Sign in to follow this  
Guest

P4 1.5 vs. P4 3.06 "disappointing results" - Pleas

Recommended Posts

thank you for posting this. im thinking of giving up on my old PII and buying a custom made CPU from dell. i had the same choice (P4 3.06) but im going with the 9700 pro card instead of the GF4. i thought about it but reading the reviews and by going over the posts here like this one i pretty much decided ATI was the way to go untill the price drops on the GFX. i do believe GFX will hold the number 1 spot on the market for awhile but dont forget ATI will be coming out with theyre next card not too long after. so who knows if itll beat the GFX or not. if you have looked at the trend lately, everytime NVIDIA comes out with a card ATI has been right in back of it with a better one. so might just be worth the wait.

Share this post


Link to post
Share on other sites
Help AVSIM continue to serve you!
Please donate today!

>i think im just going to build it myself! isnt rocket science like i thought it was!.. woohoo! :-lol Very true.I just built my first system (whoose specs are in my sig) this past summer, and was surprised at how easy it was. My motherboard came with a very comprehensive manual, and that diffenitly made the experience that much easier.Think of building a computer like putting together an advanced set of legos. ;-)Ryan-Flightpro08 :-cool VATSIM Pilot/ControllerZLA ARTCC Senior Controller (C-3)ASRC (Advanced Simulated Radar Client) Beta Tester-----------------------------My "Home Made" System Specs:Intel Pentium 4 2.2GHz ProcessorTurbo Gamer ATX Mid-Tower with 420W Power SupplyEPoX 4G4A Motherboard with Intel 845G ChipsetVisiontek XTASY GeForce4 128MB Ti4600 (Det 42.01 Drivers)512MB PC2100 DDR RAM40GB Matrox 7200RPM Hard DriveWindows XP Home Edition SP1*No CPU or GPU Overclocking*3dMark2001SE Score: 11298-----------------------------Click Here to Download my American Eagle POSKY CRJ-200!

Share this post


Link to post
Share on other sites

My personal viewpoint in these situations is fairly similar to Paul's, in that the CPU is a principal player in sim speed, but not to the total disregard of memory and graphics speed. I also tend to believe that as CPUs get faster, the improvements become exponentially smaller as well, and graphics card speed becomes more important. I actually tabulated the benchmark database here at Avsim a couple of months ago. After analysis of the top 100 results, it was clear that the gains from CPU speed improvements continue to diminish as CPU speeds increase. Unfortunately there does not exist sufficient data for over 2 ghz machines (and by that I mean at least 100 separate results for 2ghz machines or faster) to draw definite conclusions about how the graphics card influences performance at high clock speeds. I guess Paul will disagree with me here :-) but I prefer to see a large variety of results from many users rather than just relying on the analysis of a couple of people, no matter who scientific or thorough they are :-)In the case of comparing your old 1.5 ghz machine to your new 3.06 ghz machine, I would firstly make sure you are using the same settings for both when making the comparison. A lot of people complain there is little difference in performance between old and new machines until it is revealed that a few sliders were moved or checkboxes checked, or add-ons installed since placing the new machine into service. I would suggest (as an experiment only), to disable vsync, run the sim without any anti-aliasing or any anisotropic filtering at 640 x 480 x 16 resolution, and then compare the results using these settings on both machines. Just make sure the settings in the sim are identical and that AI traffic is disabled. Also make sure the test is precisely the same, with precisely the same aircraft in precisely the same place at precisely the same time of day with precisely the same weather. A pre-recorded video is obviously useful in this respect, but not neccessarily indicative of sim performance (simply because it IS a pre-recorded video rather than a real time flight). If you find the sim runs much better on the new machine in this situation, then a graphics card upgrade is most probably your answer. If you find little difference, then you have cause for complaint with the vendor, as something is defintely awry with the new machine.

Share this post


Link to post
Share on other sites

Hasnt anyone figured it out yet...INTEL SUCKS!!! Any AMD can kick the Pants of an intel. Get an AMD ATHLON XP 2800, that will fix everything

Share this post


Link to post
Share on other sites

>Hasnt anyone figured it out yet... >>INTEL SUCKS!!! Any AMD can kick the Pants of an intel. Get >an AMD ATHLON XP 2800, that will fix everything An extremely well-informed, well based and technically sound contribution to the discussion. And your fully tabulated, double blind test results performed on multiple platforms and hardware combinations would be where exactly??? :-roll

