GP2, P5 or P6? Which one is faster?

what i understand is that P6 runs no-good in 16 bit application, so i presume P5 166 would outrun P6 [in GP2], is this right? Or P6 200/150 would still outrun P5 166? Also GP2 would be Win95 compatible, does this mean it would also be NT compatible? I am intending to upgrand to NT 4.0 from Win95 as soon as it comes out. Anyone know when NT 4.0 is coming out? Chris

>what i understand is that P6 runs no-good in 16 bit application, >so i presume P5 166 would outrun P6 [in GP2], is this right? Or P6 200/150 >would still outrun P5 166? A P5 will probably run F1GP better than a P6 (assuming the same clock speed). GP2 is alledgedly going to be a 32-bit program (using the Watcom DOS4/GW extender - this doesn't instill confidence in me but that's another story...) so the P6 may well run it better than a P5 (again, at the same clock speed). NT will never be as good at running games as 95; because NT scheduling actually works correctly older games (especially non-DPMI ones) don't run at all well under it. NT 3.5 renders F1GP almost unplayable - I guess we won't know about GP2 until it's released (but if it does use DOS4GW it might be okay). You might also find that NT doesn't give you any sound. BTW, NT's resource requirements make '95 look positivevly frugal. Don't even think about running it in less than 16 meg and even 24 is tight. 32 is okay and of course more is always better. 16 is plenty for '95 and 24 is heavenly (unless you use Photoshop :-) . My, I have gone on a bit haven't I! --

Quoted text (13 lines)

>>what i understand is that P6 runs no-good in 16 bit application, >>so i presume P5 166 would outrun P6 [in GP2], is this right? Or P6 200/150 >>would still outrun P5 166? >A P5 will probably run F1GP better than a P6 (assuming the same clock >speed). GP2 is alledgedly going to be a 32-bit program (using the >Watcom DOS4/GW extender - this doesn't instill confidence in me but >that's another story...) so the P6 may well run it better than a P5 >(again, at the same clock speed). >>Also GP2 would be Win95 compatible, does this mean >>it would also be NT compatible? I am intending to upgrand >>to NT 4.0 from Win95 as soon as it comes out.

Read from Microprose (on Compuserve) that GP2 is NOT Win'95 compatible. If you want to run it under the hood, fine, but they are not writing any drivers to support DirectX, etc.

Quoted text (14 lines)

>NT will never be as good at running games as 95; because NT scheduling >actually works correctly older games (especially non-DPMI ones) don't >run at all well under it. NT 3.5 renders F1GP almost unplayable - I >guess we won't know about GP2 until it's released (but if it does use >DOS4GW it might be okay). You might also find that NT doesn't give you >any sound. >BTW, NT's resource requirements make '95 look positivevly frugal. Don't >even think about running it in less than 16 meg and even 24 is tight. >32 is okay and of course more is always better. 16 is plenty for '95 >and 24 is heavenly (unless you use Photoshop :-) . >My, I have gone on a bit haven't I! >--

<<Read from Microprose (on Compuserve) that GP2 is NOT Win'95 compatible. If you want to run it under the hood, fine, but they are not writing any drivers to support DirectX, etc.>> I think this is one of the reasons why it's been delayed so long. I can't remember where, but I think I saw somewhere that they're trying to make it Win95 compatible, but didn't have a ready version of Win95 at the time. Either way, it may still work. I run NFS, NASCAR, ICR and ICRII directly from Win95 with no loss noticable visual loss. Nick _________________________________________________________________ #6 & #94 in the *real* quest for the Cup!!! My home away from home... http://users.aol.com/ntotoro/122895.htm _________________________________________________________________

Quoted text (12 lines)

|> >A P5 will probably run F1GP better than a P6 (assuming the same clock |> >speed). GP2 is alledgedly going to be a 32-bit program (using the |> >Watcom DOS4/GW extender - this doesn't instill confidence in me but |> >hat's another story...) so the P6 may well run it better than a P5 |> >(again, at the same clock speed). |> |> Sorry, although I'm not a programmer I know for sure that using a |> DOS-Extender doesn't |> mean a program is truly 32-bit. Think about it: if it were 32-bit then, |> why can it still run on DOS? |> Making 32-bit software and using 32-bit memory access or the like is not |> the same.

Yes, its clear you don't know what you are talking about. -- Regards. Mail Station 4313, On the Information SuperHighway, I'm the guy with a POBox 64942 hat driving 50 in the fast lane with my blinker on St Paul, MN 55164-0942 and my seatbelt*** out the door making sparks.

