speeds act the same when using 512mb or 3gb r

June 17, 2009 at 14:56:13
Specs: Windows XP
i tested both rams, the 512mb by itself and the 3gb by itself. Knowing that 3GB RAM should be faster (not only bigger) they both run the same, but this happened AFTER i combined 256MB ram pieces along with the 1GB to make a little more than 3GB RAM on my computer.

Really the only difference between 512 and 1GB ram is the size. 1 bigger than the other. Not the speed or things like that. Aren't 1GB ram sticks suppose run FASTER than 256MB RAM?? I'm really in need of help.

Is this weird? because I isn't the clock speed suppose to adjust automatically no matter how many times u switch ram. Did i mess something up?


See More: speeds act the same when using 512mb or 3gb r

Report •


#1
June 17, 2009 at 15:41:11
Quantity of RAM has exactly nothing to do with CPU clock speed. The CPU speed may adjust based on the type of RAM installed, but not the amount.

How are you testing speed? Are you only looking at the reported CPU speed, or are you running some kind of benchmark? If you expect your CPU clock speed to increase or decrease based on how much RAM is installed, then your expectation is wrong and there may not actually be a problem here.

http://support.microsoft.com/default.aspx?scid=kb;en-us;555375
How to ask a question.


Report •

#2
June 17, 2009 at 15:56:00
I tested the speeds one time doing multi tasking.

When i first got 3GB RAM, i removed the 512MB ram completely and installed the 3GB (1GB each stick), then i saw a major difference.

Since I have 4 slots, i added 1 256MB stick into the remaining stick after that and i saw there was a downfall in the speed when multi tasking.

After the coincidence, I removed the 256MB stick out of the slot to keep the 3GB RAM by them selves, and there was no change.

If the computer runs at the slowest speed of the smallest RAM stick you have in combination with your BIGGER sticks, then removing the smallest stick(s) should make your computer run faster because it'll then be running at the speed of the bigger stick. And I don't think that happened.


Report •

#3
June 17, 2009 at 16:16:20
"If the computer runs at the slowest speed of the smallest RAM stick you have in combination with your BIGGER sticks, then removing the smallest stick(s) should make your computer run faster because it'll then be running at the speed of the bigger stick. And I don't think that happened."


That is sort of true, but I think you have some misconceptions here. Assuming they are the same type of RAM (e.g. PC2700, PC2100), all sticks will run at the same speed regardless of their size. A 512MB stick of PC2100 runs at the same speed as a 1GB stick of PC2100, by definition. RAM speed is determined by the type, not the size.

You originally went from 512MB to 3GB, so of course you're going to notice a big apparent difference in speed from that. This is due to the increase in amount of installed ram, and has nothing to do with the size of any individual stick. But you also say you notice a speed increase when removing the 256MB stick. Compare the 1GB sticks to the 256MB stick. Are they the same type? You can find some information about RAM types here to help you identify what you have. I suspect the 256MB stick is a slower type than the larger sticks, in which case you're probably right...the system is downclocking to the speed of the slowest installed stick. Which, again, has nothing to do with the SIZE of the stick and everything to do with the TYPE.


Now, if all installed sticks are the same type the system may "feel" faster or slower when more or less RAM is installed. This is not due to how fast the RAM runs, but rather how the system actually works when different amounts of RAM are installed. With 512MB installed, the system does a lot more paging than when more RAM is installed. So even though both amounts of RAM function at the same speed, the overall system "feels" slower. Again, this is due to the amount of RAM, not the type.


Your method of testing speed, is inherently a measure of how the system "feels". When you multitask by running several programs, the system must swap a lot when there is less RAM, and will "feel" slower as a result. This is what you observed, right?

http://support.microsoft.com/default.aspx?scid=kb;en-us;555375
How to ask a question.


Report •

Related Solutions

#4
June 17, 2009 at 16:27:59
since the system downclocks to the slower speed when adding 256MB to the rest, even when I removed it, it still felt the same as if I still had it in. I don't think that was right.

Shouldn't the system upclock if the 256MB is removed? And run or feel the same it did in the beginning when I initially installed just simply 3GB?


Report •

#5
June 17, 2009 at 16:33:48
You're talking in general & not giving any specifics about your hardware, but there's a difference between speed & performance.

A 3.0GHz CPU runs at 3.0GHz regardless of whether there's 256MB RAM installed or 4GB RAM. But the amount of time it would take to run a timed program such as SuperPi would dramtically decrease with more RAM. In other words, if it takes 30 secs to run with 256MB, it may only take 10 secs with 3GB.

However, for optimal performance, you also need to properly pair up the RAM with the CPU. For example, to be able to run dual channel mode, you need either 2 matched sticks of RAM or 4 matched sticks. By running 3 x 1GB, you're forcing the board to run in single channel mode. You also need the RAM to run at the proper clock speed, but without knowing the FSB speed of your CPU, it's impossible to say what the RAM clock speed should be set at.


Report •

#6
June 17, 2009 at 16:35:08
"Shouldn't the system upclock if the 256MB is removed?"

Yes, if only the 256MB is of a slower type, and all remaining sticks are of an identical, faster type.

But your method for testing isn't accurate. Between tests, do you wait the same amount of time after booting up to run your tests? Is the environment the same? What tasks are you performing? Has the page file size been adjusted, or is it fixed? There are too many other variables here. If you really want to know what speed the CPU and RAM operate at, get a benchmark utility or check the BIOS.

http://support.microsoft.com/default.aspx?scid=kb;en-us;555375
How to ask a question.


Report •

#7
June 17, 2009 at 17:48:36
okay. I think after the research I've done I got a few thinks correct now. I have 3GB ram running in single channel and there would be no way to make that dual channel which would run faster than it is. I would have to remove 1 of those 1GB sticks out to make it dual.

But I could have sworn one day it ran at dual memory speed (you can feel the difference vs single) when i had 3 sticks in use. And that was when the first time i ever inserted 3GB (before i even inserted a 256MB). when i inserted that 256, i noticed a change in the "feel" like it became single channel when it was dual before. I guess im wrong because it was always single to begin with.


Report •

#8
June 17, 2009 at 17:59:38
Dude, you're talking in circles. How about listing your system specs?

- What's the make/model of your motherboard?
- What's the make/model & FSB speed of your CPU?
- You've mentioned at least 4 sticks of RAM. 256MB & 3-1GB sticks...are they all the same? For example, are they all DDR2-667? If not, what are they?

As you said, you've probably never enabled dual channel mode with the RAM you have. Even so, the difference between dual vs single is only noticeable under certain circumstances. It's not like you would switch the RAM around, reboot & say, "wow, what a difference!"...it doesn't work like that.


Report •


Ask Question