Install is the same as any ram install as all the power for the LED’s are drawn from the ram, for those of you that are changing ram for the first time just shut down and disconnect the power lead, then remove old set and replace with the Ballistix Tracers. On most motherboards, the dual slots tend to share the same color as in pics below

Wow, I like the lighting effect, even the LED’s at the bottom have a nice effect and make the motherboard glow. This is the first time I have seen lighting in this way on ram I to be honest I am quite fond of it. As I mentioned earlier, all this and still the ram is standard size, Nice!

OK, on to testing.

Test setup:

First thing to check the timings of the modules, 333MHz or 667MHz DDR at 4, 4, 4, 10 is impressive to say at least.

Next, I spent a few hours messing with these sticks and endless timings to see the maximum clock I could reach. After experimenting for a while, I ended up stumped. Why, well it’s quite weird really but no matter what I done I couldn’t get the ram to clock and higher than 367Mhz and 735Mhz DDR,What made this strange is now matter what I done with the timings this was the maximum clock I could get even with 2.1 volts supplying 1.9 volt ram. Still a DDR rate of 800 MHz is nothing to be sniffed at and yes I have had Mushkin here that hit the 1000 MHz mark but with looser timings, so without hesitation we will compare the 2. the Mushkin how ever is only rated at 533mhz and timings of 4,4,4 12, but as I know it hits 1000mhz on the bus and can easily match the 677 speed with timings of 4,4,4,12, I will set it at 667spec for stock tests and then up it for clocked tests. The Ballistix Tracer DDR2 has a bit to prove here as it only hits the 800 MHz mark

The tests performed at all of the settings above are:

Sisoft Sandra 2005 pro Memory Bandwidth

PCMark 2005 (Memory Benchmarks)

Lavalys Everest Memory Benchmarks

Now there will be 2 results for each one stock and one clocked, the clocks are as stated and the tests were done on a fresh installed and updated XP pro. Below are the results.

Sandra 2005 pro Memory Bandwidth Sisoft Sandra is a popular computer hardware-benchmarking tool, that is efficient at testing ram

.

AS you can see at stock the Ballistix performs better than the Mushkin, this is to be expected though due to the Ballistix tracers have slightly tighter timings.

Well even though the Crucial Ballistix was at a lower MHz than the Mushkin When clocked and it certainly showed that the slightly tighter timings rather than the higher clock is the winner here ( but we all know that most of the time tighter timings give better results)

PCmark 2005 PCMark is another popular computer hardware-benchmarking tool. For the ram we will be benching the ram only. Bellow are the results

As you can see it clearly still outperformed the other set of ram we tested it against, even though the mushkin was clocked higher

Lavalys Everest Everest is a fairly new benchmarking and hardware information tool from Lavalys.  Their home-edition software is available for download at www.lavalys.com, which includes the benchmarks used here.

Memory Read (higher is better)

Memory Write (higher is better) 

Memory latency (lower is better)

Again, after the Sisoft results no real surprises in these tests with the slightly tighter timings at lower speeds proving to be the winner.

Next Page: Conclusion