PRODUCT COMPOSITION & PICTORIAL (cont'd)

Here you can see the large stickers which cover most of one side. The first identifies the RAM by part number, whereas the second gives you the memory's specifications

These are 184-pin DDR modules.

One of the first things you will notice is the lack of heatspreaders. RAM gets fairly hot, especially if you are overclocking your system. Because of this, many manufacturers now ship their memory with (preferably) copper heatspreaders. Seeing as the Crucial memory isn't aimed at the enthusiasts market, shipping their modules without heatspreaders is an easy way to cut that cost and offer you good memory at a lower price. If you decide that you'd like to pick up your own heatspreaders, they are not that expensive and considering the already low cost of the Crucial memory, you will still ahead.

The only question which remains is whether or not the Crucial memory can hold its own.

BENCHMARKING SETUP AND COMPOSITION

I tested all memory modules in my Albatron KX18D Pro II system. You'll notice the fan at the top of the system which was used to add a little extra cooling to 'all' of the memory modules which were used. It is my son's Memorex 80mm fan. I doubt it was needed, as the modules never were that hot to touch… but why take chances.

The system specs are as follows.

Lubic open case
Athlon XP2500+ (with 333FSB) Check Prices
HIS Radeon 9600SE
20GB Maxtor UATA133 HDD
Windows XP SP1 (and all critical patches and updates as of January 5th/2004) Check Prices
Catalyst 3.10 video drivers

Memory

· Crucial CT6464Z40B PC3200 DDR400MHz (2 x 512MB modules) Check Prices
· OCZ Platinum Dual Channel Series PC3200 Low Latency Memory (2 X 256MB modules) Check Prices
· ULTRA Dual Channel PC3200 DDR400MHz (2 X 512MB modules)

You will notice right away that the OCZ is only 512MB total, whereas the other two 'contenders' are 1GB in total. This matters very little in terms of memory benchmarking, as the applications take into consideration the speed with which the modules communicate with the motherboard… not the total bandwidth.

BENCHMARKING PROGRAMS USED

· 3DMark2001SE
· PCMark2002 (Memory)
· SiSoft Sandra Memory Bandwidth Benchmark
· GLExcess v1.2V Benchmark
· Quake 3 Demo Benchmark

PROCESSOR CONFIGURATION USED

Single Channel (represented as Single in the Benchmarks)

· XP2500+ @ 1.8GHz (DDR400) (SPD) (3-8-3-3 RAM Timings) CRUCIAL
· XP2500+ @ 1.8GHz (DDR400) (SPD) (3-8-3-3 RAM Timings) OCZ
· XP2500+ @ 1.8GHz (DDR400) (SPD) (2.5-8-3-3 RAM Timings) ULTRA

Dual Channel (represented as DDR400 in the Benchmarks)

· XP2500+ @ 1.8GHz (DDR400) (SPD) (3-8-3-3 RAM Timings) CRUCIAL
· XP2500+ @ 1.8GHz (DDR400) (SPD) (3-8-3-3 RAM Timings) OCZ
· XP2500+ @1.8GHz (DDR400) (SPD) (2.5-8-3-3 RAM Timings) ULTRA

Dual Channel Overclocked (represented as DDR4XX in the Benchmarks)

· XP2500+ @ 2.0GHz (DDR436) (SPD) (3-8-3-3 RAM Timings) CRUCIAL
· XP2500+ @ 2.45GHz (DDR446) (SPD) (3-8-4-4 RAM Timings) OCZ
· XP2500+ @ 2.55GHz (DR446.5) (SPD) (3-8-4-4 RAM Timings) ULTRA

The default timings were used for both the Single and Dual Channel modes. All modules were tested with SPD (Serial Presence Detection). I toyed with the Aggressive and Expert Timings to attain the best overclock.