Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

slicecom

macrumors 68020
Original poster
Aug 29, 2003
2,065
98
Toronto, Canada
I have a MacPro 1,1 2.0GHz and would like to upgrade the CPUs.

Is there a significant difference in every day performance between a 3.0GHz Woodcrest (quad core) vs. a 2.33GHz Clovertown (octo core)?

I use my computer mostly for graphic work in Photoshop, Indesign and Illustrator and also like to play the occasional game.

I ask because a Clovertown 2.33GHz goes for more than twice as much as a Woodcrest 3.0GHz, just want to see if its worth the price to go with Clovertown for my uses or not.
 
For most of Photoshop, all of Indesign & Illustrator and 95% of available games the 3.0 will run circles around the 2.33.

For the parts of PS that can use 8 cores the 2.33 quads will probably average about 150% of the 3.0 duals. Meaning half again as fast or as if you had 3 of the 3.0 CPUs instead of just 2. ;)

I think that's a pretty good guesstimate. Anyway, only about 30% of the the PS app, filters, and etc. are able to take good advantage of multiple cores like that. The other 70% will perform MUCH better on the 3.0GHz processors.
 
For most of Photoshop, all of Indesign & Illustrator and 95% of available games the 3.0 will run circles around the 2.33.

For the parts of PS that can use 8 cores the 2.33 quads will probably average about 150% of the 3.0 duals. Meaning half again as fast or as if you had 3 of the 3.0 CPUs instead of just 2. ;)

I think that's a pretty good guesstimate. Anyway, only about 30% of the the PS app, filters, and etc. are able to take good advantage of multiple cores like that. The other 70% will perform MUCH better on the 3.0GHz processors.

so that begs the question---- what about compared to a single processor nehalem? For example - a dual 2.66 dual core vs. a single quad core nehalem at 2.66.....
 
For most of Photoshop, all of Indesign & Illustrator and 95% of available games the 3.0 will run circles around the 2.33.

For the parts of PS that can use 8 cores the 2.33 quads will probably average about 150% of the 3.0 duals. Meaning half again as fast or as if you had 3 of the 3.0 CPUs instead of just 2. ;)

I think that's a pretty good guesstimate. Anyway, only about 30% of the the PS app, filters, and etc. are able to take good advantage of multiple cores like that. The other 70% will perform MUCH better on the 3.0GHz processors.

Great response, exactly what I was looking for. Thanks! :)
 
NP,

Another thing you might wanna look into is a used set of 3.0GHz clovertowns. That would be the best of both worlds. 8-cores total, high speed, and you can find them pretty cheap right now (used)! About $600 - $800 for the pair. You wanna look for X5365.
 
the nehalem would destroy the woodcrest.

Yeah... Two x5355 vrs one X5550 would be a better comparison. One would faster at some things than the other but the X5550 would be faster "under your mouse" so to speak.

For those who don't know the model numbers that's two clovertowns (8 cores) @ 2.66 GHz vs. one nehalem @ 2.66 GHz.

You can see it in one test anyway here: https://forums.macrumors.com/posts/7270035/ See the entry called "2.66 2006 Octad-Upgrade" and compare it with the "*** 2009 2.66 Quad" entry. :)
 
NP,

Another thing you might wanna look into is a used set of 3.0GHz clovertowns. That would be the best of both worlds. 8-cores total, high speed, and you can find them pretty cheap right now (used)! About $600 - $800 for the pair. You wanna look for X5365.

Yeah... but....
my dual 2.66 could sell for 1600 or so... add 600 for a pair, and I am in nehalem territory (I can get edu discount - and refurbs appear for 2149)...

so it probably is best to just sell and repurchase
 
The 3.2 2008 mac pro comes close as far as performance goes. When programs are written to take advantage of 8 cores, then proper comparison can be accomplished.

Huh? The 3.2 2008 DESTROYS the nehalem 2.66 octad let alone the 2.66 quad! - in every test, all the time, at everything - past, present, and future.

What???
 
You never know - some paid "reviewer" could have the 3.2 8-core and the 2.66 4-core using the same amount (3 GB) of RAM; maybe the 3.2 could get starved up for memory while the 2.66 could sail along in the 32-bit world perfectly with its triple-channel. The only time a lower-clocked quad may outrun an 8-core of a higher clock....

But that doesn't mean I would run a 3.2 with that minimal amount of RAM :rolleyes:
 
NP,

Another thing you might wanna look into is a used set of 3.0GHz clovertowns. That would be the best of both worlds. 8-cores total, high speed, and you can find them pretty cheap right now (used)! About $600 - $800 for the pair. You wanna look for X5365.

$600-$800? I am seeing that price for ONE not two.... ugh...
 
Yeah, you have to hunt daily. And when you find them you gotta act quick too. :) In Japan there's like 10 or 20 very good auction sites - a lot of which will list a "buy Price" if you don't wanna participate in the bidding. I usually just scan Yahoo auctions myself tho. I don't seem to see it for you yanks, but we in the tropical Orient get an auction service with our Yahoo:

Anyway, if you keep looking (including at old server blades) you'll find them for that price. :cool:


EDIT:
Here's one that went for $275 on E-Bay: http://cgi.ebay.com/Intel-Xeon-Quad...iewItemQQptZAU_Components?hash=item4398edf567
Or maybe the "Bid History" URL will last longer(?): http://offer.ebay.com/ws/eBayISAPI.dll?ViewBids&item=290328540519


.
 
Holy crap, thats cheaper for what I paid for 2 2.33's. But the power requirement is allot more than the the 2.33. Wasn't looking at the comparison of chips when I commented. Was thinking of the 3.2 vs the highest clocked nehalem. My bad. Send me to processor jail :eek: .
 
Pfff, either you are incredibly sensitive to time differences or you are exaggerating a little. There are no huge performance differences ("destroy", "run circles...") as you seem to believe. Photoshop performance is roughly proportional to the clock frequency in most of the cases when comparing Mac Pros with each others. Because of the Nehalem architecture being a bit more efficient, you can add 10-20% speed advantage for the same clock frequency. The difference between 4 and 8 cores does however greatly depend on the performed operation. the speed advantage of an 8 vs 4 core machine varies between 0% and 80%...

And recall the speed difference between 2.33 and 3.2 GHz seems more than it actually is, its like waiting 13.7 seconds instead of 10.... my opinion.
 
LOL! You don't really use your apps do you? Run 200 images through PS, edit 6 hours of HD, or render 2000 frames in a 3D renderer and then we can discuss it. If all you're doing is a 13sec. photo edit then get a mini or iMac... That's what they're for.
 
Yeah... Two x5355 vrs one X5550 would be a better comparison. One would faster at some things than the other but the X5550 would be faster "under your mouse" so to speak.

For those who don't know the model numbers that's two clovertowns (8 cores) @ 2.66 GHz vs. one nehalem @ 2.66 GHz.

You can see it in one test anyway here: https://forums.macrumors.com/posts/7270035/ See the entry called "2.66 2006 Octad-Upgrade" and compare it with the "*** 2009 2.66 Quad" entry. :)

your graph indicates that the 2.8 2008 octad is a great refurb deal and best of both worlds. Add an SSD boot drive, and this may be the sweet spot???
 
I'm wondering if a SLA4P and a SLAED would work together. Has anybody experience with this issue. It seems that one can occasionally buy singles at good prices but one would not allways know which stepping it is. I know that the SLA4P was earlier and more expensive and worked on a different vorltage band. That seems to indicate that the two versions would not work together in a MacPro1,1.
 
I'm not sure, but even if some say that you could use them together, I would never use CPUs with non-matching stepping.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.