The Notebook Review forums were hosted by TechTarget, who shut down them down on January 31, 2022. This static read-only archive was pulled by NBR forum users between January 20 and January 31, 2022, in an effort to make sure that the valuable technical information that had been posted on the forums is preserved. For current discussions, many NBR forum users moved over to NotebookTalk.net after the shutdown.
Problems? See this thread at archive.org.

    The WEI

    Discussion in 'Windows OS and Software' started by Peon, Feb 20, 2011.

  1. Peon

    Peon Notebook Virtuoso

    Reputations:
    406
    Messages:
    2,007
    Likes Received:
    128
    Trophy Points:
    81
    Has Microsoft ever explained how the WEI is actually calculated?

    I get a better memory score in my VM than I do on the hardware the VM runs on, which leads me to wonder if Microsoft is simply averaging out a relative score based on all of your hardware...
     
  2. RainMotorsports

    RainMotorsports Formerly ClutchX2

    Reputations:
    565
    Messages:
    2,382
    Likes Received:
    2
    Trophy Points:
    56
    Seeing as how I have had memory and cpu scores change from overclocking would lead one to think otherwise. Though one could argue they were still just looking at the current clocks. I have had bad driver releases take a score down even though the hardware id's and performance related info were reported the same so i have my doubts against people who think WEI doesnt bench the hardware.

    Fact is they do benchmark several things, those are obvious. Plenty that isn't obvious unfortunately. But all those who think microsoft just has a nice little database to compare against. There is way too much hardware out on the market then they could ever get around to even glancing at.

    And yeah while Vista and 7 scores differ a bit, I get about the same with 7 in a VM as i do on vista and its pretty close to what i had on 7 beta. I think it benches like anything else. Its just not a very good or very extensive test.
     
  3. Peon

    Peon Notebook Virtuoso

    Reputations:
    406
    Messages:
    2,007
    Likes Received:
    128
    Trophy Points:
    81
    Oh, I have no doubts that WEI benchmarks things. What I'm beginning to suspect is that they take the raw benchmark results, calculate (and potentially modify) the mean and use that as the base score, and then apply +/- offsets to the mean for each component.

    That way, they can just inflate the mean for people who have ancient computers where everything is slow so that those people won't feel too bad, while maintaining meaningful variance if one component is 10 years old while the rest of the computer is brand new.
     
  4. namaiki

    namaiki "basically rocks" Super Moderator

    Reputations:
    3,905
    Messages:
    6,116
    Likes Received:
    89
    Trophy Points:
    216
    Here's the disk result for my C300. (scores a 7.7)
    RAW OUTPUT Rw8pZyEU
    Got that from command prompt, but I bet that there's an easier way somewhere. :p

    I can't seem to get any particular numbers for the other components.
     
  5. MahmoudDewy

    MahmoudDewy Gaming Laptops Master Race!

    Reputations:
    474
    Messages:
    1,654
    Likes Received:
    744
    Trophy Points:
    131
    The guy at tweakguides.com has the complete explanation of how each and every single value of the WEI is being calculated ... download the free tweaking guide & check the WEI section & you will understand how things are done
     
  6. namaiki

    namaiki "basically rocks" Super Moderator

    Reputations:
    3,905
    Messages:
    6,116
    Likes Received:
    89
    Trophy Points:
    216
    Page 388, I guess.
     
  7. davepermen

    davepermen Notebook Nobel Laureate

    Reputations:
    2,972
    Messages:
    7,788
    Likes Received:
    0
    Trophy Points:
    205
    yes. read it on a blog on microsoft.com with detailed description and references to msdn and other microsoft sources that openly and in detail explain nearly everything about microsofts products since years.

    i'm too lazy to search the info for you together.

    but vm's can sometimes be faster than real computers, as they don't emulate a computer exactly. they can thus have different bottlenecks and different performance behaviour. mostly, they're slower. but there are occasions where they're faster.
     
  8. RainMotorsports

    RainMotorsports Formerly ClutchX2

    Reputations:
    565
    Messages:
    2,382
    Likes Received:
    2
    Trophy Points:
    56
    Interesting and im not sure it would be entirely wrong of them to impose such a limitation on score due to the size being an acutal handycap. Mind you they already have a cap thats raised each os versions so that the system does not go to hell in one day.

    32 Bit 7 and 64 bit 7 have different assume requirements, just like Vista and 7 have differences in requirements and score setup. So for them to impose a difference in score on 64 bit machines which are expected to require more ram right off the bat is fair to the system they are using, but not the hardware's performance.

    I know handicaps are not a part of normal benchmarking and seem like a horrible idea to most people.
     
  9. davepermen

    davepermen Notebook Nobel Laureate

    Reputations:
    2,972
    Messages:
    7,788
    Likes Received:
    0
    Trophy Points:
    205
    the limits make perfect sense, though. as it's a benchmark for overal system performance, it matters how much ram one has. which wouldn't in a synthetic benchmark as long as it's enough to run the test in it.

    and as the 64bit os uses more ram (and more diskspace), it makes sense that 4gb for the 64bit os doesn't get the same result as 4gb for the 32bit system.
     
  10. KimoT

    KimoT Are we not men?

    Reputations:
    560
    Messages:
    1,128
    Likes Received:
    0
    Trophy Points:
    55
    I get a 6.9 score for memory with 4 GB DDR2 800 RAM.
     
  11. Peon

    Peon Notebook Virtuoso

    Reputations:
    406
    Messages:
    2,007
    Likes Received:
    128
    Trophy Points:
    81
    The host has 16 GB of RAM, the guest had either 4 or 8 GB. I don't exactly remember as the guest was a "test a piece of software for an hour and then throw away the VM" type of situation.

    Ironically, I didn't run into the hard drive cap either. Despite the VM existing on a WD Caviar Green, the overall WEI score was 6.0 (graphics was the lowest) with the disk score being something like 6.2 or 6.3.

    Either way, I don't put much faith in WEI scores (the host's lowest score is actually the CPU, and the only faster CPUs out there are the 2 Gulftowns), but it's still very interesting to know how these scores are calculated so I'd like to thank everyone for their insightful posts.