hardCOREware - Hardware Gaming News for HARDCORE Gamers  

 

hardCOREware Newsletter!



 

 

MainReviews  -  Forums  -  Downloads  -  Links  -  About Us

Find the Lowest Price for Computer Hardware!

Ram Comparison:
is More Better?

Written By: Stephen 'Waterdog' Waits [03.19.01]



Introduction

Everyone knows that SDRAM prices are at an all-time low. With Rambus and DDR making their way into the market place, SDRAM is the slowest memory technology, the oldest, and accordingly the cheapest. Prices this inexpensive beg the question: do I have enough RAM? Should I have more? Where, if anywhere, will I notice more RAM? We all know what RAM is, what it does (at least in a general sense), so little introduction is necessary. Let's get down to it.

Test System

AMD Thunderbird 1.2GHz
ABit KT7-RAID
2x128MB PC133 SDRAM (generic)
1x256MB PC133 SDRAM (Micron)
Maxtor 20GB ATA100 7200rpm
Asus V7100 Geforce2 MX (Detonator 6.50)
SoundBlaster Live! Value
Windows 2000 Professional SP1
Windows ME

All tests were run several times to ensure no anomalous values made it into the mix. The scores were averaged, and the machine was rebooted between each set of tests. The following testing software was used.

Business Winstone 2001
Content Creation Winstone 2001
3DMark2001
Unreal Tournament (thunder.dem, available at www.3dpulpit.com)
Quake 3 Arena (demo001)
Team Arena (mpdemo1)

The tests used were chosen to represent a wide variety of tasks, from work to play. Both ME and 2k were used to give you an idea of how much of a difference you'll notice if you have the 9x or NT core on your system, as both have different ways of dealing with memory, and different initial memory footprints. Enough dawdling, onto the tests.

Disclaimer

As you've noticed, we're running some Crucial banners at the moment. This comparison was prompted by the incredibly low RAM prices we've seen in recent weeks, and ZDNet's speculation that we've hit the bottom. The fact that we're running banners from Crucial is coincidental, and you'll note that the actual testing wasn't even done with memory bought from them (though they are essentially Micron's factory-direct retail division, we got our Micron ram elsewhere). Just to be on the safe side, I'm going to pretty much let the numbers speak for themselves, bias is hard to avoid with words, but numbers are objective as they get.

Explaining the Graphs

What's really important for these benchmarks is the gain in performance achieved by adding more RAM, not the individual scores. To that end it's a waste of bandwidth to show you graphs of individual scores, so instead I'm going to show you scores in terms of the performance gained from each step up in memory. When looking at a graph you'll see gains for 256MB, 384MB, and 512MB. These are relative performance gains, based on the increase in performance over the *previous* memory size, not the base memory of 128MB.

Think of it this way, each % gain is the increase in performance you'd see from adding another 128MB to the previous score. In the case of 256MB, the % gain is the gain from 128 to 256. For 384MB, the performance gain is the gain from 256 to 384. For 512MB, the performance gain is the gain from 384 to 512. When looking at the graphs be sure to make note of the scale and actual values as things have been graphed in relation to the other values on the particular graph.

On To the Benchmarks!

Page [1] [2] [3] [4]

Upgrade your RAM with the Memory SelectorTM
Select your system and press go!  

Copyright 2000, 2001, 2002 hardCOREware. All rights reserved.
All trademarks used are properties of their respective owners.
Use IE5 or 6 to see this page properly. 1024x768 resolution is a minimum.
32 bit colour is recommended.
Our Privacy Statement

Proudly Canadian