In our first test equipped with Athlon 4000+ a single core CPU clocked at 2.4 GHz and 1 MB cache memory, two times 512 MB OCZ PC 4800 Elite platinum edition, 250 GB Western digital 8 MB cache drive and Gainward 7800 GTX card, Asus A8N SLI motherboard and Creative Labs X-Fi Fatal1ty sound card we scored 4.2 score.
Anything higher than four is good as it means you can run Aero Glass really well.
The Windows Experience Index tool won't actually give you an average number, it will pick the lowest number from the five tests it runs and set it as your final score. In my case even I had 1 GB of OCZ PC 4800 Elite platinum edition superior memory I only scored 4.2 in memory test and that was my final score. The graphic scored an excellent score of 5.9, 3D graphic score was 5.8, hard drive scored 5.3 and the CPU scored 4.2.
Vista with 4x512 MB memory
Well I decided to play a little and to plug some more memory as I logically assumed I need more memory. I really wanted a higher score.
The first step was to plug two more 512 MB modules. I picked Kingston KHX4300K2 1GB memory, an additional two modules. I wanted to enlarge my ram portion to a final 2 GB. I believed that this will make Vista happy. Oh boy I was wrong.
I know that I had to screw the memory timings to make 4x512 MB memory work together and I did so. Imagine my surprise when I re-used Windows Experience Index and scored 4.1 only. The memory performance dropped to 4.1 as the settings were little slower. Trust me forget about the memory timings 2 GB are going to perform better than 1GB of memory at least in normal windows tasks. If you run out of physical memory and with vista it is very easy to spend the whole 1GB of memory on just an OS and a single application or game, you system will start to swap and it will slow down. If you have 2 GB or more things will likely be faster.
Well Vista Windows Experience Index disagrees with me. It gets even better. I tried to use the whole 4 GB of memory jet again OCZ 4x 1 GB of DDR 466 memory at 2.5, 3, 3, 5 settings and expected to get a superb score.
We upgraded a machine little bit, plugged dual core 4800+ CPU with 2x1 MB cache clocked at 2.4 GHz and scored a satisfactory score or 4.9. Even I plugged graphically faster card Radeon X1950 XTX, which is clearly faster than 7800 GTX 256 MB card in every single game vista rated the card again as 5.9 in graphic, 5.8 in 3D graphic score. We didn't see that coming.
Vista with 4x1024 MB memory and bunch of new hardware
I upgraded hard drive from Western Digital Caviar SE 250GB SATA with 8 MB cache to Seagate Barracuda 7200.9 500GB 16MB SATA II and obviously expected some performance increase. Well I was wrong again as the index number stayed the same, 5.3.
We upgraded system with around 450 worth memory, a brand new graphic card Radeon X1950 XTX that costs roughly 399, a Seagate Barracuda 7200.9 500GB 16MB SATA II drive that costs just over 200 and a CPU that costs around 270. What did we gain from Microsoft performance tool just a lousy non realistic few points. Trust me Vista works much better with these new components and 1300+ new hardware will really make a stand out difference. Unfortunately not according to this Microsoft performance tool.
I urge you to stay away from this tool as it will just confuse you. It is as controversial as 3Dmark06 that will score better on Shader model 3.0 card than on much faster Shader model 2.0 and you will have to spend hours to explain to a less experience user why is so.
The memory and dual core upgrade sure counts a lot and even the SATA II drive with 16 MB makes a difference. In graphic of course it is faster but you wont be astonished with the difference, some twenty percent if you lucky and that's about it, but still more.
And finally Vista can address 4096 or 4095 / 4094 MB of system memory. Congratulations to the Vole, we will send you the flowers. The memory number defers from the tool you are using but we don't mind if we lose one or two MB but we mind about the gigabytes that's for sure. Take a look what Windows XP thinks about 4GB of memory. ?
Uses 20 percent less power than traditional systems
It's becoming more prevalent in car research and development
Sign up for INQbot – a weekly roundup of the best from the INQ