Do CPUs effect 3DMark scores ?

Like if I took two of the same GPUs but put one with a U or Y CPU and put the other in a quad core, would the scores be significantly different ?

 

I don't know if significant means the same to you and me; but to me: yes they would be.

The CPU + RAM (and the O/S and supporting chipset, of course) is what determines the level of a 'platform'.

The gpu, storage sub-system, and any other 'add on' is highly dependent on the basic 'platform' though.

If you notice in almost all benchmarking sites: they are running the highest level platform possible. Anything else will not be able to resolve the differences of components like gpu's as accurately nor as easily...


Also keep in mind that mobile platforms and desktop platforms (and their respective components) are not equal either... the desktop platforms have no power/battery life restrictions - much less heat restrictions - MUCH better cooling designs - and mostly no need to have 'sleep states' in use that aggressively destroy performance (even during high system loads...).


Another way to look at the above (in relationship to your question) is that even if the 'scores' were very similar (numerically/%age-wise) - that does not mean that a specific gaming title would perform the same on the two hypothetical systems you indicate... as with most things in life the answer is 'it depends' (on the specifics of the cpu/gpu and the game (and quality level) in question).


Hope this helps.

 

Yes. Since there is a CPU component (the physics test) which factors into the overall score you would score quite a bit higher with, say, a quad-core than with a dual-core at the same speeds. For example, the i7-3630QM has about twice the physics score of the i5-3230M in 3DMark 11.

 

It did thanks !

Problem for me was I was looking at a review of one of the Clevo system with the 4750HQ + Iris Pro and it had a comparison with a Asus with a Nvidia 740m GK208 but it had a duel core U CPU and it looked like the Iris beat the 740 by about a 22% average and it got me thinking if I should shell out 200$+ more bucks for that system because the 740m is the same in my quad-core Envy .

 

I see .

Crap forgot to ask, but does that include FPS in games too ?

Like if I had a U CPU vs. a Quad-Core CPU but the same graphics .

 

Depends on if the game is CPU-bound or not, but for something extremely CPU-hungry like BF3 multiplayer you would definitely see an increase in performance if the game was not completely GPU-bound in the first place, considering that the standard-power mobile quad-cores have not only twice the number of cores and amount of cache as the ULV variants but higher clock speeds to boot. With that being said, if we're talking gaming notebooks with high-end GTX-level GPU's then OEM's typically supplement that with at least the entry-level quad-core of that CPU generation. You would never, for instance, see a GTX 780M paired with some ultra-low power dual-core typically found in Ultrabooks. There have been unusual configurations, such as the MSI GX60 and GX70 with a 7970M/8970M paired with an AMD A10 APU, but these are few and far in-between and, in the case of the MSI, the severe lack of CPU power relative to GPU power to cut costs ends up unbalancing the entire system performance and substantially bottlenecking the 7970M/8970M in games.

 

I don't think the CPU will do much in this case. You say it was a GK208 740m? So best case, you get a chip that has a 64 bit bus to the GDDR5 memory. Worst case, you get the same chip but with GDDR3 memory instead. Either ways, you are looking at similar performance if you compare the 740m to the 5200. Even if that 740m had a quad core, there's a good chance it would have only been 10% better, if that.

Like octiceps said, you would be hard pressed to find a high end GPU in a system with an i5. There are a few of these in the mid tier section where you could have a 750m with an i5 (although Haswell i5s aren't out yet) and a 750m with a quad. In your case though, I doubt it would be worth it to pay $200 more and purchase the 740m. Like tiller said, the whole system is what matters. You aren't going to see a big difference in games. There could be a couple of games optimized for nVidia drivers/architectures that would run better on the 740m but the majority won't beat the Iris Pro by enough to warrant the $200. However, most of your other tasks will benefit significantly with the better CPU.

 

I see, thanks !