Try the Cinebench 10.5 benchmark. It runs about a minute which makes it easy to use a stopwatch.
-Christoph
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Overclocking Mac Pro
- Thread starter yipster222
- Start date
- Sort by reaction score
Try the Cinebench 10.5 benchmark. It runs about a minute which makes it easy to use a stopwatch.
-Christoph
Yes sure, in version 1.0 we decided to support Mac Pros at first for several reasons:
- They have an outstanding cooling system. Try smcFanControl to listen to the maximum speed of your fans.
- All Mac Pros have similar clock chips. This is unfortunately not the case for MacBooks, iMacs and Mac minis. -> more work for me
However, we plan to release newer versions, which support more Macintoshs.
-Christoph
maybe i'll try it when it get's an update 
but maybe people should specify if they have RAM in dual Channgel or Quad Channel.
Who knows this makes a big difference. In normal tests you can see an increase in performance so then each memory slot should have less change of making a mistake.
cheers and really fun to have this method. You can run a standard 2.8 as fast as a 3.2 that costs 1400+
sweet
don't think apple is gonna like it. Who knows they might even stop this method in their next update.
but maybe people should specify if they have RAM in dual Channgel or Quad Channel.
Who knows this makes a big difference. In normal tests you can see an increase in performance so then each memory slot should have less change of making a mistake.
cheers and really fun to have this method. You can run a standard 2.8 as fast as a 3.2 that costs 1400+
sweet
don't think apple is gonna like it. Who knows they might even stop this method in their next update.
Thanks for the clarification and the tool. I'm able to OC to only about 5-10 MHz over stock on a 3 GHz 1.1 before seeing fbd errors running mprime. I must have some pretty craptacular memory.
maybe i'll try it when it get's an update
but maybe people should specify if they have RAM in dual Channgel or Quad Channel.
Who knows this makes a big difference. In normal tests you can see an increase in performance so then each memory slot should have less change of making a mistake.
cheers and really fun to have this method. You can run a standard 2.8 as fast as a 3.2 that costs 1400+
sweet
don't think apple is gonna like it. Who knows they might even stop this method in their next update.
We tested lots of memory modules in the last few days. Running them in Dual-Channel-Mode tends to allow slightly higher frequencies, max. 10 MHz higher, e.g. 3.248 GHz instead of 3.241 GHz, which isn't much of a difference between quad-channel and dual-channel.
We will see, if Apple is going to like it. They still can sell their 3.2 GHz model. Since the 3.2 GHz processor has a TDP of 120 W, it should easily run at 3.6 GHz.
-Christoph
Just over clocked my 2.8 to 3.02 and itunes is skipping some. Should iTunes (or all) software be restated after you over clock?
Restarted iTunes but it's still skipping forward. Does changing the bus speed effect the playback rate of iTunes (using S/PDIF)?
Restarted iTunes but it's still skipping forward. Does changing the bus speed effect the playback rate of iTunes (using S/PDIF)?
Just over clocked my 2.8 to 3.02 and itunes is skipping some. Should iTunes (or all) software be restated after you over clock?
Restarted iTunes but it's still skipping forward. Does changing the bus speed effect the playback rate of iTunes (using S/PDIF)?
Try rebooting, but not shutting down. Apparently the system clock is overclocked as well until you do that. (won't work on 1st gen Mac Pro).
arn
I have a new 8 core 2.8. So I should over-clock and then reboot and iTunes will play fine? I use my Pro Mac as a music server to my house (using Airports) so it is most important that it serves up the music. Also coming from Vista 64 and I'm within a few hours of breaking my up time on Windows. I have 13D 12H up time and only need another couple of days. Oh well I guess I can break that record later.
Got my 2006 quad 2.66GHz Mac Pro running at 2.8GHz for about an hour, under varying processor and RAM loads, without errors. I did have the fans boosted to about 1300 RPM via smcFanControl which I'm sure helped.
I'm back to running at stock speed, however, because of the time clock issue. And I was unable to successfully (warm) reboot the machine to work around the issue. Perhaps a fix will be discovered.
But kudos to the authors for coming up with this!
I'm back to running at stock speed, however, because of the time clock issue. And I was unable to successfully (warm) reboot the machine to work around the issue. Perhaps a fix will be discovered.
But kudos to the authors for coming up with this!
tool works for me Xp in VM ware reports 3.51ghz on my 3.2 mac pro, however i do notice an instant 4-10 degree jump in temp when exporting with QT. do use with caution and monitor your temps.
