Adding RAM, any reason for more than 4 Gb?

[4string]

In an effort to keep my G5 going I would like to add some RAM. Since Pro Tools is a 32 bit application limited to 4 Gb RAM, is there any reason to add more?

My system:
MAC PPC G5 2.0 GHz Dual Core, 2.5 Gb RAM
Glyph GT 050 160 GB Audio Firewire drive
WD Caviar SE16 500 Gb Video internal drive
Logic Pro 9 (Logic Studio)
Pro Tools 8.0.1
Final Cut Express HD 3.5

I use Pro Tools for mostly audio editing, some times 30 plus tracks. I use Drumagog in the early editing stage. I try to stick to "less is more" regarding plugins: EQ, compression, reverb, and amp sims for guitar and bass DI. I am not a big user of virtual instruments. If I do use a VI, I usually record the output and then remove the plugin.

I am thinking about adding 2 or 4 Gb of RAM - is there any point to adding more that 2 to bring me up to 4.5 Gb?

Thank you
Colin

[4string]

I am on OS X 10.5.8

[albee1952]

FWIW, I have had 2 different Pro Tools instructors say that 4gig is the ideal amount for PT.

[getz76]

Since OS X can address more, it isn't a terrible idea. Pro Tools should address 4gB, while the other OS X applications can use free RAM (Finder, etc.).

[CME]

Yeah IMO more than 4GB is the point of diminishing returns. You should notice a difference on any session up to about 4GB. I like at least 4.5GB so the OS has its own 512MB, but that's just my opinion. No facts to back it up. My G5 has 6.5GB but it's because it started with 512MB, then I added 2GB, then found a good deal on 4GB. I've never taken the 2GB pair out to see how it runs, but I seen no reason to go any higher with my system. Because like you I don't use a lot of VI's. IMO VI's are what benefit the most from extra ram (over the 4GB PTLE can use itself).

[Dism]

Originally Posted by CME

Yeah IMO more than 4GB is the point of diminishing returns. You should notice a difference on any session up to about 4GB. I like at least 4.5GB so the OS has its own 512MB, but that's just my opinion. No facts to back it up. My G5 has 6.5GB but it's because it started with 512MB, then I added 2GB, then found a good deal on 4GB. I've never taken the 2GB pair out to see how it runs, but I seen no reason to go any higher with my system. Because like you I don't use a lot of VI's. IMO VI's are what benefit the most from extra ram (over the 4GB PTLE can use itself).

I'm actually kind of surprised you haven't had any problems with your system, for a few reasons. Mostly because it's a bad idea to mix RAM. Kind of like how they tell you not to mix battery brands. RAM will only run as fast as the slowest chip in the system. Meaning, that 512 chip could be slowing down your whole system. You would probably see an improvement in performance by taking out the 512.

Generally, when you put RAM in a system you want every chip to be the same speed and brand for optimal performance and stability.

[1ace1]

Originally Posted by Dism

RAM will only run as fast as the slowest chip in the system. Meaning, that 512 chip could be slowing down your whole system. You would probably see an improvement in performance by taking out the 512.

A may be a simple question, but what is the difference in speed between different quantities of RAM?

I was under the impression that 512MB could be accessed at the same speed as 2GB just not hold as much data. Is this wrong?

A:)

[Dism]

Originally Posted by 1ace1

A may be a simple question, but what is the difference in speed between different quantities of RAM?

I was under the impression that 512MB could be accessed at the same speed as 2GB just not hold as much data. Is this wrong?

A:)

You're right about the differences in storage quantities, and in theory, a 512 chip could access data at the same speeds. However, not all RAM is designed the same. Even if they supposedly have the same bus speed (which they all should to cooperate) there are still smaller factors that affect their performance, and different brands have a lot to do with it. Such as, a 2GB chip for $40 and a 2GB chip for $80 are probably going to perform differently even if they are both PC4800. So if your $40 value brand chip is slower than your $80 performance brand, your fancy $80 chip is only going to run as fast as the $40 chip. RAM also tends to be finicky in that sense. Every time I tried to mix RAM at all in my PC tower, it wouldn't even boot unless every chip was EXACTLY the same.

So what I'm saying, is if you bought 6GB of RAM all the same brand and swapped it out for your mix and match 6.5GB, you'd probably notice a significant difference in performance.

[getz76]

Originally Posted by Dism

I'm actually kind of surprised you haven't had any problems with your system, for a few reasons. Mostly because it's a bad idea to mix RAM. Kind of like how they tell you not to mix battery brands. RAM will only run as fast as the slowest chip in the system. Meaning, that 512 chip could be slowing down your whole system. You would probably see an improvement in performance by taking out the 512.

Generally, when you put RAM in a system you want every chip to be the same speed and brand for optimal performance and stability.

Dism,

Size of RAM has nothing to do with its bus speed. They are completely separate specifications.

Some systems can take advantage of matched pairs of RAM, but not all. In any case, as long as the RAM is compatible with the system, it should be fine.

[getz76]

Originally Posted by Dism

You're right about the differences in storage quantities, and in theory, a 512 chip could access data at the same speeds. However, not all RAM is designed the same. Even if they supposedly have the same bus speed (which they all should to cooperate) there are still smaller factors that affect their performance, and different brands have a lot to do with it. Such as, a 2GB chip for $40 and a 2GB chip for $80 are probably going to perform differently even if they are both PC4800. So if your $40 value brand chip is slower than your $80 performance brand, your fancy $80 chip is only going to run as fast as the $40 chip. RAM also tends to be finicky in that sense. Every time I tried to mix RAM at all in my PC tower, it wouldn't even boot unless every chip was EXACTLY the same.

So what I'm saying, is if you bought 6GB of RAM all the same brand and swapped it out for your mix and match 6.5GB, you'd probably notice a significant difference in performance.

That's a lot of misinformation based on anecdotal evidence. You likely had problems with your RAM in your PC because your BIOS were set up incorrectly.