Comparison review of the Diamond Monster 3D II and the Creative Labs 3D Blaster Voodoo2

July 6, 1998

The first thing jumping to the eyes when comparing these graphic cards is the layout of the boards themselves. As a fact, both boards are almost physically identical on every points ! So, it's not of a great surprise if their performance are quite close and only different drivers can explain the quite small performance gaps between these graphic cards.

However, the software bundled with each cards is quite different ! The Diamond Monster 3D II that I tested was an OEM version sold with absolutely nothing else than a video and a SLI cables that's it, that's all ! No user's manual, no installation information, nothing, nada, niet ! So, you will have to run the installation CD before to install the card if you want to make sure how to install it correctly ! The worst thing in all this is that there even are no indication of the back panel of the card to identify which one of the two jacks is the input and which one is the output !

On the other hand, the Creative Labs Blaster 3D Voodoo2 comes with a much better bundle. There are four CD including "G Police" by Psygnosis, "Incoming" by Rage Software and "Ultime Race Pro" by Microprose and of course the installation CD, a video cable, an SLI coupling cable, a small installation information pamphlet and finally clearly labeled connectors on the back panel. Of course all this software bundle does have its influence on the overall retail price but there is no big deal here, or is it !


Installation

Installing these cards is really not a problem. After the card has been installed in one of the available PCI slot by taking care to keep at least one free slot between your existing 2D card and the 3D card to make sure not to twist too much the video cable, you must connect one side the video cable included with the 3D card to the output connector of your 2D card and the other side to the 3D card input connector which will leave an empty connector on your 3D card to plug to your monitor cable.

Now, when installing the Monster 3D II a dialog box will prompt you to choose from several options. The option you will have to choose is "Drivers included on the manufacturer disk" and indicate you CD-ROM drive letter and give the full path of the Win95 directory on the installation CD if you have a Win95 version 4,00,950 installed on your computer. For any other OS please refer to the installation instruction included on the CD.

The installation of the Creative Labs Blaster 3D Voodoo2 is so similar to the installation of the Monster 3D II that you can refer to the installation information of the Monster 3D II as described above.

I don't know about you but one thing I've noticed after I installed both of these cards and used them to play several games is the fact that I constantly lost my original display setup as soon as I quit a game to come back to Win95. As a fact, I am constantly obligated to open the display properties and change my display properties to get back the Powerdesk mode of my Matrox Millenium card or even readjust my monitor settings to get back to my original setup ! I have no idea if this is caused by the 3D drivers of these cards, the game drivers or the DirectX5 drivers that I am suspecting very strongly but I must confess that it is very annoying and highly bothering.


The performances

How I tested.

All tests were performed by using the following setup:

Motherboard: AOpen AP58
CPU: AMD K6-233mhz
Memory: 32mb SDram (Hyundai)
Hard disk drive: Quantum Udma ST3.2A (3.2gig)
OS: Win95 4.00.950
CD-ROM drive: AOpen CD-936E (36X) 2D graphic card: Matrox Millenium
Soundcard: Turtle Beach Montego PCI A3D


The graphic cards drivers used were those included on the installation CD for both cards. No foreign drivers were used and all the options enabled were the default options. No tweaking has been applied neither in the cards control panels nor in the games control panels.

Note: The "Don't Synch buffer swaps to monitor refresh rate" option has been checked for every tests except for the 3D Winbench 98.

The results

Forsaken + AMD K6-233mhz
Graphic card 512 X 384 - 75hz 640 X 480 - 75hz
Diamond Monster 3D II 51.39 fps 51.33 fps
Creative Labs Blaster 3D Voodoo2 53.06 fps 53.71 fps

All the Forsaken FPS tests were conducted 6 times each and the mean value obtained is the one indicated in the above results.

As we can see here a difference of about 4.5% is in favor of the Blaster 3D Voodoo2 graphic card. This is making me suspect the idea that Creative Labs might have developed some slightly better drivers.


Incoming + AMD K6-233mhz 640 X 480 X 75hz
Diamond Monster 3D II 40 fps
Creative Labs Blaster 3D Voodoo2 41.7 fps

For the "Incoming Lux and Robur" framerate tests, I used to let the demo play for an hour and a half (for both tests) in order to smooth out as much as possible the mean FPS value because the FPS measure is taken a bit before this demo begin to run and the first value indicated is very high, so, by running the demo on a relatively long period of time we can reduce the effect of this first value to a non significative impact on the overall results.

Here again, a gain in favor of the Creative Labs Blaster 3D Voodoo2 can be observed by a factor of 4.2% which is beginning to show a tendency letting us suppose once again about slightly better designed drivers from Creative Labs.


Quake 2 + AMD K6-233mhz using 640 X 480 X 75hz
Graphic card Demo 1 Demo 2 Timerefresh
Diamond Monster 3D II 21.45 fps 15.95 fps 70.27 fps
Creative Labs Blaster 3D Voodoo2 25.02 fps 20.00 fps 73.28 fps

All the Timedemo tests have been performed 4 times each and the mean value obtained is the value indicated in the above table. All the Timerefresh tests were performed 6 times each and the mean value obtained is the value indicated in the above table.

The Timedemo test have been conducted as described here:

  • The computer has been rebooted
  • Quake 2 has been launched
  • The video console has been opened
  • The 640 X 480 mode has been choosed
  • The game console as been opened by typing " ~ "
  • Once in the console I typed "Timedemo 1" and "Enter"
  • The command "demomap demoX.dm2"
  • Once that the demo was finished, I used to quit the game
  • The computer has been rebooted for the next test.

    Note: X stands for 1 or 2 depending if I wanted to use demo 1 or demo 2.

