Announcement

Collapse
No announcement yet.

The true Power of the Parhelia reveiled - READ

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • The true Power of the Parhelia reveiled - READ

    last week I bought the game Halo. I thought well that it probably would run a 800x600 on my Parhelia. So after I installed it, I gave it a try. The game ran very smoothly, with every detail on max. So I thought lets pump up the resolution to 1024x768. And it still ran pretty good.

    I was thinking to myself if there are already some benchmarks on the web. I started searching and I ended up with the review at extremetech. Which benchmarked halo with the NVidia FX 5600,5900 and the ATI 9600Pro and 9800pro.

    I found how to benchmark the game and so it happend.

    My result : Average frame rate=26.39fps
    here you can find the results at techextreme: http://www.extremetech.com/article2/...1354067,00.asp

    here are my systemspecs:

    - Intel P4 1,8 -> 2.4 ghz (no hyperthreading) FSB 133
    - 512MB pc3200 running at -+ 187 (378) mhz
    - ABIT BD7-II with the 845 chipset (single channel)

    - Philips Acoustic Edge (which isn't nearly as fast as the audigy 2, even when eax is disabled).

    As you can see the Parhelia with my system is just as fast as an 9600pro and twice as fast as the FX5600 ULTRA! And my system is proberly a bit slower than their 2400 configuration.

    This weekend I received Max Payne 2 too (which also uses alot of pixelshaders) and the game runs very good at a resolution of 1024x768.

    nowadays the big review sites don't include the Parhelia anymore, because it wasn't fast enough in the games that were available back then. But remeber, those games didn't use pixelshaders yet. I am talking about games like Quake 3 based games, Half-life, etc. Games that are more dependant of the raw fill rate power of the card which the FX series and 9700 and newer had alot more.

    But now more and more games are using pixelshaders in their game. Games like Halo, Max Payne 2 and future games like Half-life 2, stalker, farcry. And it seems that the Parhelia is catching up.

    I think it was Haig that said back then that the Parhelia wouldn't have the same raw power for in those games. but when games will take advantage of pixelshaders the power requirement would shift from raw fill rate to pixelshader performance. And it looks like he is right!

    How come that in 3dmark 03 the parhelia performs poorly, I think we should ask Mado... I mean futuremark. Proberly due to lack of optimization.

    But all those big sites already bashed the Parhelia. And they don't include them anymore (even not Digit-life.com in their 3d digest). Why because it wasn't fast enough in old generation games.

    I won't say that the parhelia has some negative points (like that the FAA-x16 doesn't work always) But they did bash it. They should be ashamed.

    They said the Parhelia wasn't a gamer card, but in Halo it performs just like the 9600pro and twice as fast as the FX5600 PRO. Why don't they call those cards NON gamer cards. Agreed the Parhelia is more expensive. but thats not the point here (because you get alot of features and offcourse great support).

    My request is anyone who read this (who works for a review site) to add the Parhelia back again to their reviews with the newer games!

    thank you!
    Last edited by CaineTanathos; 20 October 2003, 12:59.
    Hey! You're talking to me all wrong! It's the wrong tone! Do it again...and I'll stab you in the face with a soldering iron

  • #2
    Hi Caine,

    good to hear that the P does well in Halo, and there are more games and benches now in which the Parhelia is performing very good (Aquamark 3 for example).
    You are totally right about the reviews, althought there were some good Parhelia reviews online (3dchip.de in germany for example) most of them and all of the pc-magazines bashed it.

    (Further, in my opinion the performace wasn't that bad even with older games, another point was the big performance increase in 3D3 with later drivers...)

    But you said the Parhelia lacks on fillrate performance, that's not true. Parhelia probably has a problem with (very ) high polygon scenes because of the low core clock, but the multitexturing fillrate performance of the Parhelia is awsome, better than any Ti4600, and the singletexturing just as good as a GF4 Ti...
    With a clock as a R9800 or something like that the P-512 Chip would stamp them all by fillrate.

    But great statement anyway, thanx
    P IV 3,06 Ghz, GA-8ihxp i850e, 512 MB PC-1066 RDRam, Parhelia 128 mb 8x, 40 + 60 gb IBM 7200 upm/2048 kb HD, Samtron 96 P 19", black icemat, Razer Boomslang 2100 krz-2 + mousebungee, Videologic sonic fury, Creative Soundworks

    Comment


    • #3
      Correct me if I'm wrong but isn't the ATI and Nvidia video cards using DX9 drivers and therefore using DX9 features that the games supports?

      The Parhelia has no DX9 drivers right?

      Dave
      Ladies and gentlemen, take my advice, pull down your pants and slide on the ice.

      Comment


      • #4
        when I use I:\Program Files\Microsoft Games\Halo\Halo.exe -use11 -timedemo
        I get 28.58fps
        Hey! You're talking to me all wrong! It's the wrong tone! Do it again...and I'll stab you in the face with a soldering iron

        Comment


        • #5
          Still the Parhelia is using the DX8 codepath (pixelshader 1.1), whereas both the Radeon and the GF FX are using the much more GPU intensive DX9 pixelshader 2.0 in the bench you provided.

