, running audio software that taxes the CPU, will the CPU be able to keep TB up or does the heat eventually (if not quickly) force the CPU to clock back down?

... it's something i can look up, i know, and i will, but if you know and feel like answering, it might help the thread as well.

As I understand it, TurboBoost has nothing to do with short bursts of power -- it disables one of the cores in order to divert more power to the other one and let it run faster. So you get one core at 2.3 GHz instead of two cores at 1.6 GHz, or whatever the numbers actually are. For non-parallel tasks, this is very useful, while for tasks that can make use of both cores the chip operates in normal mode and you get better speed that way. The chip doesn't use more power in either mode, though, so the heat should be the same.

KPOM
Jul 21, 2011, 05:41 PM
As I understand it, TurboBoost has nothing to do with short bursts of power -- it disables one of the cores in order to divert more power to the other one and let it run faster. So you get one core at 2.3 GHz instead of two cores at 1.6 GHz, or whatever the numbers actually are. For non-parallel tasks, this is very useful, while for tasks that can make use of both cores the chip operates in normal mode and you get better speed that way. The chip doesn't use more power in either mode, though, so the heat should be the same.

There is also a Turbo Boost feature that works with 2 cores. It doesn't boost as high, but it does boost. I'm not sure when and how that works.

LeakedDave
Jul 21, 2011, 07:04 PM
http://browse.geekbench.ca/geekbench2/view/442569

2011 13" MacBook Air i7 1.8GHz

64 bit test: 6918

MrMacAir
Jul 21, 2011, 11:24 PM
The new MacAirs do NOT have either Turbo or Hyper as Apple has disabled both which makes the tests impressive. You can check their store and note they do not show it on the tech specs. Probably, it is a heat issue. Also, I assume you are aware the camera is not HD.

Do you feel the i7 results justify the additional $100 for the 13"?

nebulos
Jul 21, 2011, 11:39 PM
The new MacAirs do NOT have either Turbo or Hyper as Apple has disabled both which makes the tests impressive. You can check their store and note they do not show it on the tech specs. Probably, it is a heat issue. Also, I assume you are aware the camera is not HD.

Do you feel the i7 results justify the additional $100 for the 13"?

as far as i know, Apple does not list these features on their specs page, but that doesn't mean they were disabled. in fact, look at this thread:

MBA 13" 256 GB i7 - facts and figures - now with HyperThreading and TurboBoost (http://forums.macrumors.com/showthread.php?t=1193767&highlight=hyperthreading)

MTD's Mac
Jul 22, 2011, 01:28 AM
The 13" i5 looks interesting so far, hope it runs silent most of the time.

I've been pushing my i5 all day, it rarely gets the fans going. Much cooler/quieter than my 2010 MBP.

Mikael
Jul 22, 2011, 01:39 AM
The new MacAirs do NOT have either Turbo or Hyper as Apple has disabled both which makes the tests impressive. You can check their store and note they do not show it on the tech specs. Probably, it is a heat issue. Also, I assume you are aware the camera is not HD.
Why do you write something completely baseless like this and phrase it like it's facts? It's not, the new Airs definitely have HyperThreading and all info I've seen so far indicates that turbo is working as well.

DavidC1
Jul 22, 2011, 02:23 AM
TB allows for short bursts of CPU power. what about steady CPU power?

There's a "Long boost" and a "Short boost". Though on the 17W chips, even the Long boosts are quite high.

Long Boost: lasts forever, or until emergency thermal limitations kick in, this is the typical clock speed you know
Short Boost: ~60 seconds

I'll just list the MBA 2011 ones

Core i5 2467M

Base: 1.6GHz
2C Short boost: 1.9GHz
2C Long boost: 2.0GHz
1C: 2.3GHz

Core i5 2557M

Base: 1.7GHz
2C Short boost: 2.3
2C Long boost: 2.4GHz
1C: 2.7GHz

Core i7 2677M
Base: 1.8GHz
2C Short boost: 2.5GHz
2C Long boost: 2.6GHz
1C: 2.9GHz

(2467M speeds are guesses for Long and Short)

nebulos
Jul 22, 2011, 02:37 AM
^ thanks!

would like to see some benchmarks at 'long boost' speeds.

stylinexpat
Jul 22, 2011, 03:03 AM
I've been pushing my i5 all day, it rarely gets the fans going. Much cooler/quieter than my 2010 MBP.

Try opening Skype with a Video call for say 15-20 minutes;)