Poll: In gaming, is FPS a measurement of performance of CPU & GPU, or measurement of performance of every component working together?

CPU & GPU
2 6.5%
Every component
27 87.1%
Other
2 6.5%

Status: Voting has ended

29 Total Votes

Poll Debate about definition of FPS (Frames Per Second) in gaming (Read 2025 times)

  • Avatar of reko
  • PipPipPipPipPipPip
  • Group: Member
  • Joined: Jun 11, 2002
  • Posts: 883
Verne and I were having this debate on #GamingW today, about what FPS really means. Well we obviously know it means "Frames Per Second", but what it really means beyond that. Verne insisted that FPS only measures the capacity of the CPU and GPU to produce as many frames as possible, while I insisted that it's the sum of every component working together. Verne's basis for the argument was that CPU and GPU process the frames while parts such as RAM don't really process anything, but are just used for storage. My basis was that FPS indicates that how many frames are actually displayed, and even component like RAM has direct impact on the performance since if there's too little of it swapping will occur, and if it's too slow it takes more time to transfer data with the RAM thus acting as a potential bottleneck.

So basically this boils down to question which do you think FPS is a measurement of. Overall performance of all components, or just CPU and GPU? Or maybe something else? Tell me what you think!
Last Edit: October 14, 2008, 06:30:17 am by ramirez

big thanx to dragonslayer for sig!
  • Insane teacher
  • PipPipPipPipPipPipPipPipPipPip
  • Group: Premium Member
  • Joined: Oct 8, 2002
  • Posts: 10515
oh...my...god...

this topic...its so boring...

must post to warn others...

also you're right.

also also YOU'RE BOTH WRONG ITS FIRST PERSON SHOOTER.
brian chemicals
  • Avatar of JohnnyCasil
  • Comrade!
  • PipPipPipPip
  • Group: Premium Member
  • Joined: Jan 5, 2005
  • Posts: 453
I agree with you rami, and I have seen this happen first hand at my job.  Anything and everything can affect the FPS (can we just say speed?) of a game.  Just as your example, RAM can play a major part due to swapping.  This can be made even worse if the end user is using an integrated solution that shares its VRAM with system RAM.  Hell, depending on the system something as innocuous as a sound card could potentially slow down the game due to poor drivers.
  • Avatar of Marcus
  • THE FAT ONE
  • PipPipPipPipPipPipPipPip
  • Group: Premium Member
  • Joined: Sep 28, 2002
  • Posts: 2690
Uh, frame frequency is the measure of the frequency an imaging device produces consecutive images.  30 frames per second (or more precisely 29.97) is the standard number for NTSC televisions while film uses a lower number around 24 frames.

Oh wait you guys are talking about vidya nevermind
  • Avatar of DS
  • DragonSlayer o_O
  • PipPipPipPipPipPipPipPip
  • Group: Member
  • Joined: Jul 7, 2002
  • Posts: 2668
verne is wrong as always
To Never Be Known Is The Worst Death
  • Avatar of Barack Obama
  • PipPipPipPipPipPipPipPipPipPip
  • Group: Premium Member
  • Joined: Jun 16, 2008
  • Posts: 5244
final fantasy 7 has awesome graphics
  • BAA2U
  • PipPipPipPipPipPipPip
  • Group: Premium Member
  • Joined: Aug 7, 2007
  • Posts: 1403
I can't see how it would only be CPU and GPU. Since there are so many other factors, including your drivers, your operating system, the game being played, and every piece of silicon on the motherboard that helps connect the two. Your monitor could even have a very low refresh, effectively capping your FPS while having nothing to do with your CPU or GPU.

BECAUSE of all these factors, there is no way to get a direct CPU and GPU reading of FPS, and all FPS readings will be "everything together" because the only way you're finding out the FPS is by looking at the end result.
Last Edit: October 12, 2008, 05:13:49 pm by goat
  • Avatar of reko
  • PipPipPipPipPipPip
  • Group: Member
  • Joined: Jun 11, 2002
  • Posts: 883
Yeah, there's a lot more into it than just the hardware part, but we were debating about upgrading a computer which is why I left drivers and the software in question from the equation. There are tons of other things that matter as well.. Room temperature, case air flow, component quality, et cetera. But let's keep this realistic. Anyway basically what I was trying to say is that FPS is  by no means a measurement of CPU/GPU performance like Verne thought it is (that's what we use flops for!)
Last Edit: October 12, 2008, 05:23:00 pm by ramirez

big thanx to dragonslayer for sig!
  • Avatar of Chubby Skelly
  • Got the powerup and won the game
  • PipPipPipPipPipPip
  • Group: Premium Member
  • Joined: Apr 4, 2004
  • Posts: 947
Yeah this is a pretty fucking stupid poll. The combination of all your components makes the difference; it's not just two things.
  • Avatar of Bobberticus
  • now that's what I'm talking about
  • PipPipPipPipPipPip
  • Group: Premium Member
  • Joined: Sep 20, 2003
  • Posts: 888
There's nothing really to debate here.
Verne is wrong.
fuck it all, dd is dead
  • Avatar of DS
  • DragonSlayer o_O
  • PipPipPipPipPipPipPipPip
  • Group: Member
  • Joined: Jul 7, 2002
  • Posts: 2668
wow this is like a public humiliation

unclerami helps my ass
To Never Be Known Is The Worst Death
  • Avatar of reko
  • PipPipPipPipPipPip
  • Group: Member
  • Joined: Jun 11, 2002
  • Posts: 883
ultimately it will help verne... he will know one thing more. sometimes the only thing you can do is getting more people to tell how things are before someone is willing to accept it (plus i hate not winning a debate when i know i'm right :<).

