Geekbench Score?

harrymatic · Feb 19, 2014

gavinstubbs09 said:
Do you remember how much you paid for the processor? May want to upgrade mine from the 1GHz And do you have the copper heatsink or the metal one?

I paid about £30 for it - these G4s and parts for them seem to hold their value much more in the UK. I'm using the aluminium heatsink. With Nap mode enabled I haven't seen the temperature go above 45C even under really heavy load. With normal stuff it idles at around 38C.

tevion5 · Feb 21, 2014

gavinstubbs09 said:
Mac Pro (1,1 with 8 Cores): 9781
http://browser.primatelabs.com/geekbench3/379860

Power Mac G5 (DP2.0 Early 2005): 1844
http://browser.primatelabs.com/geekbench2/2130897

Damn that 1,1 is a powerful system! Shame they aren't fully supported. It's way stronger than my 2011 MBP which is :/

Zotaccian · Feb 21, 2014

PowerBook G4 1.67GHz 2005 : 964

http://browser.primatelabs.com/geekbench2/2425789

Geekbench version is 2.1.13

eyoungren · Feb 23, 2014

Well, after bumping my processors up to 1.8Ghz I now get 1282!

desantii · Oct 9, 2014

I saw 1470 on friends 1.8 dual 7448

GraniteTheWolf · Oct 9, 2014

I get a 722 on my iMac G4 1.25ghz... not too bad.

Tangerine 300mhz iBook: 155 (ouch!)
Graphite 466mhz iBook: 216 (better...)

My 2012 6-core mac pro gets over 17,000 on the 64 bit test

128keaton · Oct 17, 2014

tevion5 said:
Damn that 1,1 is a powerful system! Shame they aren't fully supported. It's way stronger than my 2011 MBP which is :/

Same boat, running a 2.66 GHz x2 Xeon MacPro1,1 with Yosemite. 7GB of RAM, 500GB HDD, GT210.

Aameiel · Oct 17, 2014

iMac G4 (20-inch Flat-Panel) 748 score =D

supernova777 · Oct 18, 2014

im a registered geekbench2 user
i had been in contact with primate labs regarding the inability to log my scores anymore (The whole reason why i purchased geekbench2)
they were responding but he stopped responding to me,
im kind of annoyed, i wanted to include my g3 B&W on my list of scores..
has anyone else experienced the inability to upload their geekbench2 results ??
it seems to get a 404 error as though the file is no longer on the server to handle the upload.. i told them this and i got no response ;(

i paid to use the program... so what if theres newer versions of programs, the old ones still work and i basically feel ripped off because he didnt even extend to me a geekbench3 license... based on the fact that i am a licensed user of geekbench 2.. and they removed the ability for the app to function.. i feel like i deserve to be autoupgraded to 3... whats worse i dont even really want to use 3.. i want to use 2..

i assume u guys have all experienced the same error when trying to upload scores to your geekbench account?

catzilla · Oct 19, 2014

Rebuilt PB G4 17 1.33, 803. That was on battery, would AC increase the score?
I just realized this $50 ebay special has a Western Digital 250 GB Scorpio Blue.

weckart · Oct 20, 2014

catzilla said:
Rebuilt PB G4 17 1.33, 803. That was on battery, would AC increase the score?
I just realized this $50 ebay special has a Western Digital 250 GB Scorpio Blue.

Well, that paid for itself. I was chuffed enough when my £36 PB 12" had a 160GB Scorpio Blue inside.

catzilla · Oct 20, 2014

PB G4 12" 1Ghz-598.
PB G4 15" 1.67Ghz-972.

Next my Ti and my MDD.

MDD Dual 1.25Ghz-1004.
G4 iMac 800Mhz-455.

Ti is running 10.4.1 I don't think Geekbench will run on it.

bunnspecial · Oct 20, 2014

968 on my DLSD 15"(2gb ram)

1920 on my 2.0 DC G5(10gb ram)

1210 on my single 1.8 G5(4gb ram)

910 on my dual 1ghz MDD(2gb ram)

543 on my 867mhz TiBook(1gb RAM)

flyrod · Jan 30, 2015

Geekbench is a pretty awful measurement tool. Look at a plot of results for the same computer:

This should be a single point. Imagine if your speedometer worked this well.

I've found that nbench is a much better measurement tool. It gives repeatable results even when there are other processes running.

http://www.tux.org/~mayer/linux/bmark.html

There is also a wide range of posted results to compare against, with results from early computers to modern stuff:

http://www.tux.org/~mayer/linux/results2.html

If only nbench had a fancy front end and web site for it...

That being said, here are some of my ppc geek results:

http://browser.primatelabs.com/geekbench2/1688867 1.75mini 1087
http://browser.primatelabs.com/geekbench2/1601644 1.67pb 1012
http://browser.primatelabs.com/geekbench2/1481670 0.55dualcube 580
http://browser.primatelabs.com/geekbench2/1448289 1.92dualcube 1566

I just tried my dual 1.8 7448 cube got a 1544.

Dronecatcher · Jan 30, 2015

I've also found that using a different version of 2.x.x can change the result by as much as 200 marks.

desantii · Jan 30, 2015

Cube 1.8ghz 854....

Altemose · Jan 30, 2015

catzilla said:
Rebuilt PB G4 17 1.33, 803. That was on battery, would AC increase the score?
I just realized this $50 ebay special has a Western Digital 250 GB Scorpio Blue.

Changing the performance settings would. Ideally, it should have clocked up to the load regardless of being on battery or AC power.

Ih8reno · Jan 30, 2015

My most powerful Power Mac that I have left (sold some) is a dual 1.42ghz MDD 2gb ram running Leopard with a 256mb ATI Radeon 9650 with the pins taped off.

Score: 1228

ptdebate · Jan 30, 2015

flyrod said:
I just tried my dual 1.8 7448 cube got a 1544.

Your cube is a god among PowerPC macs! How do you handle cooling?

desantii · Jan 30, 2015

My cube is jealous!

Cox Orange · Jan 31, 2015

flyrod said:
Geekbench is a pretty awful measurement tool. Look at a plot of results for the same computer:...

I agree, this is what I am always saying,... Hence I do real life tasks (like copying a big file and stop the time to complete or convert a video, etc.)

Would be nice, if there was a script that could easily be launched without much command line knowledge, that starts several "real life" tasks.

For example there could be a test file, that can be downloaded and then the script can tell the Mac to copy that or open it in Handbrake and convert it with a strict preset.
A folder with all things needed (script, files of different size, one movie file, a psd photo file,...) could be prepared and presented for download, so that one has everything he needs to do the test.

(Or we would agree on some task that everyone can do and then we report back the times).

There was a graphics card benchmark that should the same results for cards with the same model name, but different clock speeds, that wasn't quiet convincing to me.

flyrod · Feb 9, 2015

ptdebate said:
Your cube is a god among PowerPC macs! How do you handle cooling?

Cooling a dual 7448 is actually pretty easy compared to a dual 7447:

http://www.cubeowner.com/forums/index.php?showtopic=10744

I measured the power use of the dual 47 card at ~100w when loaded. So the 48 is just using the previous heatsink arrangement with a plain base fan now.

To stay on topic: I helped a friend work on a 1GHz 12" powerbook and it ran geek to the tune of 625. The dual 550MHz cube got 580, so I guess their weighting system doesn't give full credit for the 2nd cpu. Comparing other numbers on their web site confirms this when looking at a 1GHz G4 vs an equivalent dual 1GHz g4.