Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Thats the point actually. Web browsing, word processing, and iTunes are single threaded apps, they will not go faster with more cores. 11% boost compaired to 11% decrease is a big change. 8 cores running slower will make all single threaded apps run slower. I for one took the route of going with dual 3Ghz quad cores over dual 2.66 dual cores...wanted to cover both angles.

I agree but what gains does 11% actually get you? Almost nothing and further if normal day to day activities are all that you are doing with a Mac pro, why do you have a Mac pro?!? These can be accomplished with a mini and cost you less energy....
 
Questionable. Any increase below 50% (some set the threshold at 100%) is hard to notice at all without using a benchmark program.

Exactly, 11% is impossible to feel in day to day operations. Which again, I question if all you are doing is basic day to day tasks, why do you even have a Mac Pro when even a Base Mini scores roughly the same as a quad core 3.0ghz 2006 Mac Mini....

Edit: The question here is rhetorical by the way!
 
Here's my after effects before and after my 2x2.66 quad upgrade. Cut my renders in half and then some.
 

Attachments

  • Screen shot 2012-07-24 at 5.31.17 AM.png
    Screen shot 2012-07-24 at 5.31.17 AM.png
    38.5 KB · Views: 312
I have the original 5160s. I wouldn't even consider swapping for 5355s as that's a downgrade in my book, no matter what geekbench says. Real world performance would be worse as I don't use multi-threaded apps that much. The only viable upgrade for me would be the 5365s where my single threaded performance wouldn't be affected.
 
I've replaced 5150s with 5160s due to my software usage. No benefit for me from more cores - I use PS, ID and AI mostly. And I'm still on 10.5. I would upgrade to SL but my scanning software won't work with it and there's no alternative except VueScan. Add price difference 5160 vs 5365 or even 5355 (2 years ago) to this and decision was straightforward.
"Slightly better" is always much better than nothing ;)
I've choosen tower because of expandability and because in 2006 it was the best choice for me (and got really good deal on it). Now in 2012 it still suites my needs, what would be much harder without all upgrades I've done through these years.

So, it really depends on software that one is using (or would be using in the future if it's possible to predict).
 
And again, for anyone reading this thread at a later date, why would you upgrade from 4 cores at @ 2.66ghz to 4 cores @ 3.0ghz. What benefit is there?
A significant increase in power and responsiveness for a very low price.

Almost none in day to day operations (11% clock speed increase is almost pointless).
Wrong.

If you really want to unleash the potential of your Mac Pros, then 8 cores at 2.33ghz or better is much better
Wrong. For the reasons given in post #21.

In day to day operations, a Xeon processors based on Core2duos at 2.66ghz or 3.0ghz isn't going to be much better.
For the price, that is wrong.

Really what is the point of spending $60+ to go to 4 @ 3.0ghz?
Really, whats the point of asking such a stupid question? If you have to ask about why one would install more powerful processors then you clearly know little about computers.

Going with your belief, then why is anyone buying 8+ core Mac Pros?
Because they are more powerful and its what they need. If you but a 4x2.66 and need more cores to do the work you want, you should have bought an 8-core to start with!

I for one took the route of going with dual 3Ghz quad cores over dual 2.66 dual cores...wanted to cover both angles.
Thats the smart route, but we are discussing cheap upgrades. Our 4x3.0 and 8x1.86 processors cost less than $70 total, just one of the CPUs you bought is more than $150.

I agree but what gains does 11% actually get you?
Der, a more responsive system. :confused:

Questionable. Any increase below 50% (some set the threshold at 100%) is hard to notice at all without using a benchmark program.
Then why does Apple charge $2400 (as much as an entire base MP) for "only" a 25% increase in performance with the 2.4-3.06 12-core model? By your logic, such a small increase in speed will not be noticeable.... :rolleyes:
 
Questionable. Any increase below 50% (some set the threshold at 100%) is hard to notice at all without using a benchmark program.

Which is why I upgraded my 1.1 rather than buy the bump they released as new earlier this year. The Mac Pro release frequency is driving me away from Macs. When a new mac pro is less than 50% more powerfull than one that is 6 years old....I am done.
 
Which is why I upgraded my 1.1 rather than buy the bump they released as new earlier this year. The Mac Pro release frequency is driving me away from Macs. When a new mac pro is less than 50% more powerfull than one that is 6 years old....I am done.

Really? Checking Geekbench, the 2006 came in at around 6000 where the newest 12-core comes in at almost 25000....
 
