PDA

View Full Version : Power Mac G5 Quad VS new Mac Mini 2.4GHz Intel Core 2 Duo




marisp
Jul 13, 2010, 03:58 AM
Hi, what do you think guys is the new Mac Mini 2.4Ghz as powerful as the old G5 Quad or not really?

Thanks



iWright
Jul 13, 2010, 05:24 AM
In terms of choice it would probably come down to the following:
1) Expandability (More in the G5)
2) Platform (Do you want/need to run software that needs an Intel chip [like snow leopard])

OrangeSVTguy
Jul 13, 2010, 07:45 AM
For everyday tasks, you will not notice a difference between the two besides the obvious physical features. Also the Mini is dead silent compared to the G5 but then again compared to any computer for that matter. It's no slower than my 2.26mhz 2009 Mini. It's still quite a good machine for being 5 years old.

noodle654
Jul 13, 2010, 08:34 AM
The 2009 2.66GHz Mini gets a geekbench score of 3724. The G5 Quad gets 3301. Doesn't mean much though, the G5 has much more expandability. If you are trying to buy one or the other, I would get the Mini, unless you can find a ridiculously cheap G5 Quad for less than $500.

zmttoxics
Jul 13, 2010, 08:34 AM
I dont know if expandability is worth arguing anymore. The only thing you would really want are bigger and more harddrives, and the new mini is fw800 capable. It is also easier to upgrade the ram in the mini now (maybe even easier then doing it in the G5).

My point is, why dump money into expanding the capability of the G5 when you are nearly at the end of software life for it. If you get a new mini there is nothing you can't run (besides classic, but that argument is waning too).

I have had both a 2005 G5 and a 2009 mini, my vote is for the mini.

Hrududu
Jul 13, 2010, 04:21 PM
If you use applications that will take advantage of all 4 cores, then the G5 will smoke the Mini. This past week I've been using a dual 2.0GHz G5 while my MBP was getting a logicboard replacement. I can honestly tell you I didn't notice hardly any difference in speed doing everyday tasks. Granted, my MBP is only 2.16GHz, but we're talking about a faster mini and a faster G5. For video encoding etc, I would think the G5 is going to be a lot faster.

Davy.Shalom
Jul 19, 2010, 12:29 AM
The Mini, just because it's a PITA to find a cheap reliable G5 quad with good specs. Also, don't fork out too much extra money on a machine that's approaching the end of it's life. If all you do is email, word, iTunes, and youtube, and editing is CS4, then the G5 is probably for you. If you want to go beyond that and also add Windows dual boot capability to that picture, go with the mini.

Buck for buck, I think the mini is a better deal just because it will be running modern software with the potential to upgrade with more modern software.

AND-if you ever want to run a game just dual boot windows, even most mac games now aren't PPC compatible. Just a thought.

mrsir2009
Jul 19, 2010, 12:37 AM
Yeah, apart from being louder, bigger and not being able to run things like Snow Leopard and any other programs that require intel. But if you don't need those... and your not bothered by the noise... go ahead! :D

DesmoPilot
Jul 19, 2010, 01:02 AM
If you use applications that will take advantage of all 4 cores, then the G5 will smoke the Mini. For video encoding etc, I would think the G5 is going to be a lot faster.

Doubtful in my opinion. That Core 2 Duo is generations ahead of the G5; combined with newer versions of software (also generations newer) and intel optimizations, I'd put money on the C2D being quite a bit faster than the Quad G5 for demanding creative tasks such as video encoding.

mrsir2009
Jul 19, 2010, 01:23 AM
Doubtful in my opinion. That Core 2 Duo is generations ahead of the G5; combined with newer versions of software (also generations newer) and intel optimizations, I'd put money on the C2D being quite a bit faster than the Quad G5 for demanding creative tasks such as video encoding.

Even so, preforming demanding tasks on a G5 quad core would be quite a enjoyable experience compared to the consumor computers of that time (G5 iMac).

marisp
Aug 10, 2010, 03:01 AM
hi, thanks for great input. the thing is that i own the G5 quad and im using for home recording processes (logic 9). I'm thinking weather to switch to mini or no.

I would agree that its better to switch to something on which you can run all the latest stuff, but my only concern is that if you open up a heavy session on mini it can still run it!

thunng8
Aug 10, 2010, 06:30 AM
hi, thanks for great input. the thing is that i own the G5 quad and im using for home recording processes (logic 9). I'm thinking weather to switch to mini or no.

I would agree that its better to switch to something on which you can run all the latest stuff, but my only concern is that if you open up a heavy session on mini it can still run it!

