• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Only 5.8 in Vista Experience Index?

dparm

Limp Gawd
Joined
Jun 26, 2007
Messages
181
The setup:

ABIT IX38 Quad GT
4GB OCZ Reaper DDR2-1066 (5-5-5-15 @ 2.1V, D9 Micron)
Q6600
8800GT SSC SLI
Seagate 250GB
HT Omega audio
NEC DVD+RW
Vista Business SP1


No overclocking or anything like that, and for some reason I'm still only getting 5.8 in the "memory" section of the Windows Experience Index. All the latest drivers & newest BIOS.

What the hell is even needed to get a 5.9? The machine is still plenty fast for me so it's not a big deal I'm not getting 5.9 all the way down. Is there anything simple I'm missing?
 
Windows Vista score means nothing in the real world performance. Simple ignore it and more on to better things and enjoy your pc.

Good Day
 
you can view the details on your score -- then you'll know why you got a 5.8

likely it is your hard-drive.
 
You misread the first post then.

I got 5.9 in every area EXCEPT memory, where I got only a 5.8
 
That's ok, my C2D E8500 @ 3.3Ghz is only rating 5.7 in Vista64 Ultimate.
Everything else is at 5.9 -- Go figure?

One thing I know however is that if I flash the BIOS on my Asus Rampage Formula to a newer rev. Vista will report a new CPU score at 5.9

I could really care less though and would rather have a more stable BIOS to run on :)
 
I have 4GB OCZ Reaper HPC 800mhz 4-4-4-15 and I get a 5.9 Vista score. Try lowering your timings and see if it changes.
 
is it not running 1066? Is it using a lower divider?


jamiee
As for E8500 I think you need 4.0Ghz or more to hit 5.9 on the CPU score. However a stock quad will rank 5.9. Its a stupid rating system. Dual core takes a huge hit just because it's not a quad.

I, of course, hit 5.9s on everything and didn't quit till I did ;)
 
I have 4GB (4x1) 800MHZ ddr2 55516 and it's 5.9 My core 2 3ghz is at 5.7. OC to 3.6 and it's 5.8.

Ignore people who say "it doesn't matter" because software is moving towards using rating systems like this instead of the typical ratings. If you install newer games and look at your games folder, it shows you what the requirement and recommended is based on windows experience index. Get used to it because it's not going away.
 
I have 4GB (4x1) 800MHZ ddr2 55516 and it's 5.9 My core 2 3ghz is at 5.7. OC to 3.6 and it's 5.8.

Ignore people who say "it doesn't matter" because software is moving towards using rating systems like this instead of the typical ratings. If you install newer games and look at your games folder, it shows you what the requirement and recommended is based on windows experience index. Get used to it because it's not going away.


It pretty much is meaningless. When I had 8GB, I scored 5.9 across the board (which by the way is the highest you can go right now). I pulled out 4GB to RMA it, and it dropped to 5.8. Still dual channel, still running DDR-800 5-5-5-15.

And the only games I've seen using the Windows score are made by MS.


edit: my bad, Spore Creature creator uses it and requires a 2.0 System. My laptop can handle that.
 
Rainbow 6 and The Witcher are 2 games I have installed right now that use the index.

Anyways, for as bad off as the PC gaming industry is right now, you guys should be happy that someone is trying to make it easier on people to get into PC gaming. Don't push it away because it's not what your used to.
 
Rainbow 6 and The Witcher are 2 games I have installed right now that use the index.

Anyways, for as bad off as the PC gaming industry is right now, you guys should be happy that someone is trying to make it easier on people to get into PC gaming. Don't push it away because it's not what your used to.

If it were actually useful / meaningful, we wouldn't.
 
I have 4GB (4x1) 800MHZ ddr2 55516 and it's 5.9 My core 2 3ghz is at 5.7. OC to 3.6 and it's 5.8.

Ignore people who say "it doesn't matter" because software is moving towards using rating systems like this instead of the typical ratings. If you install newer games and look at your games folder, it shows you what the requirement and recommended is based on windows experience index. Get used to it because it's not going away.
It doesn't matter, particularly since the difference here is between 5.9 and 5.8. I'm sure some game is going to care that his memory has a score of 5.8 instead of 5.9. The scores don't mean anything.
 
I have 4x 1GB sticks. Moving from 2x1 to 4x1 made no difference.

I set 5-5-5-15 @ 2.1V since that's what the memory's EPP profile is. And yes, it is indeed running at 1066. Actually I think it's running at 1088, since that's exactly a 2:1 ratio.