I'm scared to run this when my current rig cost over 5K.
Is anyone else out there scared of seeing the magic smoke?
Is anyone else out there scared of seeing the magic smoke?
Here is my method. I have my 2.8 now over clocked to 2.9 and will run this way for a week or so. Then I will go to 3.0 for another week or so, then 3.1,.... Will probably stay at 3.2.
Remember Intel uses the same chips in the 3.2 as in the 2.8. They are just set to a faster clock.
to those of you unsure why people are interested in doing this, let me put it as simply as I can:
If you bought an 8x2.8 but you wanted an 8x3.2 and could not justify the cost difference, you were right to get the 2.8. Because with this utility and decent RAM, you now have a 3.2 for the same price.
If you bought an 8x2.8 but you wanted an 8x3.2 and could not justify the cost difference, you were right to get the 2.8. Because with this utility and decent RAM, you now have a 3.2 for the same price.
Actually, the argument is even more compelling than that. A 2.8 GHz mac pro overclocked to 3.2 GHz is actually *faster* than the equivalent stock 3.2 GHz MP from Apple due to the fact that in an overclocked MP the system and memory bus are overclocked as well, feeding the CPU with data at a faster rate. A stock 3.2 GHz MP only has its multiplier increased, the system and memory bus stay at their original values.
-Robert
Hi there, I just wanted to share my early experiences this evening after downloading and installing the OC tool.
First off, I have a 2008 Oct 2.8, 2GB + 8800GT. I use the Mac for the folding@home project and currently do this under WinXP 32bit. I tweaked the speed up to 3178, applied and rebooted. The screen didn't come back up and I had to do a hard power off.
I re-read the instructions which hinted that you should be able to reboot at this speed. So I figured to lower my clocks a little back to 3150, saved, rebooted and OSX came back up again. So I rebooted again into XP and tested there. CPU says 451 / 1800FSB and once I put a load on the processors, I found the multi-X went from 6 to 7, this gave me 3150.
I setup folding smp in two dos sessions, got Linux running in vmware also running smp folding, got the 8800GT folding. So far so good.
I will report back tomorrow if this crashed etc. Oh and to answer to those people out there why the overclock? In the PC world, we have been able to ge these overclocks for many years. consider it a free upgrade if you're willing to take the time and be patient. I have been trying to overclock my MP since I bought it.
Yes, I do use the power of 8. From me, a big Thanks to the guys who wrote this as it will hopefully help me folding a little quicker now. Also for those really interested in folding, my Mac Pro with the 8800GT now does over 10K PPD alone.
-casw1000
fold4life.com
First off, I have a 2008 Oct 2.8, 2GB + 8800GT. I use the Mac for the folding@home project and currently do this under WinXP 32bit. I tweaked the speed up to 3178, applied and rebooted. The screen didn't come back up and I had to do a hard power off.
I re-read the instructions which hinted that you should be able to reboot at this speed. So I figured to lower my clocks a little back to 3150, saved, rebooted and OSX came back up again. So I rebooted again into XP and tested there. CPU says 451 / 1800FSB and once I put a load on the processors, I found the multi-X went from 6 to 7, this gave me 3150.
I setup folding smp in two dos sessions, got Linux running in vmware also running smp folding, got the 8800GT folding. So far so good.
I will report back tomorrow if this crashed etc. Oh and to answer to those people out there why the overclock? In the PC world, we have been able to ge these overclocks for many years. consider it a free upgrade if you're willing to take the time and be patient. I have been trying to overclock my MP since I bought it.
Yes, I do use the power of 8.
From me, a big Thanks to the guys who wrote this as it will hopefully help me folding a little quicker now. Also for those really interested in folding, my Mac Pro with the 8800GT now does over 10K PPD alone.-casw1000
fold4life.com
Tried the overclocking thingy, but.....
Hi!
Having a 2.0 Ghz Mac Pro I naturally wanted to tried to overclock it, and It sure did go faster, but timing or whatever is making this not useable.
When playing music, the sound skips.
And when I tried playing WOW, all the monsters disappeared.
The console is reporting: WARNING: AppleUSBAudio has detected that clock_get_uptime () value changed radically from previous values.
If there is a "hack" to just change the CPU speed, I'm sure that would make my 2.0 go alot faster.
Hi!
Having a 2.0 Ghz Mac Pro I naturally wanted to tried to overclock it, and It sure did go faster, but timing or whatever is making this not useable.
