Processor limiting or graphic card limiting?

How do you determine if you processor limiting or graphic card limiting?

 

I'm not sure I understand what you're asking...

 

Depends on what you do, and how you do it. You have to give a specific case - like are you doing video editing, playing games, compiling C++ code, etc... It is a case by case thing.

 

i'm not sure exactly what you're asking but i think you are referring to the power settings (schemes) that affect a notebooks CPU, GPU, etc throttle speed.

You can download Notebook Hardware Control to determine the dynamic speed of your CPU, as well as knowing the temps of your CPU, HDDs, etc. Its Freeware thats a must have.

By default, Windows XP will set your Power Scheme to "Laptop"... which would not put your CPU at full throttle all the time, but will dynamically increase the speed when it is required, such as Video Conversion, Gaming, and other heavy multitasking situations.

Hope this helps,

-Gophn

 

He might be talking about being cpu/gpu limited in a game. What this means is that, depending on the game, either the gpu is waiting for the cpu to catch up (cpu limited) or the cpu is waiting for the gpu to catch up (gpu limited). In all cases you want to be "gpu limited" to a certain extent, because this means that the graphics card is being used to its full potential. This is all within reason, of course. When a game is cpu limited the graphics card has to wait for the cpu to send information before it can draw the next frame, therefore limiting FPS.
That is why it is important if you have a dual/quad graphics card set-up that you feed the beast with a very fast cpu.

 

These days processors are fast enough to run games. Even a low end processor is fast enough. Please don't get me wrong, i'm not saying that a Sempron or Celeron will run games at high fps. Most of the time, the graphics are the limiting factor in laptops.

 

This is partially right, for laptops anyway .

Anyway, let me add my 2 cents.

In most cases, both have to work in conjuction to get the most possible FPS in game. GPU do all what's pixel rendering and that's almost all. All what's mathematical calculations (vectors), physics, AI and I forget probably some, is put on the CPU.

That's why don't bother putting a 7900gtx on a PIII system cuz the CPU won't keep up with the GPU. The pixel rendering throuput will be much higher than the calculation that have to be made by the CPU.

I remember reading a nice article on Toms Hardware about this.

Now, in case of laptops, it may be different, as usually the GPU is left out for slower models becuase of energy consumption and heat mostly. So you'll find quick CPU with a basic GPU. In that case, the GPU will be holding a system.

In the case of Sager DTR thouhgh, this shouldn't be a concern as higher range models packs the top CPU with the top GPU.

As long as you pack some 2-3 Ghz processor with a 7xxx nvidia card, there shouldn't be a problem. Plus, if you have a recent processor and you don't get high enough FPS in a game, upgrade the GPU.

So let say you have a Pentium 4 around 3ghz, or a core duo processor (for a laptop), CPU won't be the limiting factor.

Now I don't think there is a way to measure this. 3D mark measures CPU processing power, but not sure how you could use those results.

 

I was taking high-end. In which case, high-end laptops are around the performance of a mid to high end gaming machine. The only difference being that the GPUs might have a slighly lower clock and one or two less pipes. It doesn't apply to thin-and-light-ultra-portables, but it does apply to DTR machines.

 

3DMark06 has the only worthwile CPU tests - '05's tests really did nothing that a CPU would have to do in real life. '06 does a good job of simulating next-generation games and AI.

 

I should have been more specific, only looked back just now.

Robgunn is right about what I was asking. Interesting article on:

http://www.digital-daily.com/video/processor_dependency/

What I have is the same as a sager 5760 with teh 7900xt 512 card and the 1.67 dual core so I don't think would be a problem. I was wonder how to tell the limiting part

 

not sure if you included this under CPU, but you are also waiting on fetch cycles when you read memory from ram or worse, from disk. then you may also have to wait on paging when you do memory-to-memory read/write, which is even slower. ask anyone who has studied microprocessors and their design, the FSB is the slowest part of the computer. you waste a lot of your 3.4billion cycles per second waiting for data to be ready.