Voodoo5 6000 und Voodoo5 6000 by Anthony (Snowhite) compabilty

    Zitat von Löschzwerg
    Zitat von Gold Leader

    3D Mark 2001SE is pretty useless,

    It's interesting from a technical standpoint and for testing the different limitations / bottlenecks. But that's surely just something for hardware nerds like Raff and myself ^^

    it only shows a random number on how that shady 3D Mark works, which has no real value on how the system actually would perform in the games or simulations it's made for.

    I never needed a shady program to tell me how my system performs, my games & flight simulations give me all the accurate & proper information I need to see how my system performs and if it's not that, I use Quantum3D's Open GVS instead, far more accurate software to use.

    If you really want to stress the system, just run Freespace 2 fly in to the Nebula and see your system struggle.

    Einmal editiert, zuletzt von Gold Leader (20. Februar 2022 um 23:24)

    The key values of 3DMark00/01 lie within the feature tests (fillrate, pixel/vertex...), that's where you can see limits/bottlenecks. If you know how to handle these values 3DMark is a very nice and convenient tool. It easily shows if you are in the ballpark when it comes to performance.

    I guess hat Obi Wan means, is that 3dmark 2001 is a test meant to be failed by the Voodoo cards. 3dmark01 is a using the DX8 render path the Voodoo cards have not been designed for. They dont have any hardware T&L or pixel shaders. All these requiered hardware features are done by software and therefore slow as fuck. So it is not 'showing' appearing bottleneckes, it is showing things the Voodoo is not capable of. Its like throughing a car into the ocean to proof how bad it swims.

    Also what poeple are compareing are the final scores and not certain areas of 3d mark benches. It would make sense the share values of game 1 or 2 scenarios or bandwith tests after tweaking things, but this is not what happens.

    To showcase the limitations of bandwith etc. you can perform these tests using Q3, for example, pretty well. Just turn of the texture compression and increase the resolution and you will find the sweet spot quiet fast where the AGP1x interface reaching its limitations.

    Yes, it is a banchmark, but in the end it shows how bad a car is swimming and not how fast it can drive.

    Well that's totally missing the point about the initial question from Raff. The whole idea about a "modern" Voodoo system is getting more performance in scenarios where a beefy CPU is of benefit. And this if when a tool with specific tests like 3DMark comes in handy. Simple as that.

    To figure out the impact of potent cpus or platforms in general 3d mark 2001 is not the perfect test to use for Voodoo cards either. It could be usefull if stronger cpus would show a linear increase of points till a certain point where limitations break through. And then with a focus on specific tests you could read when and where a hardware limitation has been reached. But this is not the case in 3d mark 01.

    The point is that a stronger cpu will inceases points on high polygon tests due to software T&L, but on the downside you will miss improvements where the Voodoo cards already pasts their hardware limitations, because of missing hardware support.

    For example, for tests including pixel and vertex shaders the Voodoo will be slow down a lot and wont show any or only a very few increased framerates while in cases without those shaders a faster cpu would lead into noticable higher framerates. So what will be the specific conclusion out of these results when you cant tell, if the low results were now caused by software rendering or hardware bottlenecks?

    I might be missing something, but honestly I never saw any particular and detailed results or comparisons specifically made out of 3d mark 01 testing. Or can you provide any report or knowledge about the Voodoo cards only available thanks to 3d mark 01? Also cant remember Raff himself using any 3d mark results within his Voodoo articles (but might be lack of memory).

    If you want to figure out the impact of potent cpus on the gpu, you will beginn using low resolutions and increase or use stressy scenarios like high count bot matches. Otherwise it would be legit to test todays Radeons under the use of all Nvidia specific hardware features to show their bottlenecks.

    Ofcourse it is nice to increase points in 3d mark benchmarks. It is like beating the time shown on navigation deviced in cars, but in both cases the success does not mean much at all.

    Zitat von DerBiber

    I guess hat Obi Wan means, is that 3dmark 2001 is a test meant to be failed by the Voodoo cards. 3dmark01 is a using the DX8 render path the Voodoo cards have not been designed for. They dont have any hardware T&L or pixel shaders. All these requiered hardware features are done by software and therefore slow as fuck. So it is not 'showing' appearing bottleneckes, it is showing things the Voodoo is not capable of. Its like throughing a car into the ocean to proof how bad it swims.

    Also what poeple are compareing are the final scores and not certain areas of 3d mark benches. It would make sense the share values of game 1 or 2 scenarios or bandwith tests after tweaking things, but this is not what happens.

    To showcase the limitations of bandwith etc. you can perform these tests using Q3, for example, pretty well. Just turn of the texture compression and increase the resolution and you will find the sweet spot quiet fast where the AGP1x interface reaching its limitations.

    Yes, it is a banchmark, but in the end it shows how bad a car is swimming and not how fast it can drive.