    The Timerefresh tests have been conducted as following:

  • The computer has been rebooted
  • Quake 2 has been launched
  • The video console has been opened
  • The 640 X 480 mode has been choosed
  • A new game has been started at the easy level
  • After the explosions, the game console has been opened by typing " ~ "
  • Once in the console I typed "Timerefresh"
  • Once the demo was finished, I used to quit the game
  • The computer has been rebooted for the next test.

    Quake 2 + AMD K6-233mhz using 800 X 600 X 75hz
    Graphic card Demo 1 Demo 2 Timerefresh
    Diamond Monster 3D II 21.82 fps 15.97 fps 62.02 fps
    Creative Labs Blaster 3D Voodoo2 24.32 fps 17.18 fps 63.87 fps

    All the Timedemo tests have been performed 4 times each and the mean value obtained is the value indicated in the above table. All the Timerefresh tests have been performed 6 times each and the mean value obtained is the value indicated in the above table.

    The Timedemo test have been conducted as described here:

  • The computer has been rebooted
  • The graphic card control panel has been opened
  • The 800 X 600 mode has been choosed
  • Quake 2 has been launched
  • The video console has been opened
  • The 800 X 600 mode has been choosed
  • The game console as been opened by typing " ~ "
  • Once in the console I typed "Timedemo 1" and "Enter"
  • The command "demomap demoX.dm2"
  • Once that the demo was finished, I used to quit the game
  • The computer has been rebooted for the next test.

    Note: X stands for 1 or 2 depending if I wanted to use demo 1 or demo 2.

    The Timerefresh tests have been conducted as following:

  • The computer has been rebooted
  • The graphic card control panel has been opened
  • The 800 X 600 mode has been choosed
  • Quake 2 has been launched
  • The video console has been opened
  • The 800 X 600 mode has been choosed
  • A new game has been started at the easy level
  • After the explosions, the game console has been opened by typing " ~ "
  • Once in the console I typed "Timerefresh"
  • Once the demo was finished, I used to quit the game
  • The computer has been rebooted for the next test.

    Unreal + AMD K6-233mhz
    Graphic card 640 X 480 - 75hz 800 X 600 - 75hz
    Diamond Monster 3D II 19.10 fps 17.63 fps
    Creative Labs Blaster 3D Voodoo2 22.65 fps 20.01 fps

    The Unreal timedemo test is an average frame rate benchmark. To have a valid reading with this game benchmark you have to let the flyby intro screen play for a while. So, I used to play the flyby intro screen for half an hour to have a very well stabilized frame rate value and this is the value indicated in the tests results.

    If a graphic card has some weakness it is with this very complex textures loaded game that it will show up. As a fact, Unreal is the most incredible game I've seen to date when talking about complex textures and special effects of all kind. Under such a heavy load, we can clearly see that the Monster 3D is lagging behind the Blaster 3D by about 18.6% in the 640 X 480 mode and by about 13.5% in the 800 X 600 mode, so, make your own conclusions !


    Wizmark + AMD K6-233mhz
    Graphic card Number of frames Elapsed time FPS Benchmark value
    Diamond Monster 3D II 1059 58.65 sec. 18.06 20160.78
    Creative Labs Blaster 3D V2 1059 58.49 sec. 18.11 20291.73

    Wizmark is a 3D performance benchmark created by 3Dfx Interactive, Inc. that measures the performance of Microsoft's Direct3D(tm) and 3D hardware accelerators. The benchmark utilizes essential 3D rendering features which provide enhanced visual realism. These features include: Gouraud shading, perspective correct texture-mapping, Gouraud modulated texture mapping, texture bilinear filtering, texture LOD mip mapping, alpha-blending, and Z-buffering. WizMark is based on models and artwork from 3Dfx's "Wizard's Tower," a first person perspective walk through of a series of connected rooms.

    The value indicated are the mean values obtained after I ran the test 6 times in a row for each cards.

    Again, a slight difference is in favor of the Blaster 3D Voodoo2 by about 6/10 of 1%. However, this difference is so small that we could say that we have a drawn game here with this benchmark.


    3D Winbench98


    Even if this Benchmark is rather controversial, I personally choosed to use Ziff Davis 3D Winbench98 to bring additional information. As a fact, some rumors have circulated (and are always circulating) that some graphic cards manufacturers would design their drivers to have a better score with this Benchmark. Of course this would give unrealistic performance with real world applications but this is precisely where are coming in help the games Benchmarks by giving a more precise picture about how are really performing graphic cards under a real world environment.

    With this last test we see no significative differences between both these Voodoo 2 based graphic cards and as a fact they come to a drawn game once again so we can at least know for sure now that none of them is using drivers that would artificially inflate its results with this benchmarks !


    Conclusions

    My conclusions are that all of this only reveal that none of these Voodoo2 based graphic cards are bad and it would have been totally unlogical to come to this conclusion because as a matter of fact both cards are identical ! So, if the Monster 3D has showed some weaker performances than the Blaster 3D it is only because of drivers probably not as well designed as those of the Blaster 3D. In my book, this is only good news because it means that Diamond will probably correct the situation on their next drivers updates so if you own a Monster 3D II keep an eye on Diamond drivers library because it could mean a 5% to close to 15% increase in the performance of this card under some cirumstances.

    Finally, there is another factor to consider when choosing between these cards and it is the price they are sold. So, one might prefer to buy an OEM version without any software bundled to pay less while some other could prefer to get the full retail version and have a good software bundle ready to use even if this means that they will have to pay a little more for it.


    Click here to go to the Monster 3D II page on Diamond Multimedia web site for additionnal informations.

    Click here to go to the Blaster 3D Voodoo2 page of Creative Labs web site for additionnal informations.

    Diamond monster 3D II specifications

    Blaster 3D Voodoo2 specifications


    		•