          For an apples-to-apples comparision (all cards using Pixelshader1.1, all on a P4@2.4GHz, in 1024x768 resolution) it would look as follows:
          21 fps GF FX5600
          29 fps Parhelia
          31 fps Radeon9600 Pro
          48 fps GF FX5900Ultra
          55 fps Radeon9800 Pro
          Last edited by Indiana; 20 October 2003, 16:09.
          But we named the *dog* Indiana...
          My System
          2nd System (not for Windows lovers )
          German ATI-forum

          Comment


          • #6
            Originally posted by Helevitia
            Correct me if I'm wrong but isn't the ATI and Nvidia video cards using DX9 drivers and therefore using DX9 features that the games supports?

            The Parhelia has no DX9 drivers right?

            Dave
            They're being compared with PS1.1 so it's an equal playing field.

            Comment


            • #7
              Halo benchies

              OK in response to your post, which I thought was a great post btw I just tried the Halo benchmark myself. With a humble Athlon 1800XP I scored 26.15 FPS with my GeForce 4800SE. Now that's more or less equal to your P, only with a much slower CPU (Athlon XP @1.5GHz v your P4 @ 2.4GHz, unless I misread your post).

              Did you have sound enabled or disabled? I guess from reading your comments about your soundcard you have sound enabled. For the record I ran using sound enabled with my old Creative Labs Live! 1024. Since I noticed errors previously with hardware sound enabled I am running with hardware sound diabled.

              Someone with more knowledge than me will have to comment on the differences in shader version between P and GF4, the timedemo.txt reports I am using shader version 1.3.

              The GF4 cost less that 1/2 a Parhelia when I bought it, if I wanted to I could *upgrade* to a Radeon 9600Pro today and still have change. That's why review sites say the P is not a gamer's card, based primarily on bangs for the buck (as opposed to other features which a pure gaming user is not necessarily looking for). (not saying they are right, but that is what they would say).

              Let me know if you check your settings and I will run again for comparison as it is very interesting.

              Thanks

              Jon
              XP1800+ MSI 746Ultra 256Mb Corsair PC2700 MSI GF4800SE SBLive! WD 80GB 7200 DVD CDRW

              Comment


              • #8
                hmm, am I missing something here? Based on the PS 1.1 benchmarks, the P that CaineTanathos has is ahead off all of the cards and tied with the Radeon 9800?



                EDIT: Clarifying what I am trying to say...
                Last edited by Helevitia; 20 October 2003, 14:56.
                Ladies and gentlemen, take my advice, pull down your pants and slide on the ice.

                Comment


                • #9
                  sound enabled, but eax is disabled, I have an acoustic edge, which is pretty slow in games
                  Hey! You're talking to me all wrong! It's the wrong tone! Do it again...and I'll stab you in the face with a soldering iron

                  Comment


                  • #10
                    helevitia switch the 1600 results with the 1024

                    so its still faster than the FX5600
                    Hey! You're talking to me all wrong! It's the wrong tone! Do it again...and I'll stab you in the face with a soldering iron

                    Comment


                    • #11
                      Yes, the Parhelia is about on a level with the Radeon9600Pro. Unfortunately the 9600Pro can be had for about 150€, though.
                      You could even get a R9600 non-Pro for ~105€ and oc it to >9600Pro levels while the Parhelia (according to reports from those that've tried) is an absolutely lousy oc'er.

                      But still, despite its low price, a Radeon9600 Pro is in no way a slow card. So in conclusion:
                      Yes, the Parhelia delivers playable framerates in most games and thus IS usable as a gaming card.
                      BUT it's unfortunately much too expensive when you just want it as a gamers card and don't need the extra features like triple head, dual overlays (should I add the banding here? )
                      Last edited by Indiana; 20 October 2003, 16:19.
                      But we named the *dog* Indiana...
                      My System
                      2nd System (not for Windows lovers )
                      German ATI-forum

                      Comment


                      • #12
                        If Matrox had just priced it at $200 from the beginning, it would have been immensely more successful. If they had continued 6months later with P8x at $200 and dropped P to $150 or so, Matrox would be sitting on a very nice share of the gamer's market right now.

                        Comment


                        • #13
                          Might I add that were Parhelia priced lower, the market share, the developer support and surround gaming installed base would take off, thus creating momentum and marketshare.

                          Comment


                          • #14
                            I highly doubt that. but then again, i have my reasons.
                            "And yet, after spending 20+ years trying to evolve the user interface into something better, what's the most powerful improvement Apple was able to make? They finally put a god damned shell back in." -jwz

                            Comment


                            • #15
                              Perhaps with this new info. Matrox should push to get the P back into the reviews. Perhaps they could write off all the P512 chips of as 'bulk' supply and run a new batch with the increased yield and a higher clock. They may be able to push 260mhz with a new run. Label it as a Parhelia MAX and push it out the door. If this MAX version could compete then they may sell a lot of 'bulk' stock.

                              Dont forget Matrox were very good at pushing EMBM into the hearts of developers when they had a competitive card in the G400 MAX.

                              Comment

                              Working...
                              X