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

Tyroler

macrumors member
Original poster
Jan 2, 2011
97
0
Hatfield, UK
http://www.ebay.co.uk/itm/220874472...WAX:IT&_trksid=p3984.m1423.l2649#ht_846wt_936

Was asking the Seller if Cooler is leaking and he says "No,its new!" (nearly) Apple exchanged in 2009 Logicboard and Cooler.
But im not happy to bid on it,as i dont know how long it will be till he starts leaking again.
But here maybe someone who knows how you treat such Cooled Machine and how to maintain it.
Think im more for a 2.3Ghz watching around
 
I honestly wouldn't trust any G5 system when it comes to anything that resembles longevity. If you can't control your need to buy one then buy the most reliable model which is the fan cooled 2.3GHz.

A G4 MDD (dual 1.25+) would be reasonably close in performance and be much more reliable than even the 2.3 G5. It would also use at least 50% less power if not 80% at times.
 
I know plenty of people that are happily still using their PowerMac G5's. I understand they are not the most reliable systems out there but they are not built like **** either.
 
I know plenty of people that are happily still using their PowerMac G5's. I understand they are not the most reliable systems out there but they are not built like **** either.

I have direct extensive experience with approx. 107 different PowerMac G5's over the past 7-8 years and over half of them are no longer running. All of these were used in lab situations and every single one was running on state of the art power conditioning from the time it came out of the box till death. That means they were fed 100% clean power 100% of the time so any failure was their own and not the power.

These 107 were mostly dual 2.0GHz but about 30 were newer 2.3GHz and quad systems. Not that all of them die but they are BY FAR the most unreliable Mac EVER made. Period.
 
I have direct extensive experience with approx. 107 different PowerMac G5's over the past 7-8 years and over half of them are no longer running. All of these were used in lab situations and every single one was running on state of the art power conditioning from the time it came out of the box till death. That means they were fed 100% clean power 100% of the time so any failure was their own and not the power.

These 107 were mostly dual 2.0GHz but about 30 were newer 2.3GHz and quad systems. Not that all of them die but they are BY FAR the most unreliable Mac EVER made. Period.

I can second this with similar experience with about 60 PowerMac G5s. My primary Mac at home is a last-gen 2.3GHz (DC) and I only partially trust it. I've got spare parts, but yeah. I hang on to it because I don't have the budget to upgrade at the moment. Anecdotally, it's part of the class of machines that I didn't see many failures of beyond video cards with those, but they're still G5s. Great machine for UT2004, does okay with Flash, but I'm glad I got it through a trade and didn't pay cash for it.

I would strongly advise against paying cash for any G5 unless it's for a case by itself.
 
I wouldn't be buying any liquid cooled G5, they everyone else has said they just aren't reliable. Having said that, if you can find a non-liquid cooled G5 that is still running without problems, chances are it's one of the good models. I've got a dual 1.8 G5 that works fine.
 
Zen, mate most of the time your spot on, but your last is just utter crud..

"A G4 MDD (dual 1.25+) would be reasonably close in performance and be much more reliable than even the 2.3 G5. It would also use at least 50% less power if not 80% at times."

That last bit is true fer sure, but your initial here is just plain crap mate :)

The MDD dual 1.2 scores around 1085 Geekbench and the Late 2005 G5 2.3 scores around 2084.. so the G5 is round 2x as quick as the MDD 1.2..

ive owned and run both and yeah the MDD is a great box, but slow in comparison with the late 2005 G5..

http://www.primatelabs.ca/geekbench/mac-benchmarks/

and i am no bench slut either, but even encoding the G5 PM 2.3 has it all over the MDD.. You are however spot on about the power usage.. the G% SUCKS big time in that regard...
 
I honestly wouldn't trust any G5 system when it comes to anything that resembles longevity. If you can't control your need to buy one then buy the most reliable model which is the fan cooled 2.3GHz.

A G4 MDD (dual 1.25+) would be reasonably close in performance and be much more reliable than even the 2.3 G5. It would also use at least 50% less power if not 80% at times.

The G4 and G5 systems aren't really comparable in performance. You can run quite a lot on a G5 that would choke the G4, and the G5 can accept super cheap ram upgrades.

Keep in mind here while they can run okay that's still a very dated machine. A macbook or mini today can run circles around it assuming it's configured properly. Tiger could still benefit from more than 3GB of ram. If you went with Leopard and around 6 you're probably maxed on performance at that point. Ram for those is quite cheap.
 
