What is bottlenecking?

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.

symbi0nt

Distinguished
Mar 15, 2006
220
0
18,680
I think the apples to apples section of that review is more critical, since most gamers use 1280x1024. There it doesn't look like the Conroe has a huge lead, and even that lead would lessen over time as GPUs get better and more graphical features are added?

To me, that would ease my mind about buying an AMD for a gaming build. [Before I think about the prices, anyway!]
 

ikjadoon

Distinguished
Feb 25, 2006
1,983
44
19,810
OK, Conroe has 1-3 frames better than the FX-62. If that is what you call a substantial lead, sure. I quote

Oblivion is the only game in our testing suite that we found had any kind of gaming experience differentiation between the Intel’s Core 2 and AMD's FX-62. Even then, it was only a small difference.

Linkage

If you have an X2 right now, it is fine. If you have something older and want a new system, go with Core 2 Duo.

~Ibrahim~
 

ikjadoon

Distinguished
Feb 25, 2006
1,983
44
19,810
You are right. But that is JUST gaming. Core is faster than the current X2's in other things, such as multimedia and the sort. But right now, buying a Conroe system over a X2 system just for games is not a nessecarily a good thing or a bad thing. The difference is small in gaming, as would expect from the heavy-GPU usage in games. So if you ONLY game, it won't matter much.

But who JUST games? There are so many other things that the CPU does and Core 2 Duo seems to be faster. Currently, I'm waiting for Hard OCP's second edition, which benchmarks Core's performance in everything, but games....

~Ibrahim~
 

xsandman

Distinguished
May 26, 2006
232
0
18,680
There it doesn't look like the Conroe has a huge lead, and even that lead would lessen over time as GPUs get better and more graphical features are added?

NO!!!!! This is COMPLETELY WRONG!!!

The exact OPPOSITE is true! As GPUs get better, the lead Conroe has would INCREASE!! If you use a better GPU, then the GPU is less of a bottleneck, and the game becomes more CPU dependant. Does this make sense?
 

RichPLS

Champion
There it doesn't look like the Conroe has a huge lead, and even that lead would lessen over time as GPUs get better and more graphical features are added?

NO!!!!! This is COMPLETELY WRONG!!!

The exact OPPOSITE is true! As GPUs get better, the lead Conroe has would INCREASE!! If you use a better GPU, then the GPU is less of a bottleneck, and the game becomes more CPU dependant. Does this make sense?

Xsandman is correct... :trophy:
 

RichPLS

Champion
My point is that everyone does not have that setup and play at that res., and even so, might not want to disable detail quality to make game operational, as in that review...

But if you get better graphics, the Conroe lead gets much greater also.
And if your graphic card is lower end, then a fast CPU is less important.

After a certain point, 2GHz or so and at higher resolutions, the bottleneck is the GPU... But as you invest in higher performance graphic cards, or dual graphic cards then the faster CPU is needed to increase the FPS.

And Conroe boots and loads much faster than AMD too, which actual user experience is tied to general quickness of the PC, which is hard to benchmark but is evidenced by Conroe users statements...

Strangest thing is, if this was AMD with this killer chip, NO ONE WOULD BE STATING THAT EXTRA CPU POWER JUST DOES NOT MAKE A DIFFERENCE!!! :roll:
We all know there are other bottlenecks that affect performance, it is just humorous how the beaten ones feel they need whole articles dedicated to dispell Conroe's performance by selecting a system and resolution, and ignoring the other resolutions as a reference... that is biased.
 

ikjadoon

Distinguished
Feb 25, 2006
1,983
44
19,810
We all know there are other bottlenecks that affect performance, it is just humorous how the beaten ones feel they need whole articles dedicated to dispell Conroe's performance by selecting a system and resolution, and ignoring the other resolutions as a reference... that is biased.

Are you referring to me? I am not ignoring any article run at low resolutions, just pointing out that they are not realistic in real performance, though they are pretty good at noting CPU performance, which is in favor of Conroe.

As I mentioned, Conroe has a very slight lead in gaming, but that is just gaming. That slight lead is the CPU running faster than the X2's. That increase can only be attributed to the CPU because it is the same GPU. Xsandman is correct, as GPU's mature, Conroe-based systems will start to show faster FPS, which is the way you want. It is going to benefit plenty, probably starting with G80 and the R600 who debut later this year Looking at these benchmarks, who are using CPU-intensive, we can see Conroe with a large lead:

Core 2 Duo Multimedia Performance

This shows some raw processing power, in which C2D is in the lead.

Most benchmarkers will now have to resort to Core 2 Duo systems over their AMD counter-parts to release any bottlenecks.

~Ibrahim~

P.S. There is a good chance I'll be getting a Core system.
 

RichPLS

Champion
But contrary to HOCP's review, Conroe does indeed have a large lead in gaming if you unconstrict the GPU bottleneck by using next gen faster video cards or current xFire enabled dual X1900XT's and test at 1280x1024 or even 1600x1200 or 1920x1200...

that article just omits these facts...
 

xsandman

Distinguished
May 26, 2006
232
0
18,680
But contrary to HOCP's review, Conroe does indeed have a large lead in gaming if you unconstrict the GPU bottleneck by using next gen faster video cards or current xFire enabled dual X1900XT's and test at 1280x1024 or even 1600x1200 or 1920x1200...

that article just omits these facts...

You are both correct. It sounds like you are both arguing the same conclusion. You are both aware of what HOCP does with their benchmarks and that it does show a GPU bottleneck, and that as the GPU gets better, the difference of the CPU will come out.

It is just certain people (of which do not include u 2) that use the HardOCP data to show the CPU has NO IMPACT at all in real world performance, which is incorrect.
 

ikjadoon

Distinguished
Feb 25, 2006
1,983
44
19,810
It is just certain people (of which do not include u 2) that use the HardOCP data to show the CPU has NO IMPACT at all in real world performance, which is incorrect.

Exactly. That is absurd to think that a CPU will not change real world performance. I should know, I use 900Mhz and below computers every day. Coming home to a P4 2.8Ghz is bliss, let me tell you.

~Ibrahim~