Share this post


Link to post
Share on other sites

>Thank you Michael. I said the sim is CPU dependent, ram >dependent and GPU dependent. I have also read over and over >and over of the success stories of members who have upgraded >their GPU and discovered substantial performance increases >using the same CPU. It all depends on where the bottleneck >is. Doesn't mean squat if its not apples to apples, you can also read about amazing performance increases with every new driver and version of DX that comes along :-roll, doesn't make it so.>Paul; Perhaps you should learn how to read what is writtenI quoted you directly MG, here:"it's also GPU dependent and ram dependent. Just like all games are. Some more than others. FS2k2 certainly is no exception."Uh, just like all games are?Sorry, you know not of what you speak. "What you're reporting is no surprise to me."Funny though, it is a surprise to most everyone on this board, including those that build gaming rigs for a living. :-roll>rather than just dismiss anything that tends to contradict >your CPU only theory. Since the first Fs2k2 beta came out more than 18 months ago I have always said that Fs2k2 performance was about 4 to 1 in favor of the CPU, when Fs2k2 final was released with the help of whole team of developers I was able to pin the performance down to:55-60%CPU 15-20%memory-subsystem 10-15%video card, the rest is variable to BIOS&settings,OS&settings, drivers, sound cards and way to much more to pin down. This was accurate enough to nail what kind of performance increase could be had in Fs2k2 with the then to be released GF4TI just based on its specs alone because of its similar architecture to the GF3.all of those post still exist here in the archive if you ever want to challenge it.>While a good CPU is important with >FS2k2, it ain't nothing without the rest of the system.so True, we can agree on that. >I've been flight simming since '98 with four different >systems. I think I've figured out what's important and >what's not. and every year you get closer to figuring it out! :-lol, Just kidding MG! BTW; they haven't made a CPU yet that can run >FS2k2 in all its glory with demanding aircraft and 100 >percent AI. Perhaps a 5.0 gig machine will show us the >light. well now, actually there are some extreme P4's running on air up in the 3.4GHz range using the 9700P and even while in the CS727 VC at LAX with full PAI-atc on stayed up over the 18 FPS mark and that is really exceptional, I can do 16+fps AAx2Q in the CS727 VC at most airfields with 100%PIA-atc but not at places like LAX.So, yeah I'll take a 5GHz machine! :)

Share this post


Link to post
Share on other sites

>INTEL SUCKS!!! Any AMD can kick the Pants of an intel. Get an AMD ATHLON XP 2800, that will fix everything Very constructive comment. :-rollWhy does every, mature, very imformative thread about computer hardware have to have somebody come along and post something like that? Ryan-Flightpro08 :-cool VATSIM Pilot/ControllerZLA ARTCC Senior Controller (C-3)ASRC (Advanced Simulated Radar Client) Beta Tester-----------------------------My "Home Made" System Specs:Intel Pentium 4 2.2GHz ProcessorTurbo Gamer ATX Mid-Tower with 420W Power SupplyEPoX 4G4A Motherboard with Intel 845G ChipsetVisiontek XTASY GeForce4 128MB Ti4600 (Det 42.01 Drivers)512MB PC2100 DDR RAM40GB Matrox 7200RPM Hard DriveWindows XP Home Edition SP1*No CPU or GPU Overclocking*3dMark2001SE Score: 11298-----------------------------Click Here to Download my American Eagle POSKY CRJ-200!

Share this post


Link to post
Share on other sites

John,I would agree 90% of what you said (well based on word count it's closer to 96% :-lol)My caculations are based on many systems as well as other users who know FS and PC's Builders and developers, many personaly setup to test Fs2k2 by me, all kinds, if I had to count, it would be close to fifty for FS2k2 alone in the past year or so(about thirty soon after the release of Fs2k2.As for the "CPU speed improvements continue to diminish as CPU speeds increase" that is as you point out hard to nail as it would seem that way... and then a new MB-chipset comes out that changes the results for the same given CPU, take for example the latest P4-b MBs as well as the new GforceII boards I have personally seen increases using the exact same CPU that are up near 10-15% better as a result, still the fact does remain that there is a diminishing return if you look purely at "well here is a 300mhz gain" yeah, but now it is only 15% increase for a 1.7Ghz to a 2Ghz, comparing to a PIII@600 to 900 was almost a 35% increase, well I think you know what I mean..So if you still look at a 10% increase in CPU cycles in the same box you can still find a 5-6% increase in raw unfiltered power for FS2k2.Make sense?

Share this post


Link to post
Share on other sites

PaulLO1 is right on the money!!! When I upgraded to a new Box P4-1.5 to an AMDXP2100 the performance gains were very nice. Whether I ran FS2002 with a GF3 or GF4 vid-card did not really matter when looking at FPS. The CPU horsepower did....when I built my brother his P4 2.53 533 FSB system it just screamed with FS2002. The evidence is cold and hard check the threads in this forum. What is up with your system I do not know. But you can always send me the chip!!!!!Tony Ascaso, RN

Share this post


Link to post
Share on other sites

>Hasnt anyone figured it out yet... >>INTEL SUCKS!!! Any AMD can kick the Pants of an intel. Get >an AMD ATHLON XP 2800, that will fix everything :-lol :-lol :-lolwell if anything Sucks, its the VIA chipsets certainly not Intel!Get real man! :-lol