Sorry, although I'm not a programmer I know for sure that using a DOS-Extender doesn't mean a program is truly 32-bit. Think about it: if it were 32-bit then, why can it still run on DOS? Making 32-bit software and using 32-bit memory access or the like is not the same. It's like Win95 and OS/2 (or NT): The latter are true 32-bit systems and thus exploit a PPro, Win95 uses some 32-bit techniques and is partially 32-bit, but it's slower on a PPro 150 than on a P150. This is the same with games using extenders. If any game should run faster on a PPro I think it wouldn't be written for plain DOS. So, don't buy a PPro for games, it will be significantly slower than a Pentium of the same speed (a german games mag has tested this with various actual games liek Wing Commander 4 etc.). Buy a PPro only if you need it to run a true 32-bit operating system (like OS/2, NT, Linux, but NOT Win95). Greetings, Tjark P.S.: If you want to reply by PM, please use the address

It will RUN under Win'95, but it's not a Win95 game. It will still run more quickly under DOS than under Win95 as do all DOS games. Cheers! John _________________________________ __ _____| |_____ __ _________| |__| :| John Wallace | |__| |_________ \ :| |::| :| Team WW Racing TSW | |::| | / > :| |::| :|_________________________________| |::| | < / :|__|::|____:/ Sim Racing News \.____|::|__| \ /_______:/ \::/ http://www.dcn.ed.ac.uk/pulse/index.htm \::/ \._______\

Vallevand) schreibt:

Quoted text (10 lines)

>writes: >|> >A P5 will probably run F1GP better than a P6 (assuming the same clock >|> >speed). GP2 is alledgedly going to be a 32-bit program (using the >|> >Watcom DOS4/GW extender - this doesn't instill confidence in me but >|> >hat's another story...) so the P6 may well run it better than a P5 >|> >(again, at the same clock speed). >|> >|> Sorry, although I'm not a programmer I know for sure that using a >|> DOS-Extender doesn't >|> mean a program is truly 32-bit. Think about it: if it were 32-bit

then, >|> why can it still run on DOS? >|> Making 32-bit software and using 32-bit memory access or the like is not >|> the same. >Yes, its clear you don't know what you are talking about. Well, then prove it instead of just being negative. There is a difference between real 32-bit programs (and operating systems) and programs which utilize certain 32-bit techniques. For example, Win95 is advertised as a 32-bit-OS, but parts of the graphic interface and the mutimedia interface are still co5", Andrew Schulman and many others). Some system calls are passed through to the underlying DOS. Worse, some GDI parts are not re-entrant, so 16-bit applications can hinder the pre-emptive multitasking when entering that code. For the difference between using DOS4/GW and true 32-bit, see the post "GP2-flawed logic" from somebody else in this group. He has explained it technically better than I could. Regarding games: Some results from the PPro against Pentium test in the german games mag "PC Player 4/96": All values are Frames per Second (fps): Duke Nukem 3D Start of first mission PPro 150: 9 P 150: 19 P 166: 19 Duke Nukem 3D standing still after start PPro 150: 12 P 150: 23 P 166: 25 Duke Nukem 3D security monitor view in cinema entrance PPro 150: 9 P 150: 19 P 166: 20 Wing Commander IV flight through station Bluepoint PPro 150: 7 P 150: 7 P 166: 8 Only differences between the test systems were the mainboards and the CPU (of course...). So, if you want to buy a PPro for ***, do it. But not only would you need real 32-bit games written for the PPro, you would also need an underlying OS that can actually exploit the features of the PPro, I think. Greetings, Tjark

Quoted text (25 lines)

>Vallevand) schreibt: >Regarding games: Some results from the PPro against Pentium test in the >german games mag "PC Player 4/96": >All values are Frames per Second (fps): >Duke Nukem 3D Start of first mission >PPro 150: 9 >P 150: 19 >P 166: 19 >Duke Nukem 3D standing still after start >PPro 150: 12 >P 150: 23 >P 166: 25 >Duke Nukem 3D security monitor view in cinema entrance >PPro 150: 9 >P 150: 19 >P 166: 20 >Wing Commander IV flight through station Bluepoint >PPro 150: 7 >P 150: 7 >P 166: 8 >Only differences between the test systems were the mainboards and the CPU >(of course...).