Then why does Apple charge $2400 (as much as an entire base MP) for "only" a 25% increase in performance with the 2.4-3.06 12-core model? By your logic, such a small increase in speed will not be noticeable.... :rolleyes:
The increase in cost is not linear correlated to the increase in performance. However, most buyers of this option make money off their machine and then every minute saved on CPU-intensive projects is worth pure money and the additional costs are soon made up for!

You don't notice it unless you measure it - be it by a dedicated benchmark program or a good old stopwatch making the highly CPU-intensive program a benchmark program on its own!
 
Really? Checking Geekbench, the 2006 came in at around 6000 where the newest 12-core comes in at almost 25000....

Dude, GB means squat to most users. Sure, it's nice to brag about huge numbers, but name me a few apps that most people would use that use all the 12 cores. Yeah, I thought so. My MP actually does close to 7000 GB, that's just under 2000 per core. The new 12-core does just over 2000 per core.

And to answer your next question, the point of having a pro machine like the MP without using pro apps, is that people are picking 1,1 machines for dirt cheap these days and looking to get the maximum potential out of it. I was waiting for a new Mini last year and came across this so cheap I had to get it. I love the expandability and performance.
 
Really? Checking Geekbench, the 2006 came in at around 6000 where the newest 12-core comes in at almost 25000....

Ahhhh...I see what you did there...take the bottom of the line 2006 and compair it to the top of the line 2012 and come up with more than 50%...clever...

I see some 1,1's comming in above 11000 and some 12 core's comming in around 17000.
 
Last edited:
Ahhhh...I see what you did there...take the bottom of the line 2006 and compair it to the top of the line 2012 and come up with more than 50%...clever...

I see some 1,1's comming in above 11000 and some 12 core's comming in around 17000.

Yeah the 1,1's with 11000 are the ones that were upgraded to octo 3.0ghz which per everyone else in this thread isn't "worth it".... Also remember, that when the 1,1's came out top of the line was only quad core @ 3.0 ghz (which scores less than 7K) with the standard model of quad core 2.66ghz.

Oh and 11000 (which wouldn't even have been available in 2006) to 17000 (which would be the slowest 12 core) is still more than 50% faster..... 11000 * 1.5 = 16500.... So yeah an upgraded (maxed) 2006 is still 50% slower than the slowest 12-core.

Listen I get what you are saying about geekbench, would you rather me provide rendering numbers? Would that appease you? I mean in this thread alone we have seen rendering times drop by half by going from the 2.66ghz quad core to the octo 2.66ghz. That's at a geekbench of 11,000. Shoot up to one of the new 12 core and I would imagine the rendering time of 1 hour that dropped to 29 minutes would be at 12 minutes maybe even 10 (I could probably go find some numbers if you really care).
 
Stock 1,1's are listed in database at a low of 3700 to a high of 6000. The rest are altered from stock. Real world is even more dramatic as there is no HT or turbo bins on the Core 2 architecture. I am amused everyone is so happy with their older Mac's but they are no match for even the 2009 stock. All of it does not mean they are no longer useful just slower than new stuff by a fairly wide margin. In the Mac past users were more than happy to pony up money for new Mac's that were on average only a few hundred MHz faster (see G3/G4/G5) and maybe not even 1000 geekbench points between. We are getting rather spoiled with these massive yearly gains. But all this takes perspective.
 
Stock 1,1's are listed in database at a low of 3700 to a high of 6000. The rest are altered from stock. Real world is even more dramatic as there is no HT or turbo bins on the Core 2 architecture. I am amused everyone is so happy with their older Mac's but they are no match for even the 2009 stock. All of it does not mean they are no longer useful just slower than new stuff by a fairly wide margin. In the Mac past users were more than happy to pony up money for new Mac's that were on average only a few hundred MHz faster (see G3/G4/G5) and maybe not even 1000 geekbench points between. We are getting rather spoiled with these massive yearly gains. But all this takes perspective.

While I agree with you that this explosive growth really didn't happen until recently let look at a few geek bench scores.

'03 MDD GB ......727
'05 G5 quad ......3594

thats pretty good growth especially given and particularly noticeable especially with the modern web.

'05 G5 quad ......3594
'10 MacBook.......3646
'10 MBP with i5...5529

You will honestly probably not notice a difference between these three (I haven't) in normal use. the core i will handbrake better but because of the slower drives it takes some away even compared to SATA 1.