I just did a benchmarking session on the Powermac Quad G5 vs a 2.4Ghz 13" Macbook Pro Core 2 Duo (exactly same CPU your proposed mini)

Cinebench 11.5 (Higher is better). Macbook used the 64 bit version
Quad G5 - 1.99
Macbook Pro - 1.39

Canon DPP Raw 3.8 - Convert 10 5D2 raw files to JPG (2 tasks)
Quad G5 - 2m 10 sec
Core 2 Duo - 2m 39 sec

HandBrake 0.94 encoding from 22min divx to ipod touch/ipod preset. Macbook used the 64bit version
Quad G5 - 5m 43sec
Core 2 Duo - 5m 13sec

Blender 2.53Beta rendering. Macbook used the 64 bit version
Quad G5 - 26.3s
Core 2 Duo - 27.1s

Photoshop using Radial blur filter on high quality. Macbook used CS5 64 bit while G5 used CS4
Quad G5 - 1m 7s
Core 2 Duo - 1m 20s

Quicktime 7.66 Export for a 720p movie trailer to ipod preset
Quad G5 - 1m 10sec
Core 2 Duo - 1m 24sec

So overall, the Quad is faster at all these tasks except for Handbrake which I suspect is heavily optimized for x86. But the speed difference isn't huge. The mini would be a slight step back in performance, but it would used a lot less power (watts).

MACENZ
Aug 10, 2010, 07:01 PM
Geekbench means nothing (its never accurate) As a Graphic Designer, I know what it means to put a machine to work!! I have owned a Mini and you will be disappointed! If you work quick, you will need a machine that will always be right with you. The Quad is still a beast and will be for many more years to come. I would go with the Quad any day!

If you want an Intel, then I suggest you buy a Macbook Core Duo to get your feet wet!! LOL

Good Luck.

thunng8
Aug 10, 2010, 09:38 PM
The 2009 2.66GHz Mini gets a geekbench score of 3724. The G5 Quad gets 3301. Doesn't mean much though, the G5 has much more expandability. If you are trying to buy one or the other, I would get the Mini, unless you can find a ridiculously cheap G5 Quad for less than $500.

Just ran geekbench 2.16 (32-bit as I don't have a license) and was surprised again that the G5 Quad 2.5Ghz pipped the Core 2 duo 2.4Ghz.

G5 quad - 3507
Core 2 Duo - 3428

From the benchmarking that I've done in the last few days, I can say that:

- G5 Quad most definitely faster than a Mac mini 2.4Ghz Core 2 duo
- G5 Quad would still be faster than the high end Mac mini 2.66Ghz
- likely that the 2010 Macbook Pro i5/i7 series should be faster than the Quad G5

Overall, I think the Quad G5 is holding up well in terms of performance for a 5 year old computer.

Only reasons you would want to switch to a Mac Mini 2.4Ghz are:
- Software compatibility (eg. Snow Leopard, CS5 suite, Lightroom 3). Going forward, more and more software will be intel only.
- You care about energy consumption. Mac mini should use approx. 10x less watts than a Quad G5.

marisp
Aug 25, 2010, 05:44 AM
thanx for these great posts! It definitely makes things clearer! so seams like the latest mini with 8gb ram would still lose to G5 quad.

Thanks again for posts

zmttoxics
Aug 25, 2010, 05:46 AM
thanx for these great posts! It definitely makes things clearer! so seams like the latest mini with 8gb ram would still lose to G5 quad.

Thanks again for posts

Only in numbers. You win by it not being a PPC IMHO. On the heals of 10.7 - living in 10.5 sounds crazy.

Giuly
Aug 25, 2010, 06:48 AM
I dont know if expandability is worth arguing anymore. The only thing you would really want are bigger and more harddrives, and the new mini is fw800 capable. It is also easier to upgrade the ram in the mini now (maybe even easier then doing it in the G5).

My point is, why dump money into expanding the capability of the G5 when you are nearly at the end of software life for it. If you get a new mini there is nothing you can't run (besides classic, but that argument is waning too).

I have had both a 2005 G5 and a 2009 mini, my vote is for the mini.
If you want to run Classic, get a PowerMac G4 MDD and run MacOS 9.2 native.
Else get a Mini, the 2.66GHz unibody Mini has a geekbench score of ~4060. this is 1/4 more than the G5.

thunng8
Aug 25, 2010, 06:56 AM
If you want to run Classic, get a PowerMac G4 MDD and run MacOS 9.2 native.
Else get a Mini, the 2.66GHz unibody Mini has a geekbench score of ~4060. this is 1/4 more than the G5.

I don't really care what Geekbench says. On the applications I use, the Quad will still be a bit faster than the 2.66Ghz mini.

Giuly
Aug 25, 2010, 07:21 AM
I don't really care what Geekbench says. On the applications I use, the Quad will still be a bit faster than the 2.66Ghz mini.
Well, your benchmarks are very close to the 2.4GHz MacBook, the 2.66GHz is still a bit faster. I can bench it, if you like.

But, what happened to the "2.5GHz G5 is faster than 2.5GHz x86"? Was that compared to Pentium 4? And why the hell is a 2.5GHz Quad-Core G5 about as fast as a 2.4GHz Dual-Core C2D? Has it the IPC really gone that much higher on the Intel CPUs?

If you have the time, it would be interesting to see how single-threaded applications run on both machines, like unrar, so we can compare the power per core. It should take 1.5-2x as long to uncompress the same file according to those benches, but I guess it looks different in reality.

thunng8
Aug 25, 2010, 07:52 AM
Well, your benchmarks are very close to the 2.4GHz MacBook, the 2.66GHz is still a bit faster. I can bench it, if you like.