Share this post


Link to post
Share on other sites

hi Jon,well, first of all i can assure you that the settings Both in the game and in the video card control panel was the same in Both systems during the testing. by checking and double checking and, not to mention the fact that i pretty much from day one (talking about the old system here) ran fs2k2 with All sliders maxed except visibility - set at 60 miles and the water slider set to "none". this is another reason why i chose the settings for these two sliders as well as the other settings for the rest of the sliders and the video card settings as well for the tests. i was very familiar with the fps i was getting with that system and those settings. to me it was a fair comparison between the two. also, for me the trade off was for better visuals over more performance. especially when fs2k2 ran much smoother at much lower fps than fs2000 ever could. my thinking for the testing was...we all prefer to run fs with better "visuals" and to do that means stressing the video card as well as the CPU. also, if your going to try and get a pretty good idea of what your system is made of especially when it comes to gaming, why not stress as much of the system as possible rather than just part(s) of the system. thanks for the info. happy flyin, fm

Share this post


Link to post
Share on other sites

hmm well iam not surpised at the FPS level here. When cranking up everything to max in Fs2002 it will kill the strongest of systems and this proves it. If anything you should notice a great deal of improvemnts with just turning AI down to 50% .Fs2k2 AI can't handle alot planes well anyways.I have learned to just settle for a nice medium between good looking graphics and playablity. Now I only have a 2.1 ghz but I can get a between 17-25 fps in most cases and with weather unless it's a big set of gray clouds which drops the performance to a slide show due to bugs .I would do the followingFirst get the 41.09 drivers Get nvrefresh or refreshlock to fix 60hz bug in winxp and win2kturn off all services you are not using to free up memory. most of the stuff that is turned on in 2k and xp is not need by default . now turn on fs2k2bring the AI to 50%bring the clouds to 65%bring visabilty to 90 miles turn aircraft shadows off FPS locked at 30now before going and cranking up the resolution and FSAA etc just see what the performance is like at 1024x768 at 32 . You should be able to get a solid 30 at megis with the default cessna with that set up.Now you can start to tweak things to a better level. I run fs2k2 at 1152x864 32 bit color refresh rate 100hz FSAA 4x+9 tap AA 2x I can get 25 in most cases although with heavy AI and storms it can go to 10-15 as stated above. I would thinking you should be able to get a solid 20 in all cases as long as certain aspects of FS2k2 state within a usable range.AI being the major one.If you use project ai only you will see great improvments .I do not use FS2k2 AI default planes at all.This is what works best for me :) Capt.Richard Dillon (KATL)www.jetstarairlines.comhttp://hifi.avsim.net/activesky/images/wxrebeta.jpghttp://jdtllc.com/images/RCsupporter.jpg"Lets Roll" 9/11 -----------------------Specs AMD 2400 XP MSI KTV4 512MB DDR 2100Asus GF4 ti 4200 128MBSB Audigy Gamer Ch Products Yoke and Pedals(usb)Windows 2000 Serivce Pack 3

Share this post


Link to post
Share on other sites

try setting or unsetting "T&L acceleration" in the hardware options, and see if it makes a difference.regards,

Share this post


Link to post
Share on other sites

Just do your own experiment. Go to BestBuy or Circuit City and purchase a 9700 and try it out. You can always return it with no problem. Robb

Share this post


Link to post
Share on other sites

Paul01,I wonder what you think of the recently uploaded FS2002 benchmark test. The benchmark is contained in the file#24975, FS2002 Utilities, FS2002 benchmark v.1.2 by Maurizio Losso.I'd personally like to see a thread started with the results of this test posted, if the test is valid and repeatable. Paul, my experience is that FS2002 performance is very dependent on CPU, just as you said. (Having just spent 2200$ for a Asus P4PE 2.4 GHz, ti4600 128 MB, 1GB DDR, from a 1 GHz P3 384 SDRAM G2GTS 64 MB, the somewhat less than satisfying experience is fresh on my mind! {;-))Larry JonesFlorence, MT

Share this post


Link to post
Share on other sites

Yes, it does make sense (you always make sense :-)) I guess the key is that the figures need to be kept a little rubbery to allow some lattitude here and there. I mean I'm sure the introduction of the Barton, for example, is going to boost figures # 1 and # 2, and make figure # 3 (the GPU) a little less critical, at least for that archtitecture. I certainly agree with chipsets throwing spanners in the works as well. My feeling is that future versions of Geforce FX technology (say by the end of this year) and ATi technology around the same time will well and truly put the emphasis back upon the chipset/cpu/memory trio. I guess because FS2002 really only starts to look good once you hit resolution, AA and anisotropic filtering reasonably hard, current generation cards have their work cut out a little (much less so the 9700 Pro though). Even in my own case, my only performance complaint is the effect of 4x AA / AF or higher on frame rate. I'm sure these concerns won't even be an issue at all with any state of the art GPU available in the near future. Certainly in my own case, I am more than happy with the horsepower of the Athlon XP 2400 + to run FS2002. Where I would like to improve my FS2002 experience is by running higher resolution and higher level AA/AF. So in my particular case I only need concern myself with a new GPU.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this