Jump to content
HWBOT Community Forums

DrSwizz

Members
 View Profile  See their activity

  • Posts

    111

  • Joined

  • Last visited

 Content Type 

Posts posted by DrSwizz

    Processor specifications request: Embedded Pentium MMX 266

    in Support

    Posted · Edited by DrSwizz

    can you add a pic of that embedded cpu then?

     

    I already have a picure somwhere on a HDD. I'll dig it up for you.

     

    since cpu-z can't tell the difference between the mobile and embedded ones, there will wrong results end up in that category as well i think...

     

    Huh? You mean people will post results in the wrong category? I very much doubt that would be a large problem since these particular CPUs are rather unusual.

     

    Edit: Have uploaded the picture here:

    http://imageshack.us/photo/my-images/840/intelembeddedpentiummmx.jpg/

    Please add hardware to the database

    in Support

    Posted

    i think under "GPU" in everest you can see the difference in specs as well.

     

    i think the only difference between the "dvd" one and the other is, that the "dvd" has "dvd extensions" which put less load on the cpu back then so you could watch dvds more fluid...

     

    Well, that is what I thought too until I ran benchmarks on them. The differences are larger then that. The drivers treat the GPU differently and their performance are are differnt.

    Please add hardware to the database

    in Support

    Posted

    ok, i'll add it. but they only run 3dmark 99 and 00 i think?

     

    Yes, only those.

     

    and btw. can the tools (everest) etc. show the difference between those cards? (ATI 3D Rage II and ATI 3D Rage II+DVD)

     

    Well, in those screenshots I actually included three programs for identification:

    Rivatuner & CPU-Z only shows whatever that the drivers presented to them.

    Powerstrip shows GPU type + memory type & size.

    My benchmark submission got reported despite that I did follow the rules.

    in Submission & member moderation

    Posted

    Please Dr.Swizz, any other soft like Aida or everest , that shows the clocks of the gpu ?

     

    That would be enough, I think, by common sense, we approve this submission ;)

     

    Okay, I will run this benchmark again then and use Powerstrip to shown the clock frequecies. I cannot do this right now though as I have disassembled the rig that I used for the benchmark being discussed here, but I will do it some time in the future.

     

    That being said I have submitted several other benchmark results in past where I was unable to use GPU-Z & where I have not showed the clock frequencies since I was not aware of that this requirement. I also have a few old benchmark results that I have not yet submitted (still waiting for hardware to be added to the HWBot database) that where rather difficult to run. I really would not like to be forced to rerun all those benchmarks too as it would require a significant amount of work (old, oddball hardware is easily 10X more time consuming to benchmark than new hardware).

     

    Also please update the rules so they do excplicily state that clock frequencies must be included in the schreenshots, otherwise how are people supposed to know about this requirement?

    Especially considering that for many of the combinations of hardware & software where GPU-Z fails, Rivatuner does not work 100% either and it is not possible to view & adjust the clock frequencies from within Rivatuner.

    My benchmark submission got reported despite that I did follow the rules.

    in Submission & member moderation

    Posted

    GPU clock speeds are missing.

     

    Well, Rivatuner does not report that.

     

    hmm, did you try an older gpu-z version. maybe it will work?

     

    Yes, I have tried a few older versions + I tried using different settings for kernelex. The problem with GPU-Z seem to be that it want to install some sort of driver that is made for WinXP.

     

    Anyway, since the rules clearly state that Rivatuner can be used instead of GPU-Z when GPU-Z does not work, is there any reason why my result should not be accepted?

    My benchmark submission got reported despite that I did follow the rules.

    in Submission & member moderation

    Posted

    Hello!

     

    One of my benchmark submission got reported as lacking information of what graphics card I was using for the benchmark. However I did clearly show the graphics card information using Rivatuner in my screenshot (I hade to use Rivatuner instead of GPU-Z as I was running the benchmarks using Windows 98 SE and GPU-Z does not work with that OS).

    Could someone please look into this? I tried to explained all this in the "comment" entry box for my reported submission, but nobody seems to care.

     

    This is the benchmark result in question:

    http://www.hwbot.org/submission/2248068_drswizz_3dmark2001_se_rage_128_pro_1852_marks

     

    The type of graphics card + the device ID of the card are clearly visible in the screenshot.

    [RANT] How to make sure your awards/reviews are totally meaningless

    in Offtopic

    Posted

    Awards: completely useless measurement

     

    Last year I was considering buying an Asus 890G motherboard (I forget which model). I did some googling, but found very little information about the motherboard as it was a relatively new product. On the official Asus page for that particular motherboard there was 2-3 awards given by some obscure hardware websites. I decide to take a look at those sites. It turns out that none of those websites had actually reviewed the motherboard in question. On one of the websites there was a review of one motherboard with similar sounding model name, but with different hardware. On the other site(s) I found no review whatsoever of any motherboard even similar to the one I was considering buying. I never bought that Asus motherboard

    Remove SiSoft Sandra?

    in Benchmark software

    Posted

    @Massman

     

    One of the reasons there are so few benmchmark submissions SiSoft Sandra is probably because the rules specifies SiSoft Sandra 2007, a version which does not seem to be available for download anymore. I did some googling for it a few weeks ago, but could not even find it on any third-party websites.

    Bug reports : UCBench bugfix request

    in Support

    Posted

    Ticket ID: 1315

     

    Priority: Low

     

    Please fix so that UCBench works on CPUs that lack support for CPUID.\r\n\r\nThe benchmark itself works just fine (it runs the generic 386 computation core), but the CPU detection fails and no output file is produced.\r\nHere is a screenshot of what it looks like when I run the benchmark on a Texas Instruments 486 class CPU:\r\nhttp://img688.imageshack.us/img688/2177/ucbenchoncpuwithoutcpui.png\r\n\r\nUCBench is rather well suited to run on old & slow CPUs that lack CPUID because it does not take hours to run the benchmark, so this is a bug that is worth fixing imho.\r\nYou might also want to add one more number after the decimal point to there score to add some more granularity (from x.y pwds/s to x.yz pwd/s).\r\n\r\n\r\nAlso, please do not forget this:\r\nhttp://hwbot.org/forum/showthread.php?t=27348

    [BUGS] hwbot: Revision 4 up and running!

    in HWBOT Rev.4

    Posted

    HWBot broken, data corruption?

     

    Please forgive me if this is an issue that the HWBot crew already is aware of, I have not actively been following this thread.

     

    I was idly browsing the results database and I have seem to have come across a serious bug or possibly database corruption.

    I was viewing CPU-Z results of old CPUs and when I clicked on the validation link for one of Christian Neys results the validation link was for a the wrong result. I figured he must have made a mistake and I made a mental note to send him a PM later. Then I came across another CPU-Z result that he had made with a link that also was wrong, in fact both links where to the same result. This made suspicious; Perhaps v4.x of HWBot has broken something? So I started browsing his CPU-Z results and it turns out that plenty of them have links are wrong (all Pentium-class CPUs?).

×
×
  • Create New...