Again, I could care less about the number, I'm just curious how Windows is arriving at a 5.8.


Perhaps the Intel X38 chipset's memory handling isn't good? I don't really know.
 
I had the same problem. My RAM got 5.9, and then I re-installed Windows and for some reason it went down to 5.6. It has something to do with when Vista assesses your hardware during Windows installation. You just have to clear out all the files in the \Windows\Performance\WinSAT\DataStore\ folder and run it again.
 
Weird. I'll have to try that when I get home tonight, thanks for the tip.
 
Mine went from 5.7 to 5.9 when I matched the memory speed with my FSB.
Overclocked my Q6600 to 3.0Ghz (8x375)
UNDERCLOCKED my memory (2x375 = DDR2 750)

Noticably faster and boosted my score :D
 
Mine went from 5.7 to 5.9 when I matched the memory speed with my FSB.
Overclocked my Q6600 to 3.0Ghz (8x375)
UNDERCLOCKED my memory (2x375 = DDR2 750)

Noticably faster and boosted my score :D

Something's not quite adding up here...if there's any improvement at all from clocking your RAM lower to match your FSB, (and I doubt there is), it definitely wouldn't be "noticeable," as you say.
 
Again, I could care less about the number, I'm just curious how Windows is arriving at a 5.8.
No one I have talked to knows how they derive their scores. Yet another reason it is a meaningless number thrown out there.

As noted, ignore it.
 
From the Windows Vista Team Blog. Doesn't help much, of course. :-p



The CPU Score

The CPU score was created to measure the processor performance when tasked with common Windows usage activities. The processor is assessed on the following items:

1. Compression and decompression using the LZW compression algorithm

2. Compression and decompression using the Windows Vista compression algorithm used for hibernation files, ReadyBoost and other features

3. Encryption and decryption assessment

4. Computing hashes

5. Encoding of video

The results are normalized and weight averaged in order to arrive at the final CPU sub score.
The Memory Score

The memory score measures the bandwidth of moving data into and out of memory in Mega Bytes per Second. The higher the bandwidth, the better the memory.

Not having enough memory is a limiting factor on performance. As a result, the amount of memory in the system constrains the score value. The amount of system memory is determined by the overall system memory minus any memory reserved for graphics (if any).

The amount of memory limits are:

Amount of memory


Highest possible score

Less than 256 MB


1.0

Less than 500 MB


2.0

512 MB or less


2.9

Less than 704 MB


3.5

Less than 960 MB


3.9

Less than 1.5 GB


4.5
The Graphics Score

This score is intended to reflect how a system will run Aero (desktop composition) and playback Windows Media Video. It measures video memory bandwidth (in mega bytes per second) and converts it to a score between 1.0 and 5.9.

If the graphics hardware does not support DirectX 9 (DX9) graphics, then the system receives a graphics score of 1.0 regardless of driver type. If the system supports DX9, but does not have a WDDM Driver (Windows Vista Display Driver Model) the system will then receive a graphics score of 1.9 at the most.
The Gaming Score

The gaming score measures the frames per second the graphics card can handle for various textures.

Additional notes:

· If the graphics card does not support D3D 9 then it receives a Gaming score of 1.0.

· If D3D 9 is supported, the card is DX9 capable and has a WDDM driver, it will score at least 2.0.

· If the score is greater than or equal to 5.0 and the graphics sub-system does not support Pixel Shader 3.0 then the score is limited to 4.9
The Disk Score

The disk score measures disk bandwidth (in Mega Bytes per Second). The conversion to an index number is set up in a way that all modern disks will score at least 2.0.
 
I have 4x 1GB sticks. Moving from 2x1 to 4x1 made no difference.

I set 5-5-5-15 @ 2.1V since that's what the memory's EPP profile is. And yes, it is indeed running at 1066. Actually I think it's running at 1088, since that's exactly a 2:1 ratio.

Again, I could care less about the number, I'm just curious how Windows is arriving at a 5.8.


Perhaps the Intel X38 chipset's memory handling isn't good? I don't really know.

Try pumping the FSB while keeping the ram at stock speed (if you want). You can lower the multi if you dont want to overclock the chip. If you want to prove it to yourself download sandra or everest and run the memory benchmark, then pump the fsb up to 300+ and try it again. I get pretty good bandwidth at only 844mhz on the memory with 422FSB.

*note, yes pumping the FSB probably does void your warranty, even though there is no reason it should.
 