One thing that you have to consider is the chipset that these PPro motherboards used. It's most likely that they had the Orion stepping A1 chipset with that PCI Write problem(sorry, I don't know the exact details). This problem causes the PCI video to slow to a crawl, 4MB/s or so. I would bet that the P6 that PC-Gamer used had this A1 chip set. Somebody had written a utility to change some of the PCI chipset settings on the A1 chipset that more than tripled the video performance; This was on a Gateway P6. I should be receiving my Supermicro P6SNF motherboard with PPro 200MHz CPU this wednesday and will do some testing in ICR2 and NASCAR. This motherboard has the Natoma chip set which supports Concurrent PCI. This should also provide some improvement since the video card and sound card are constantly getting attention from the CPU. =========================================================== Alex Fernandez MTS Lucent Technologies No. Andover, MA email: tel: (508)960-6510 ===========================================================

I have just purchased my Pentium Pro 200. I have tested to see that the Orion chip set is a B step. A P6 will do a very good job in ICR2 and EF2000. How good? How does it compare to a P5/166? I don't know. I'm very happy with it's performance on both sims. With the loaner Mach 64 card, I ordered a Matrox 4 Meg, I ran the FPS test under the following conditions. All cars visible in front, 2 in back. Number of cars that I hear is the default. Everything but skid marks on. Skid marks are auto. I was getting 20 FPS before the skid marks would sometimes wink out. I'll report back when I get some laps on my Matrox. It has improved the EF2000 frame rates, so ICR2 rates should also improve. Do I need to do anything weird to get the most from my Matrox? Do I need to turn off Univbe? Should I get Univbe 5.2 and use it instead of my Matrox supplied VESA driver? Larry

> Do I need to do anything weird to get the most from my Matrox? Do I >need to turn off Univbe? Should I get Univbe 5.2 and use it instead of >my Matrox supplied VESA driver? > > Larry Congratulations on your P6! I found that I could get maybe 1 or 2 more fps by using the UniVBE 5.2 driver with the Matrox 2.0 BIOS upgrade. If your a NCR fan, you won't believe the frames you get with the MGA special version on your CD-ROM. I run 30fps with all details set to auto except for skidmarks (ON) and track texture (OFF). Clark

Quoted text (16 lines)

> I have just purchased my Pentium Pro 200. I have tested to see that the >Orion chip set is a B step. > A P6 will do a very good job in ICR2 and EF2000. How good? How does it >compare to a P5/166? I don't know. I'm very happy with it's performance on >both sims. > With the loaner Mach 64 card, I ordered a Matrox 4 Meg, I ran the FPS >test under the following conditions. All cars visible in front, 2 in back. >Number of cars that I hear is the default. Everything but skid marks on. >Skid marks are auto. I was getting 20 FPS before the skid marks would >sometimes wink out. > I'll report back when I get some laps on my Matrox. It has improved >the EF2000 frame rates, so ICR2 rates should also improve. > Do I need to do anything weird to get the most from my Matrox? Do I >need to turn off Univbe? Should I get Univbe 5.2 and use it instead of >my Matrox supplied VESA driver? > Larry

That's good to hear. Now it makes sense to purchase a P6 b/c not only to you get comparable performance, but you also have it's benefits when software is actually developed to take advantage of it. Thanks for the report. Mike

Quoted text (13 lines)

> > Do I need to do anything weird to get the most from my Matrox? Do I > >need to turn off Univbe? Should I get Univbe 5.2 and use it instead of > >my Matrox supplied VESA driver? > > Larry >Congratulations on your P6! I found that I could get maybe 1 or 2 more fps >by using the UniVBE 5.2 driver with the Matrox 2.0 BIOS upgrade. If your >a NCR fan, you won't believe the frames you get with the MGA special version >on your CD-ROM. I run 30fps with all details set to auto except for skidmarks >(ON) and track texture (OFF). >Clark

I have a 200mhz PentiumPro on order, it should be here in about 2 weeks. I read in a magazine that the PPro performs about the same as a Pentium of the same megahertz rating, I hope that's true. I'll be sure to post a performance report here when the machine arrives. P.S. the machine cost me $3500US from Quantex. 32megs RAM, 2.5gig HD, 8x cd-rom drive, Ensoniq sound card, 2meg Diamond Stealth, 17" MAG 1795 monitor.

>I have a 200mhz PentiumPro on order, it should be here in about 2 weeks. I read >in a magazine that the PPro performs about the same as a Pentium of the same >megahertz rating, I hope that's true. I'll be sure to post a performance report >here when the machine arrives. Definitely let us know. I for one am VERY keen to hear how well they go in real use. Stuart -- Stuart Booth Somewhere in Godalming, England, UK