The advantage was -10 to 40%, with average around 20%. So by my estimation a Core 2 duo of ~2.9Ghz should be able to match the PowerMac Quad - which is still higher than the fastest Mac mini.


But, what happened to the "2.5GHz G5 is faster than 2.5GHz x86"? Was that compared to Pentium 4? And why the hell is a 2.5GHz Quad-Core G5 about as fast as a 2.4GHz Dual-Core C2D? Has it the IPC really gone that much higher on the Intel CPUs?

Note that not all the benchmarks listed there use the full capability of the 4 cores. For example, for Canon DPP, if I run simultaneous batches on the conversion jobs to max out the 4 cores (versus simulatenous to max out 2 cores of the core 2 duo) I get these results for 30 RAW 5D2 conversions:

PowerMac Quad: 4m 57s
Core 2 Duo 2.4Ghz: 7m 8sec

an advantage of around 44%

Yes, the Core 2 Duo has much higher IPC than the PowerPC 970MP. While 970MP has much higher IPC than Pentium 4. I estimate a single core ~1.6-1.8Ghz Core 2 should be equivalent to a 2.5Ghz 970MP single core and Nehalem based cores are even faster.

But you have to keep in mind the PowerPC 970MP design dates back to 2003 (2001 if you count the POWER4 which the 970 is based on). Core 2 is a 2006 design and and Nehalem was released in early 2009. Looking at the latest PowerPC design the POWER7 (released early 2010), you will see that IBM's latest has much higher IPC than any Intel chip.


If you have the time, it would be interesting to see how single-threaded applications run on both machines, like unrar, so we can compare the power per core. It should take 1.5-2x as long to uncompress the same file according to those benches, but I guess it looks different in reality.

unrar isn't a good benchmark - too dependent on hard drive performance. Rar compression would be better. Not sure how multithreaded it is though. I can test tomorrow.

QuantumLo0p
Aug 25, 2010, 12:39 PM
The Mini, in real world usage, is a little faster overall but definitely not by a huge margin. The Mini's disk is a horribly slow, stinky, turd of a drive and my old PowerMac, with raid 0, totally smokes the Mini. A SSD disk would definitely be faster. I plan to put a Momentus XT in mine; for me its a good compromise between speed and price. As for the Mini's graphics, which should overall be better than my old 6800 in my PowerMac, occasionally suffers from sever lag playing games which lasts for minutes unless I allow it to catch up. A couple times i had to re-start WoW to remedy the problem. I also plan to to upgrade the ram to matched 2GB or 4GB sticks.

It totally depends what your priorities are. For me, I opted to get a Mini when my PowerMac gave up the ghost. A big thing for me is having an Intel cpu so I can get all the new software which is becoming more and more limited on the Power platform.

thunng8
Aug 28, 2010, 05:18 PM
unrar isn't a good benchmark - too dependent on hard drive performance. Rar compression would be better. Not sure how multithreaded it is though. I can test tomorrow.

I didn't do the Rar compression since I didn't have it installed, so I used stuffit Deluxe v14 to compress the iwork '09 folder to .sitx archive. To my great surprise the PowerMac G5 was faster. 16m 20sec vs 21m 40s. And this benchmark never uses more than 100% CPU usage.

Stuffit compression to .sitx overall is very slow, but I wonder why it is so slow on Core 2 Duo machines? Anyway, another convincing win for the G5 on this one.

dellar
Dec 25, 2010, 04:24 AM
I had the G5 Quad a few years ago but sold it when I was moving house (it was like moving a house in itself). I own a unibody 2.66Ghz Mac mini that I use primarily as a HTPC (imagine that for a Quad G5!), but quite regularly I plug it into a Cinema Display and use it as a desktop mac for music composition. I get a Geekbench 64bit score of just under 4100 (which is higher than I thought considering my friend's 21.5" iMac brand new from last year (3.06Ghz) has about the same score. I wouldn't want to be stuck with PPC solely because I'd be stuck at 10.5 forever.

macdudeguy
Dec 25, 2010, 05:11 AM
Can you people please stop bringing up Geekbench? It has nothing to do with real world performance, it's moody enough that the test score of the same machine can vary by hundreds of points from very little things, and its composite score is never a reflection of the true median performance offered by all hardware combined.

The fact that the only benchmarks conducted for this thread show the exact opposite of Geekbench results; decisive wins for a Quad G5 against the Core 2 Duo in question, should be all the evidence you will ever need to remove Geekbench from your computers and never speak of it again.

occams razor
Dec 25, 2010, 06:19 AM
Geekbench is for little pussies that wanna brag about **** that doesn't matter. Get a Mac mini. They are gonna stop supporting the ppc machines then what are you gonna do? The mini will run everything you need and or want. Better yet build a hackintosh and call it good. This thread needs to be closed cause it's a pissing match cause everyone has their own opinions. Do your own research and come to your own decision instead of wasting our time

California
Dec 25, 2010, 12:23 PM
This is an old thread.


That said, the OP already owned a Quad.

As for me, I've owned a handful of Quads. Even with the Nvidia Quadro graphics card and 16gigs of ram.

Nice machine, those Quads. If calibrated right, they are pretty silent.