When playing music, the sound skips.
And when I tried playing WOW, all the monsters disappeared.
The console is reporting: WARNING: AppleUSBAudio has detected that clock_get_uptime () value changed radically from previous values.
If there is a "hack" to just change the CPU speed, I'm sure that would make my 2.0 go alot faster.
My Experience:
It's cool to see everyone having fun with this. BUT!!!
I'm officially giving up on this tool.
My system: I am running 2x2.66 dual cores 8 gigs RAM.
Here's why:
1. There's no way to verify the actual speed increase, or effect this has on your system, other than the syslogs and the temperature of your cores. IE, when you go to "About this Mac", the core speed still reads 2.66, not 3.0, which is where I got my to run stable.
The fact that you can measure speed increases with a "stopwatch", makes no sense to me. If the cores are faster, they are faster, and should have measurable benchmarks that indicate them to be as such.
This is the way it's done (and I have personally done) in the PC community.
2. Due to a lack of a proper bios, overclocking the system beyond its specs causes other strange, seemingly benign errors.
Like one poster mentioned (linked to a wikipedia article), overclocking *my* system sped up my clock drastically. When I woke up this morning, looked at my clock, and realized it was 8:30, I panicked. How am I going to get my train into NY? ****!
Then I looked at my watch, and realized my Mac Pro was running in a time warp, and that I had just run around my house like an idiot, rushing about, because my clock was wrong.
This tool is nothing but an effort from the PC community to make us all incredibly confused. (evil grin)
And yes, I don't care what anyone says, this tool can destroy your computer.
If you run a higher clock speed, you increase heat. Yes, the processors can handle a high level of heat, yes the Mac has a good cooling system, etc. etc. To the person arguing about voltage, yes, you're correct: however, for the sake of argument, the increased clock speed DOES in fact increase core temperatures and subsequently reduce system stability and processor life.
If it wasn't for the Kernel Panics induced by the overclocking, we'd all be screwed.
As someone who has personally fried about five or so processors through overclocking in my day, I can attest to the fact that no, this had nothing to do with voltage, and everything to do with multipliers, FSB speed, and core clock speed.
So, have fun with this tool. But don't miss your train, screw up your timestamps, don't fry your system, and don't think that it's completely harmless to keep getting your system into a kernel panic.
It's cool to see everyone having fun with this. BUT!!!
I'm officially giving up on this tool.
My system: I am running 2x2.66 dual cores 8 gigs RAM.
Here's why:
1. There's no way to verify the actual speed increase, or effect this has on your system, other than the syslogs and the temperature of your cores. IE, when you go to "About this Mac", the core speed still reads 2.66, not 3.0, which is where I got my to run stable.
The fact that you can measure speed increases with a "stopwatch", makes no sense to me. If the cores are faster, they are faster, and should have measurable benchmarks that indicate them to be as such.
This is the way it's done (and I have personally done) in the PC community.
2. Due to a lack of a proper bios, overclocking the system beyond its specs causes other strange, seemingly benign errors.
Like one poster mentioned (linked to a wikipedia article), overclocking *my* system sped up my clock drastically. When I woke up this morning, looked at my clock, and realized it was 8:30, I panicked. How am I going to get my train into NY? ****!
Then I looked at my watch, and realized my Mac Pro was running in a time warp, and that I had just run around my house like an idiot, rushing about, because my clock was wrong.
This tool is nothing but an effort from the PC community to make us all incredibly confused. (evil grin)
And yes, I don't care what anyone says, this tool can destroy your computer.
If you run a higher clock speed, you increase heat. Yes, the processors can handle a high level of heat, yes the Mac has a good cooling system, etc. etc. To the person arguing about voltage, yes, you're correct: however, for the sake of argument, the increased clock speed DOES in fact increase core temperatures and subsequently reduce system stability and processor life.
If it wasn't for the Kernel Panics induced by the overclocking, we'd all be screwed.
As someone who has personally fried about five or so processors through overclocking in my day, I can attest to the fact that no, this had nothing to do with voltage, and everything to do with multipliers, FSB speed, and core clock speed.
So, have fun with this tool. But don't miss your train, screw up your timestamps, don't fry your system, and don't think that it's completely harmless to keep getting your system into a kernel panic.
As you said, it makes the clock go faster. If you measure something with a fast clock it will appear to still be taking 5 minutes but really it now only takes 4 (for example).