For RAM, I'm using 2x1Gb @375Mhz, 1:1 ratio to match my e6750 which is sitting at 3.0Ghz lately. 5.9 on memory, 5.7 on CPU. Somehow I have a higher memory subsystem score than people using 4Gigs at higher speeds? I'm in the camp of "Vista Experience Index is completely meaningless."

(Now, a combination with, say, a miniature version of Sandra to develop this score, and not limit it the way they are, that'd might actually be pretty neat)
 
For RAM, I'm using 2x1Gb @375Mhz, 1:1 ratio to match my e6750 which is sitting at 3.0Ghz lately. 5.9 on memory, 5.7 on CPU. Somehow I have a higher memory subsystem score than people using 4Gigs at higher speeds? I'm in the camp of "Vista Experience Index is completely meaningless."

(Now, a combination with, say, a miniature version of Sandra to develop this score, and not limit it the way they are, that'd might actually be pretty neat)

With 1:1 at 375, if you are running tight timings... you probably have pretty good bandwidth. Run a sandra then drop your FSB to 200 with the ram at 2:1 and compare ;)
 
It's still not really... you know, useful. In any way. Rather hard to fault him.

Agreed when the audience is computer gurus...

But for a lot of people who know nothing about computers??!!?? I'd say it's much better than nothing. Think how many people know nothing of graphics cards and don't know why their game doesn't work. They don't understand what they need, but if the box says they need a graphics card with a rating of 4.0 to play and theirs is a 2.0 they understand in a quick glance that is the component they need to upgrade.
Some of the benchmarks are actually a computational benchmarks, that will change when your CPU is overclocked etc --- and not just a static rating. Think of it like a cheapo version of 3d mark with an easier to understand rating output. I think something like this personally is overdue in the industry. It's not perfect, but it's a pretty good start.
 
Agreed when the audience is computer gurus...

But for a lot of people who know nothing about computers??!!?? I'd say it's much better than nothing. Think how many people know nothing of graphics cards and don't know why their game doesn't work. They don't understand what they need, but if the box says they need a graphics card with a rating of 4.0 to play and theirs is a 2.0 they understand in a quick glance that is the component they need to upgrade.
Some of the benchmarks are actually a computational benchmarks, that will change when your CPU is overclocked etc --- and not just a static rating. Think of it like a cheapo version of 3d mark with an easier to understand rating output. I think something like this personally is overdue in the industry. It's not perfect, but it's a pretty good start.

So when they go to upgrade their graphics card, how do they know which card to get?
 
Mine went from 5.7 to 5.9 when I matched the memory speed with my FSB.
Overclocked my Q6600 to 3.0Ghz (8x375)
UNDERCLOCKED my memory (2x375 = DDR2 750)

Noticably faster and boosted my score :D

This. 1:1 is what hit 5.9 for me. It's not that my goal was to hit 5.9, it's just that I always try to run 1:1. I have a Q6600 at 3.2 (8x400) and 2x2GB Corsair XMS2 DDR2-800 memory at 5-5-5-18 2T.
 
So when they go to upgrade their graphics card, how do they know which card to get?

Good question as very little if any of the rest of the computer industry knows about, cares about, or even uses this numbering system.

LMAO, can you imagine asking a Best Buy school boy employee to find you a GPU that does at least a 3.0 on the Vista scale :confused::confused::confused:
 
Good question as very little if any of the rest of the computer industry knows about, cares about, or even uses this numbering system.

LMAO, can you imagine asking a Best Buy school boy employee to find you a GPU that does at least a 3.0 on the Vista scale :confused::confused::confused:

If they stratified the number system by making it more precise and with a much higher cap then got AMD, Intel, Nvidia, and the RAM manufactuers to go along with it... that would be awesome!
Wouldn't make any difference to us, but the average idiot would stop buying crap and actually understand value vs performance. Would be a lot harder to sell ultra video cards and extreme edition chips to the dummies but ohwell.
 
Yeah but, can you imagine asking a hardware maker to assign a low number to certain products? They would be cutting your own throat on selling certain items just to adhere to some almost random number system from an OS that is being replaced in late 2009. I guess it is a decent effort at some sort of standardization but it's not suitable for the retail side of the world. Perhaps it would be helpful to SIs and OEMs.
 
I see a simple solution here, forget the Vista score and use proven benchmarks, such as Everest, PCMark and 3DMark. Don't trust Microsoft! ;)
 
change the timing of your memory. I have 5.8 with 6 gb of ram. I change the timing and it move up to 5.9
 
Back
Top