The G4 and G5 systems aren't really comparable in performance. You can run quite a lot on a G5 that would choke the G4, and the G5 can accept super cheap ram upgrades.

Keep in mind here while they can run okay that's still a very dated machine. A macbook or mini today can run circles around it assuming it's configured properly. Tiger could still benefit from more than 3GB of ram. If you went with Leopard and around 6 you're probably maxed on performance at that point. Ram for those is quite cheap.

It is a common knowledge that the latest intel offerings have much higher clockspeeds and more powerful processors.

Besides it is a PPC thread.

So let us talk PPC not intel.
 
VanneDC:

It's not utter crap at all. You can't base everything on benchmark numbers. I'm talking about real world computing. When it comes to common tasks there is not that drastic of a difference at all.

I have experience with those 100+ G5's plus I am the former owner of 2x MDD and one was the dual 1.42 so I actually know this from experience. My 1.42 held it's own against G5's as does my single 1.8GHz G4 7448. In xbench for example my 1.8GHz G4 scores a 97 vs. the base of 100 for a dual 2.GHz G5. The G4 7448 is actually a newer CPU than all the G5 chips. Freescale released the 7448 in 06 after Apple had left PowerPC hardware behind.

My only thought for how different our thinking is here is that you must have used a lot of 64-bit software on your G5 which would certainly make a big difference in performance.
 
Last edited:
There is no way Zen, let us dig deeper, how can a 167MHz bus speed for the latest G4 be faster than a 1.25GHz on a 2.5GHz PMG5?

FSB is still a major factor in this comparison.
 
Zen, this rime you're wrong about 1.25 G4 vs. 2.3 DC. They'e not even close each other, especially in real world. In any CPU, RAM or HDD intensive task G5 is better. Period (as you use to say ;)). I mean comparing the comparable configs, i.e. 2GB RAM each, the same HDD (in G4 via SATA card ofc) for instance.

BTW, your Xbench score (97 in CPU section) is based purely on CPU frequency, not on "power". Xbench sucks there (and not only there :D)

@Macintosh"Master" - stop trolling man. If you don't like zen and his comments report him to the mods or so, just don't spread your crap all over the forum. More respect for other readers, your'e not alone here...
 
Zen, this rime you're wrong about 1.25 G4 vs. 2.3 DC. They'e not even close each other, especially in real world. In any CPU, RAM or HDD intensive task G5 is better. Period (as you use to say ;)). I mean comparing the comparable configs, i.e. 2GB RAM each, the same HDD (in G4 via SATA card ofc) for instance.

BTW, your Xbench score (97 in CPU section) is based purely on CPU frequency, not on "power". Xbench sucks there (and not only there :D)

In the real world experiences I have at common tasks there wasn't that massive of a difference. Hence use of the words "reasonably close in performance". When it comes to CPU and RAM intensive stuff there would be a much more noticeable difference in performance. Even more so with 64-bit software.

I'm not delusional here.. :) I am speaking from the actual experiences I have had over many years.

As for xbench I do understand it's not the best bench out there but it isn't all based on clock speed. Many of the 1.8GHz 7447 upgrades only score in the 70's. I remember you once being impressed with my 7448 xbench score and now you try to say it doesn't matter. Make up your mind my friend. ;)
 
As for xbench I do understand it's not the best bench out there but it isn't all based on clock speed. Many of the 1.8GHz 7447 upgrades only score in the 70's. I remember you once being impressed with my 7448 xbench score and now you try to say it doesn't matter. Make up your mind my friend. ;)

You're geting me wrong. I mean that CPU section is based on clock. Score in this part increases in linear way with CPU clock increase. No matter what FSB is. That's why my 1.92 7447 eMac beats your 1.8 7448 chip (which is obviously better than eMac's CPU).
 
It is a common knowledge that the latest intel offerings have much higher clockspeeds and more powerful processors.

Besides it is a PPC thread.

So let us talk PPC not intel.

Sorry I didn't mean to derail it toward intel. The G4s probably had a lower failure rate. Neither is reasonably serviceable at this point from the perspective of buying a machine today. At the very least I'd ensure you price it out the way you want to use it. Any changes in hard drive or ram configuration and a small budget for potential replacement parts should be factored in when you're determining how much to pay for an older machine. The tasks you want to run on it do matter. You'll want to ensure you're buying something suitable for your needs. Keep in mind a G5 will impact your power bill more than other options if this machine is going to be running constantly.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.