'05 G5 quad ......3594
'06 MP 2.0.........4465
'06 MP 2.66.......5873
'08 MP 3.0.........8858
'12 MP 3.06 12...24,190

Having owned a 1,1 2.66, a 3,1 3.0, and a G5 quad (currently) I cannot tell the difference between the 3 in day to day use until I render or encode. My hackintosh with SATA 3 and a GB of 18,000 wasn't enough faster day to day not to be replaced with a quad.

IMHO anything above 3k will not be noticeable in day to day usage. I will not deny the speed nor do I believe it's replaceable I do deny that it's really noticeable. Please keep in mind that the Hack was at 5.0Ghz so it should be noticeably faster in day to day that anything Apple makes currently until 6 to 12 real cores are needed.
 
I was waiting for a new Mini last year and came across this so cheap I had to get it. I love the expandability and performance.

I also like the fact the CPUs run at 90*f all the time instead of 200*f like my 2011 i5 mini did.

Yeah the 1,1's with 11000 are the ones that were upgraded to octo 3.0ghz which per everyone else in this thread isn't "worth it".
Correct. If you're going to spend $300 to upgrade to an octo, it might as well be to add the $300 to the base price and buy a 2009 2.8 octo MP that can run the most modern OS version and fully 64-bit software.

Listen I get what you are saying about geekbench, would you rather me provide rendering numbers?
xbench provides a more real-world benchmark since it accounts for disk drive, GPU and UI performance as part of the equation.
Simply installing an SSD for example will produce a big jump in the score, just as an SSD improves performance in the real-world.

IMHO anything above 3k will not be noticeable in day to day usage.
The GB issue with the G5 isn't so much power, its compatibility. My MacBook browses the web easier than my Quad G5 did simply because the G5's software is years behind in updates.
 
Correct. If you're going to spend $300 to upgrade to an octo, it might as well be to add the $300 to the base price and buy a 2009 2.8 octo MP that can run the most modern OS version and fully 64-bit software.
.

But it doesn't cost $300 to upgrade to an octo anymore. I spent $100 to upgrade to an Octo @ 2.66ghz in my 1,1..... And the price to upgrade to 4 cores @ 3.0ghz apparently is $68. To me, the rendering/encoding time drop that 8 cores at 2.66ghz makes the additional $32 worth it. Especially since 11% gains in single threaded apps isn't going to feel any faster, but my encoding times will be cut in half.

And now you are talking about adding an SSD and how that affects the numbers, but technically that has nothing to do with processing power. So now I have to ask, what are we comparing? Day to day tasks, adding an SSD will have a greater impact than upgrading to 3.0ghz or octo would it not? But that wasn't at question here, the discussion here is whether upgrading from 4 cores @ 2.66ghz to 4 cores @ 3.0ghz would even matter.
 
I also like the fact the CPUs run at 90*f all the time instead of 200*f like my 2011 i5 mini did.


Correct. If you're going to spend $300 to upgrade to an octo, it might as well be to add the $300 to the base price and buy a 2009 2.8 octo MP that can run the most modern OS version and fully 64-bit software.


xbench provides a more real-world benchmark since it accounts for disk drive, GPU and UI performance as part of the equation.
Simply installing an SSD for example will produce a big jump in the score, just as an SSD improves performance in the real-world.


The GB issue with the G5 isn't so much power, its compatibility. My MacBook browses the web easier than my Quad G5 did simply because the G5's software is years behind in updates.

That would be a bad browser choice.
 
But it doesn't cost $300 to upgrade to an octo anymore. I spent $100 to upgrade to an Octo @ 2.66ghz in my 1,1
Thats because you got the slow ones. But as has been said many times now, you won't get any actual increase in performance except with a few specific highly multithreaded apps. At least with the octo-3.0's you get benefits of both.

..... And the price to upgrade to 4 cores @ 3.0ghz apparently is $68.
Which will produce a tangible increase in real-world app performance.

And now you are talking about adding an SSD and how that affects the numbers, but technically that has nothing to do with processing power.
If you would read the thread, you'd see that we are talking about real-world performance of a computer, not benchmarks. Since the OS has tens of thousands of sub-1MB files an SSD will significantly improve responsiveness of the entire computer.

----------

That would be a bad browser choice.

That is a stupid statement.
 