    That exactly every damn review I read has the same boring 3D mark tests in them why don't we get to see how the latest glide games actually perform games like Descnet3, Freespace2, Wing Commander V, Dark Forces II Jedi Knight, Star Wars X-Wing Alliance, Falcon 4.0 + Free Falcon 3.2, SSI's Flanker 2.0 with 2.51 3dfx Voodoo5 patch, not your everyday games & Simulations I am sure there are tons more than UT or Q3A, some different tests would truly be welcoming Need for Speed 3,4 & 5 and some other good Racing & Rally games not sure if we had a Colin McCray's Dirt Rally tho, a test of 50+ games of all 3dfx Glide cards V1 to Voodoo5 6000 that would be something really worth reading and looking in to over the same boring 3D Mark UT & Q3A tests.

    Just sick and tired of reading the same review every year I just stopped putting time in to them because it's the same old song over and over and offers nothing important or new to read.

    Thanks for understanding me bud :)

    DerBiber You get nice theoretical values for very specific tasks. Those aren't necessarily relevant in games, if at all. I don't disagree. They are however relevant to see every bit of system performance, no matter practical useful or not. Why? Why not! Is there anything else to see or test what we haven't seen already? UT @ 640x480 and a highend CPU to see the benefit? Sure, why not :)

    And to be honest, we all have seen so many game tests and also 3DMark values aside... If someone runs 3DMark and gets a nearly identical score, one can probably estimate that certain game will run fine as seen in the test/review. That's the overall benefit of a synth benchmark.

    Ofcourse it is interesting to see how the Voodoo 5 is able to handle next gen software (DX 8 etc.) and also to see playable framerates in UT2003. Nothing wrong about that.

    So maybe it is more correct to say that these results on 'modern' tests do not necessary speak for timeline correct gaming performance and vice versa.

    just irrelevant data for me every year will be the same and boring as these reviews tell you basically nothing new, a true 3dfx Glide shootout never happens it's always the same shit different day kind of result.

    Just a total waste of time

    Hallo zusammen,

    ich finde so eine Liste echt sehr gut! Durch Anthonys Nachbau sollte der Bedarf ja dementsprechend ansteigen :)

    Allerdings ist mir etwas aufgefallen, was man evtl. genauer beleuchten sollte:

    Zitat von tyrannus

    Super Sache! Ich hatte schon ein leicht schlechtes Gewissen, dass ich quasi den Vorstellungsthread zu Anthonys 6k Version "missbraucht" habe mit meinem Kompatiblitätsthema.

    Eine Zusammenfassung der mittlerweile in verschiedenen Threads hier zusammengetragenen, sehr wertvollen Infos an eine Stelle ist jedenfalls ne top Idee. Und es hat mir schon eine weitere wichtige Info für mein potentielles System gegeben, laut dem link von @cicara hat Amigamerlin selbst mal das MSI KT3Ultra-II mit ner 6k getestet und für gut befunden. Das lässt doch hoffen, dass es nach einer Messung der AGP Spannung nutzbar ist.

    Ja, die 6k läuft super im MSI KT3 Ultra-II. Leider gab es anscheinend auch wenige Boards mit dem KT333-CF Chipsatz (umgelabelter KT-400), was natürlich inkompatibel ist.

    Jedoch selbst wenn ein "echter" KT333 verbaut ist, gibt es einige Berichte dass eine V56k bei längerer Nutzung darin stirbt!

    Aus diesem Grund hab ich davon abgesehen und werde ein anderes Board verwenden. Gibt es da noch mehr Erfahrungsberichte? Falls sich das bestätigen sollte, müsste man das Board leider von der Kompatibilitätsliste streichen.

    Viele Grüße,

    Wuuki

    - hier könnte Ihre Signatur stehen -

    Darum gehts mir nicht. Ich habe schon ein Chaintech CT-7VJL bei Backfire zum Recappen, was für die 56k herhalten darf :)

    Ich wollte nur darauf hinweisen, dass ein KT3 Ultra II entgegen der Kompatibilitätsliste bei längerem Betrieb eventuell doch inkompatibel sein könnte.

    Gruß,

    Wuuki

    - hier könnte Ihre Signatur stehen -

    It`s already on the Voodoo5 6000 compatibility test. Anyhow, I tested today Anthony`s Snowwhite on a Gigabyte GA-7VTXE (KT266A, Rev 1.1, latest Bios F9), works perfectly fine as expected.

    Zitat von Tormi2k

    It`s already on the Voodoo5 6000 compatibility test. Anyhow, I tested today Anthony`s Snowwhite on a Gigabyte GA-7VTXE (KT266A, Rev 1.1, latest Bios F9), works perfectly fine as expected.

    WHat list? Not the one I made i am sure, just hope it's stable 3 a 4 months from now ;) And hope for no issues.