PDA

View Full Version : 'Geekbench 2' Brings Benchmarking to the App Store




MacRumors
Jun 21, 2010, 10:22 PM
http://www.macrumors.com/images/macrumorsthreadlogo.gif (http://www.macrumors.com/iphone/2010/06/21/geekbench-2-brings-benchmarking-to-the-app-store/)


http://images.macrumors.com/article/2010/06/21/232107-geekbench_iphone_1.jpg

http://images.macrumors.com/article/2010/06/21/232107-geekbench_iphone_2.jpg

Primate Labs, the company behind the popular Geekbench (http://www.primatelabs.ca/geekbench/) benchmarking application for a number of operating systems, has now come to the App Store with the release of Geekbench 2 [App Store (http://itunes.apple.com/us/app/geekbench-2/id377657310?mt=8), $4.99]. Compatible with the iPhone and iPod touch, Geekbench 2 utilizes the same benchmarks (http://www.primatelabs.ca/blog/2010/06/geekbench-2-for-iphone/) as those used on other platforms.What’s great about Geekbench 2 for iPhone is that it runs the same benchmarks as Geekbench 2 for other platforms, which means it’s possible to compare benchmark results not just between iPhones and iPods, but also between iPhones and iPods and Macs and PCs!The application also integrates with Geekbench Result Browser, allowing for easy uploading of results for comparison with other users. While many users are likely to balk at the application's $4.99 price tag, the Mac OS X and other traditional versions of Geekbench carry a price tag of $19.95 for full unlocking of the application's features.

Back in 2007 after the release of the original iPhone, Primate Labs did compile (http://www.macrumors.com/iphone/2007/08/16/geekbench-on-the-iphone/) a basic version of Geekbench for the iPhone and shared the results, but the new Geekbench 2 represents the company's first App Store release of its benchmarking tool.

Article Link: 'Geekbench 2' Brings Benchmarking to the App Store (http://www.macrumors.com/iphone/2010/06/21/geekbench-2-brings-benchmarking-to-the-app-store/)



Eric.
Jun 21, 2010, 10:43 PM
What's the point in having a benchmark tool on a platform that will always have the same capability?? I guess its nice to compare between every variation of iPhone and iPod Touch hardware, but for an individual I don't see much use for $5.

appleguy123
Jun 21, 2010, 10:44 PM
I'm confused. Won't it give the same number on each model? Only one person per device needs to buy it.

sonicwind
Jun 21, 2010, 11:00 PM
Interesting that it has processor speed and memory. I thought those stats were only available from prohibited APIs. And yeah, considering there is absolutely nothing anyone can do to change their results other than running in a freezer or an oven. $4.99 seems over priced by about $4.99.

Anonymous Freak
Jun 21, 2010, 11:56 PM
Yeah, the only variables I can see are making full use of multitasking, and running on Jailbroken devices; to see how much of an impact those have.

kockgunner
Jun 22, 2010, 01:05 AM
I'm confused. Won't it give the same number on each model? Only one person per device needs to buy it.

Lol exactly.

ethernet76
Jun 22, 2010, 01:06 AM
Interesting that it has processor speed and memory. I thought those stats were only available from prohibited APIs. And yeah, considering there is absolutely nothing anyone can do to change their results other than running in a freezer or an oven. $4.99 seems over priced by about $4.99.

You don't need to access prohibited APIs.

Since the memory and processor speeds are the same, you only need to know the model. You could even just detect available APIs to determine what the model is.

IE gyroscope == 4, compass, but no gyroscope == 3gs, no compass, but gps == 3g, no gps == original iPhone.

opticalserenity
Jun 22, 2010, 01:09 AM
I agree, what a stupidly priced app. Yes, it's great they brought this to the iPhone, but wouldn't have it been smart for them to actually release this for free and then tell people that there is a desktop version for OS X so you can test the rest of your systems?

Wow.

iphones4evry1
Jun 22, 2010, 01:15 AM
So, here's what should be benchmarked:

Speed (time it takes to complete certain tasks) of iPhone 3G before updating to iOS 4,

and speed of iPhone 3G after updating to iOS 4.

doctoree
Jun 22, 2010, 03:33 AM
The problem with Geekbench in general is thats it's engine is totally outdated, makes ZERO use of OpenCL and your GPU to speed up its measurements.

Kenndac
Jun 22, 2010, 05:16 AM
Wait… this was approved? Wow. I thought Apple doesn't publicly disclose this kind of information (CPU, RAM, etc) for a reason — users shouldn't have to know (or care) what specs their phone has. I'm very surprised this made it through.

