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.
Upgrading MP 1,1 CPU
- Thread starter Omnius
- Start date
- Sort by reaction score
You simply swap the processors. It will still say it's a 1,1 mac pro though - I'm not sure if flashing anything might change this but it won't make any difference to the performance of the computer.
Some people have reported problems after the swap until they flashed to 2,1 firmware. Others - like myself - did not experience problems with the 1,1 firmware. I suspect it may be connected to CPU's with older steppings being used by the people who where forced to flash.
If you want to be on the safe side, flash before you swap.
In either case you should go for CPU's with G0 stepping.
If you want to be on the safe side, flash before you swap.
In either case you should go for CPU's with G0 stepping.
I did an upgrade to BSEL 1.86 processors, and while the daily routine stuff like iCal, Safari, Mail, Acrobat, etc. isn't a dime's difference, all the high-processing stuff definitely improved. Video work was cut in half after the upgrade.
Geekbench went from circe 5100 to circa 8000, if I remember right. I think I actually got much higher in 64-bit.
If you can upgrade some cheap processors and BSEL mod them, you can really zoom. I'll probably do the 5365 upgrade in another year or two and keep this tower as a spare machine for side working and as a video server.
If I'm not mistaken, BSEL mod doesn't work for the x5355 or x5365 correct?
----------
I was interested in some sort of benchmark scores. The difference between 4 cores and 8 cores will be substantial enough for my purposes but I was thinking a benchmark difference between
4 core 2.66
8 core 2.66
8 core 3.0
----------
I was interested in some sort of benchmark scores. The difference between 4 cores and 8 cores will be substantial enough for my purposes but I was thinking a benchmark difference between
4 core 2.66
8 core 2.66
8 core 3.0
As you mentioned Geekbench i thought you'd be aware of Primatelabs' website with its Geekbench result browser.
The 4-core 5150 (i.e. standard 1,1) scores on average 5.289, the 4-core 5160 @3GHz averages at 5.912 and the 8-core x5365 (3GHz) averages at 9.769.
I could not find official results for the x5355@2.66GHz (they may be there though - did not search thoroughly), but i can tell that those were just shy of 9.000 (32bit) on my own 1,1. I once managed to get around 9.300 somehow (no programs running), but i'd say a solid 9.000 is a given.
In 64bit (test software has to be purchased) afair the scores were a couple hundred points higher, so some people managed to break the 10.000 barrier.
The 4-core 5150 (i.e. standard 1,1) scores on average 5.289, the 4-core 5160 @3GHz averages at 5.912 and the 8-core x5365 (3GHz) averages at 9.769.
I could not find official results for the x5355@2.66GHz (they may be there though - did not search thoroughly), but i can tell that those were just shy of 9.000 (32bit) on my own 1,1. I once managed to get around 9.300 somehow (no programs running), but i'd say a solid 9.000 is a given.
In 64bit (test software has to be purchased) afair the scores were a couple hundred points higher, so some people managed to break the 10.000 barrier.
I say you should be good with a 9 inchet. You can always cut it to your liking. With a tool that ends up shorter than needed, you will have a hard time making it longer. In most cases longer is better.