Stock 1,1's are listed in database at a low of 3700 to a high of 6000. The rest are altered from stock. Real world is even more dramatic as there is no HT or turbo bins on the Core 2 architecture. I am amused everyone is so happy with their older Mac's but they are no match for even the 2009 stock. All of it does not mean they are no longer useful just slower than new stuff by a fairly wide margin. In the Mac past users were more than happy to pony up money for new Mac's that were on average only a few hundred MHz faster (see G3/G4/G5) and maybe not even 1000 geekbench points between. We are getting rather spoiled with these massive yearly gains. But all this takes perspective.

I've always felt that it comes down to if the things you want to run will perform without lag. When we look at scores today, a G3 seems unusable, yet it covered many of the same functions. In some cases gains in software requirements haven't been too aggressive over the past few years. It obviously varies. When you start stepping into software that goes into 4 or 5 figures, it's assumed that up to date hardware can be purchased to run it. Below that a lot of stuff has been running in place to a degree allowing it run reasonably well on lighter systems. If you owned something like a 1,1 ans your workflow hasn't changed from what it was + machine isn't lagging, I could see how there wouldn't be much motivation to replace it. People were most likely willing to pay for some of the older upgrades because software demands escalated faster.
 
I use my 1,1 MP for rendering, photoshop, and video encoding. The $100 I spent for the x5355's was worth every penny. Couple that with $100 for an SSD makes a big difference. My jump was from the 4600's to 9600's on geekbench. My rendering speeds have tripled, and I haven't even added more RAM yet. The CPU temps have never gone over 50C and I didn't upgrade the firmware to 2,1. I did have to bump the fan speeds to 1000, but the fan noise change is slight.

The 3.0GHz 8 core CPU's get too hot so I wouldn't recommend those and you have to upgrade the firmware.

The only reason to upgrade the CPU's on a 1,1 is for multi-threaded applications. I would buy or build a new PC if you want to use the MP for gaming. The PCIe 1.0 slots limit the GPU you can use and long term will not be cost effective. A $700 PC is much more cost effective than a $2500 MP.

Photoshop use is a toss up. RAM and SSD would help you more than doing a CPU upgrade. I have noticed much better performance with the SSD (because of the scratch file on the SSD) than with the CPU swap.
 
  • Like
Reactions: splifingate
I know this is an old thread, but I'm looking to upgrade my 1,1 2x Dual Xeon 2.66 to an Octo machine, just not sure which and what to go with.


How much faster would the 3.0Ghz X5365's be over the much cheaper X5355's in single threaded and day to day use?



Which X5355's (if I go with these) do I go with? SLY9, SLAC4, or SLAEG? I read SLAEG is more power efficient, so I bookmarked a pair of those from eBay.


EDIT: Found this: https://forums.macrumors.com/threads/1210403/

SLAEG it is, but what about the other, faster 5365's? Worth the price difference?
 
Last edited:
I know this is an old thread, but I'm looking to upgrade my 1,1 2x Dual Xeon 2.66 to an Octo machine, just not sure which and what to go with.


How much faster would the 3.0Ghz X5365's be over the much cheaper X5355's in single threaded and day to day use?



Which X5355's (if I go with these) do I go with? SLY9, SLAC4, or SLAEG? I read SLAEG is more power efficient, so I bookmarked a pair of those from eBay.


EDIT: Found this: https://forums.macrumors.com/threads/1210403/

SLAEG it is, but what about the other, faster 5365's? Worth the price difference?

I did the upgrade about 6 months ago and opted for the 5355's instead of the 5365's. Frankly 2.66ghz to 3.0ghz is only about 12% but when I was looking the cost for 2-2.66ghz was about $100 were as 2-5355's was well over $200. I just couldn't justify 100%+ more in cost for only 12% more speed. In single threaded applications, it won't be any faster than your old 4 core machine, but I find that I can more virtual machines, do more stuff like handbrake, etc now with the an Octo than I could with just quad cores. Some will tell you, you are better off with 5160's since single threaded apps will run faster, but it seemed silly to me to only go from a 2.66ghz quad core to a 3.0ghz quad core. Yes single threaded apps would be faster, but that's about it.

In short, my opinion is go with the 5355's and just be happy that you went from roughly a geekbench score of 5200 to over 9000 for about $100....

Edit: Don't forget, you are spending money on a 6 year old machine. It theoretically could "give up the ghost" at any time, so for all you know a week after yo do the upgrade it could die. Keep that in mind when deciding how much money to put back into it.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.