Fraaaa
Jun 22, 2010, 05:27 AM
So, here's what should be benchmarked:

Speed (time it takes to complete certain tasks) of iPhone 3G before updating to iOS 4,

and speed of iPhone 3G after updating to iOS 4.

Unfortunaly iOS 4 only :(

ltcol266845
Jun 22, 2010, 07:15 AM
So, here's what should be benchmarked:

Speed (time it takes to complete certain tasks) of iPhone 3G before updating to iOS 4,

and speed of iPhone 3G after updating to iOS 4.

Spoiler Alert: It makes the iPhone 3G sooooo slow by eating up even more of the phone's RAM. I used to run somewhere around 15-30MB of free RAM on 3.1.2, but with iOS4, I'm lucky to have 5MB free even after a reboot :(

RedTomato
Jun 22, 2010, 08:41 AM
Does this actually measure phone speed or just call up the figures from a list of iPhone models?

It would be incredibly useful if it did measure speed, as I'd like to know what effect things like having Location Services or mail.app running in the background have on the phone speed.

I have an original iphone 2G and I have to turn off as much as I can to keep the speed up.

By next year, when iPhone 5 and a more heavy iOS 5 come out, millions of people with the 'old' iPhone 4 will want to know the impact of multitasking and other background services as well as the new iOS 5 on their phone.

ugp
Jun 22, 2010, 09:01 AM
269

iPhone 3GS 16GB iOS 4.0 after normal usage of several Apps. Doing a Hard Reset right now.
http://browse.geekbench.ca/geekbench2/view/262270

After Hard Reset my score dropped to 261.
http://browse.geekbench.ca/geekbench2/view/262277

mabaker
Jun 22, 2010, 09:32 AM
Last generation Powerbooks get still respectable 800-900 points.

iPhone has tremendous performance in one tiny package but a far cry from a Powerbook.

pmz
Jun 22, 2010, 09:54 AM
any iPad results yet?

Haoshiro
Jun 22, 2010, 10:10 AM
You can't bench the iPad, as it is not running iOS4

iPod2,1 - Score 198
iOS 4.0
Processor 532 Mhz
Memory 116 MB

Halon X
Jun 22, 2010, 10:38 AM
269

iPhone 3GS 16GB iOS 4.0 after normal usage of several Apps. Doing a Hard Reset right now.
http://browse.geekbench.ca/geekbench2/view/262270

After Hard Reset my score dropped to 261.
http://browse.geekbench.ca/geekbench2/view/262277

I'll be with water cooling you could hit 300...

iSee
Jun 22, 2010, 11:14 AM
As this rolls out to other devices this will be more interesting.

It would be fun to compare, e.g., evo to iphone 4, etc.

(Of course this is a synthetic benchmark that ignores, e.g, the effect the app SDK and OS has on real-world performance of app, but all benchmarks have their strengths and weaknesses -- you just need to understand them to understand the results.)

It's too bad about the limitation to iOS 4. I would love to see the effect of different OS's would have. Also, I thought iOS 4 won't install at all on older phones and touches.

firewerk
Jun 22, 2010, 12:37 PM
I'll be with water cooling you could hit 300...

Hahaha definitely made me laugh.

It does seem pointless on an un-customizable device... They could have just told us the results instead of spending time on the App Store.

hogo
Jun 22, 2010, 03:25 PM
This would be useful if you could measure other systems with it. Like an external geek benching device that you could plug in to other computers via usb and test their systems.

dukebound85
Jun 22, 2010, 03:31 PM
dont see the point. cant upgrade the phone so id rather just read results of others who run this app

wolfenkraft
Jun 22, 2010, 03:37 PM
dont see the point. cant upgrade the phone so id rather just read results of others who run this app

Exactly.

RedTomato
Jun 22, 2010, 04:33 PM
I reread the article, and it does seem they're running a benchmark suite.It does seem pointless on an un-customizable device... They could have just told us the results instead of spending time on the App Store.

iPhone is very much customizable. You can turn various background services on and off. BT, wifi, location services, push notifications, music streaming etc etc. All of them take up some system resources and slow down other apps. Exactly how much is open to debate.

This would be useful if you could measure other systems with it. Like an external geek benching device that you could plug in to other computers via usb and test their systems.

