RAM upgrade - Performance Average. Help?

[willywilson]

Hey guys, got a question for all you out there reading!

I recently installed an extra 2gb of RAM to my computer to make total of 3gb of RAM. My computer is nearly 2 yrs old. XP Home 64 bit and all that jazz. I expected a fair bit of a performance upgrdae but it really doesn't seem to have changed much.

My question is thus: Is there a way to make the performance of the RAM run better through software etc. and is there any other way I can utilise the extra RAM so it uses its full potential?

Thanks for any help!

[Piggy]

There should be a decent performance boost going from 1 gb -> 3 gb.

Take out the new ram and see how it is without it.

[gwai lo]

If you're not using the computer to the point where the extra RAM is used, then you won't see a performance difference. It's kind of like buying a 4870 over your older FX5200 and expecting windows to load faster.

[ralph_tran]

If you're not using the computer to the point where the extra RAM is used, then you won't see a performance difference. It's kind of like buying a 4870 over your older FX5200 and expecting windows to load faster.

yeah, but adding more ram would make windows load faster and do things faster no matter what apps hes running. Its not like the 1gb stick is used, then when it is filled the 2gb stick is used.

[gwai lo]

I wouldn't expect it to be significant though, *shrug*. A guy I lived across the hall from had 1GB of RAM v. my 2GB of RAM. We had comparable hard drives and processors and for working with Windows we hardly saw a difference until we started running more applications. While I was waiting on a motherboard, I was running 2?512MB v. my 2?1GB now and using Windows was essentially the same for me...however it is all relative. I saw gains as I opened up my usual number of applications, where the multitasking was a little quicker...but yeah, just saying that if the OP only opens up one IE Window to check his email, 1GB v. 3GB won't really change much. In this case, we can assume what we want about what the OP does.

edit: Okay to address the sticks thing, dual channel does provide gains, but they're not really that significant. You can try this one yourself, move one of your sticks to another DIMM and do what you normally do;);)

edit2: Oh and we both run XP 32.

[ViperAFK]

Meh over 1 gb doesn't make much of a difference on xp at all, exepct for gaming and heavy multitasking. Now vista you would definitely see an increase though. I upgraded from 1 to 2 gb in xp and didn't really notice much difference.

[redvamp128]

When you upgraded did you make sure to use the same memory that matches the first stick? Same part number- Depending on the Motherboard it will run the memory at the lowest memory setting. Try this- Take out the old stick and put in the 2 new sticks and see how it performs- I have seen this before when you mix and match memory.

[Zanaffer]

I agree, RAM is not the cure-all for the performance ailments. Also, regarding software to manage RAM to make things go faster, you have it backwards. The LACK of software makes things run faster. Look into disabling some unnecessary programs and services and that will speed the computer up. No need for extra fancy software to optimize things... chances are it would actually slow things down.

[gwai lo]

When you upgraded did you make sure to use the same memory that matches the first stick? Same part number- Depending on the Motherboard it will run the memory at the lowest memory setting. Try this- Take out the old stick and put in the 2 new sticks and see how it performs- I have seen this before when you mix and match memory.

This also pretty much only appears in synthetic benchmarks, FSB and RAM speeds matter little in real world performance.

[redvamp128]

Actually Some motherboards can run for example ECC and Non ECC memory and when you mix and match them you get poor performance (just as an example). The computer sees all the memory. Another example I have seen is a system with 600mhz memory in slot 0 and 800mhz memory in slot 1 and 2. The Bios saw all memory and ran all 3 sticks at 600mhz. But as soon as the person took out the stick in slot 0 the system responded faster.

So has nothing to do with benchmarks.

Also manufacturer specs for memory list them as 1 thing and they are fractions off... but fractions off can cause some issues with performance that is why they like to sell memory in sets-

An example of that is some "40gig " hard drivers are really 39gig or some are more.

The same can also be true for example with slot 0 having unbuffered memory and slot 1 and 2 having buffered memory can cause poor performance.

Edited July 13, 2008 by redvamp128

[gwai lo]

Okay ECC and non-ECC I can see maybe being weird...but what do you mean by 600 MHz and 800 MHz sticks...? Like DDR2-667 v. DDR2-800? I can't think of a setup that ran RAM at 600 MHz as of right now.

Also, what do you mean by responded faster? Was there a measured difference? Or something "felt", as the latter is quite prone to the placebo effect and human error.

Wrt to hard drives, that's just marketing working on base 10 and computers being base 2.

[redvamp128]

That was just an example-

Mhz was the wrong example but PC would be the correct specs

PC 600

http://www.pricewatch.com/system_memory/pc600_512mb.htm

PC 800

http://www.pricewatch.com/system_memory/pc800_512mb.htm

Same memory and if you mix them the system will run the faster memory at the slower speed.

[gwai lo]

Oh, well if you're talking about platforms that old then I don't have the knowledge/experience to speak on that. I thought you were talking about a more...recent system.

However, with regards to DDRI to present, RAM and FSB matter little.

[redvamp128]

However if he mixed the memory it could cause slow down issues or no performance gains.

That is why I suggested him to remove the old memory To see if the 2 gigs would perform faster-...

Sorry about that antiquated example but recently I ran into a server that had that issue and that solved the issue by changing it all over to 4 slots with the PC800 and that one was at my Corporate Home Office.

Someone before me had tried to upgrade the memory and mixed them- and it was very slow. But now it is faster.

Edited July 13, 2008 by redvamp128

[gwai lo]

Yeah I can see how that could be an issue with older systems, which brings up to this next point....we don't have the OP's specs either, so we can just speculate for another page or two.

Anyway, that's kind of interesting to see how different platforms react so differently to....well different things. Those older systems were actually affected by RAM speed that significantly...the A64's saw up to 10% in performance deltas with timings...now we have Conroe based systems that pretty much don't get affected by RAM significantly (although 1:1 is still considered ideal).

[redvamp128]

He also posted here gwai lo

https://www.neowin.net/forum/index.php?show...=646094&hl=

I think there we will have more of a clue as to the system specs.

[gwai lo]

Ah, that was a good idea. *smacks forehead*

Yeah with that platform there are no issues at all with slower sticks, the architecture has an onboard memory controller so RAM always runs a divider against the memory controller/chip.

So that's one thing to scratch off our list now.