big thanx to dragonslayer for sig!
  • aye ess dee eff el cay jay ache
  • PipPipPipPipPipPipPipPipPipPip
  • Group: Premium Member
  • Joined: Jun 24, 2005
  • Posts: 5149
hi here's some reality for you: without a gpu you're not going to see ANY frames per second. the gpu is the greatest limiter to actual frames per second and thus fps only measures the capacity of the gpu directly, the rest is measured secondarily.
I USE Q'S INSTEQD OF Q'S
  • Avatar of King Arthur
  • Heavenly King
  • Group: Premium Member
  • Joined: Jan 29, 2003
  • Posts: 20
If playing back videos on the computer has shown me anything, it's all the components of the computer working together that dictates the FPS you get in the end. Agreeing with rami.
Last Edit: October 13, 2008, 05:48:42 am by King Arthur
  • Genericism is the phoenix of originality.
  • Pip
  • Group: Member
  • Joined: Apr 18, 2008
  • Posts: 190
All I know is that Oblivion runs shitloads faster now that I have 2gb of RAM rather than 512mb. 'Nuff said.
Upon arriving the attending doctor could find no abnormal physical symptoms other than extremely dilated pupils. After spending several hours terrified that his body had been possessed by a demon, that his next door neighbor was a witch, and that his furniture was threatening him, Dr. Hofmann feared he had become completely insane.
  • Avatar of Carrion Crow
  • I need to watch things die
  • PipPipPipPipPipPipPipPipPip
  • Group: Premium Member
  • Joined: Apr 5, 2006
  • Posts: 3516
All I know is that the rain in spain is mainly on the plain.

Sorry... was that irrelevant?
  • Comrade!
  • PipPipPipPip
  • Group: Member
  • Joined: Sep 1, 2003
  • Posts: 525
I voted for every component since that how it works in practice and especially games, but really its the incorrect answer. You're assuming that you're measuring FPS in a game or a poorly coded program. Games are one type of program, that are usually coded somewhat inefficiently. If your computer is fully functional and we're not talking games, then Verne is mainly correct as it'll come down to the CPU and GPU what your FPS effectively comes out as while rendering 3D. This is especially true if you're running benchmarks, where the FPS is the main indicator of your computer's performance in an intensive graphical situation.

Really the initial question is flawed. But by definition Verne is correct.
"the way you speak about rm2k3, like "modules, arrays, pointers" sounds like
you're some badass c++ programmer stuck in the past and only has rm2k3 to
work with"
-Gutts
  • Avatar of headphonics
  • sea of vodka
  • PipPipPipPipPipPipPipPipPipPip
  • Group: Member
  • Joined: Dec 24, 2003
  • Posts: 6432
i posit that your all gay
  • Avatar of reko
  • PipPipPipPipPipPip
  • Group: Member
  • Joined: Jun 11, 2002
  • Posts: 883
I voted for every component since that how it works in practice and especially games, but really its the incorrect answer. You're assuming that you're measuring FPS in a game or a poorly coded program. Games are one type of program, that are usually coded somewhat inefficiently. If your computer is fully functional and we're not talking games, then Verne is mainly correct as it'll come down to the CPU and GPU what your FPS effectively comes out as while rendering 3D. This is especially true if you're running benchmarks, where the FPS is the main indicator of your computer's performance in an intensive graphical situation.

Really the initial question is flawed. But by definition Verne is correct.
Uh, no. Your computer can be fully functional with say, 256 megs of RAM, but if game requires a gig to run well, that doesn't mean it was coded inefficiently. It means that you have an insufficient setup, and that directly affects the performance. So no, I'm not assuming that you're measuring FPS of a poorly coded game. The only thing that Verne is right about is that CPU and GPU are the biggest factors. You can't get a huge increase by upgrading RAM if you already have a sufficient amount of it and at sufficient speed, but you can by upgrading CPU or GPU. That's all.

And like said, there are tons more factors that come to play, from heat production to your OS and drivers, that are independent of how the game is coded originally. The game can't help if the user got a shitty drivers (yet still functional, just poor), yet that can still result in noticeable drop in performance.
Last Edit: October 14, 2008, 05:09:20 am by ramirez

big thanx to dragonslayer for sig!
  • Comrade!
  • PipPipPipPip
  • Group: Member
  • Joined: Sep 1, 2003
  • Posts: 525
Thanks... for expanding the first statement in my post? I said every component affects games, but the question was not about games, it was about FPS. Game performance != FPS.

I agree that RAM matters ofcourse, its just kind of silly to state it like that. Your FPS is not an indicator of how much ram you have, unless your rating system is from "not enough" to "enough".
Last Edit: October 14, 2008, 05:47:07 am by Vanit
"the way you speak about rm2k3, like "modules, arrays, pointers" sounds like
you're some badass c++ programmer stuck in the past and only has rm2k3 to
work with"
-Gutts
Locked