Already have. It runs a standard geekbench suite. You can download the free version to your PC or Mac and run that and compare the numbers directly (in theory).
dont see the point. cant upgrade the phone so id rather just read results of others who run this app
See above. You won't know what background services they've got running unless they say so. If you're interested in speeding up your own phone for your own mix of background apps, this will help. Safe prediction: by mid 2011 or 2012, people will be complaining how slow and weak the iPhone 4 is.
269 iPhone 3GS 16GB iOS 4.0

iPod2,1 - Score 198
iOS 4.0
Processor 532 Mhz
Memory 116 MB

Interesting. Wild guess - I expect the iPhone 4 to score about 400. Maybe even 500. My 3 year old Macbook 2.2ghz scores around 2600.

400-500 is a decent score for something that's 1/4 of the size and weight of just my macbook battery. The way things are going, iPhones will be as powerful as laptops in 5 or 6 years time.

Benchmarking a phone. Huh. 3 years ago I would have laughed at the idea.

iMacDragon
Jun 22, 2010, 05:10 PM
Interesting. Wild guess - I expect the iPhone 4 to score about 400. Maybe even 500. My 3 year old Macbook 2.2ghz scores around 2600.

400-500 is a decent score for something that's 1/4 of the size and weight of just my macbook battery. The way things are going, iPhones will be as powerful as laptops in 5 or 6 years time.

Benchmarking a phone. Huh. 3 years ago I would have laughed at the idea.

Considering 3GS is ~260 give or take, I'd not expect it to jump that high. Though, if the memory performance is as increased as rumoured, I guess it just might.

Mystikal
Jun 22, 2010, 07:17 PM
Someone with an iPhone 4 needs to run this now!!!

Burglnar
Jun 22, 2010, 11:34 PM
Someone with an iPhone 4 needs to run this now!!!

^+1!!

One of you lucky early delivery folks who have $5 to waste, let us know what you score with no apps running (clear the taskbar) and one with all 8 save states filled- preferably with various system and entertainment apps like pandora etc..

Anyone?? Please! I can't wait another day and half! Haha

RedTomato
Jun 23, 2010, 07:06 AM
Considering 3GS is ~260 give or take, I'd not expect it to jump that high. Though, if the memory performance is as increased as rumoured, I guess it just might.

You might be right. I'll scale my guess back to 350-odd. Most likely 340.

The bottom end of http://www.primatelabs.ca/blog/mac-benchmarks/ is very interesting. Seems the iPod is already more powerful than all G3 macs, and the iPhone 4 looks to be about the same as a PowerMac G4 533MHz. If the benchmarks can be compared of course.

As someone said in another thread, 1GHz processor, 512MB ram, 100MHZ bus, where have we heard this before?

Nevertheless, 50% processor power increase per year per phone model for 3 years straight is not bad. Hot stuff.

AJerman
Jun 23, 2010, 10:55 AM
Spoiler Alert: It makes the iPhone 3G sooooo slow by eating up even more of the phone's RAM. I used to run somewhere around 15-30MB of free RAM on 3.1.2, but with iOS4, I'm lucky to have 5MB free even after a reboot :(

What?! You might want to get your phone checked out if that's the case. My 3G got 50-60 MB of free memory at a max before iOS 4, and with iOS 4 (with multitasking, wallpapers and the battery percentage all enabled) I currently have 55 MB free.

ericschmerick
Jun 24, 2010, 02:21 AM
Interesting. I am actually tempted fork over 4.99. I think that makes me an idiot.

I guess they'll make their money mostly off guys like me, who get that all the phones are _supposed_ to be the same and not customizable, but "surely my phone is running slower than normal - seems a little slower than Bob's over there". Just want to download it, confirm that I get the same score, and then probably never run it again. :)

Anyway, I have the original 3G. It's painful to see that score of 138. The 3GS is something like 260? iPhone 4 is ~ 350 or something? Ug, no wonder my phone feels so slow running the same OS as those other two. That plus the big fat 116MB of ram, vs 256 and 512.

How painful. I need a new phone.

RedTomato
Jun 24, 2010, 05:44 AM
Geekbench scores for the iPhone 4 are on the front page.

http://www.macrumors.com/2010/06/23/iphone-4-is-faster-than-3gs-and-slower-than-ipad-in-early-benchmarks/

350-odd. Respectably faster than the 3GS and about the same score as a Pentium 3, or Celeron or early G4 mac.

derek4484
Jun 24, 2010, 02:42 PM
Interesting that it has processor speed and memory. I thought those stats were only available from prohibited APIs. And yeah, considering there is absolutely nothing anyone can do to change their results other than running in a freezer or an oven. $4.99 seems over priced by about $4.99.

Exactly. You buy it for $5, then run it once and find out what your score is and then you never need it again. Seems pretty steep. Maybe should be $0.99.