Ram as a bottleneck?

I always figured that as long as I had 6+GB of ram that there would be no reason for it to bottleneck in games or benchmarks. However, it was recently suggested to me that the ram speed could be a bottlenecking feature (such as 1333ddr3 performing much worse than 1866ddr3). Also I was wondering about desktop ram vs laptop ram (1066 ddr3 desktop vs 1866 ddr3 laptop).

This may play a factor in which laptop i purchase in the future.

 

I would ask the person who suggested this to you to detail, very carefully and specifically, exactly what they mean. Then post their explanation here.

I always like to start the week with a good laugh.

 

In case you missed newsposter's point... RAM speeds will never bottleneck you in gaming, so it should not factor into your laptop choice at all

 

Well it started out as a discussion of the Clevo x7200 vs the M18x. I've always been partial to Sager and he's an Alienware guy.

He pointed out that some of the M18x's were getting higher 3d11 scores with 6970's in CF than the x7200 with the i990 and 6990's in CF. His reasoning was that this was due to the higher ram clock speed on the M18x.

 

That's a benchmark, in benchmarks memory speeds do make a difference. Benchmarks don't mean much in the real world though. In actual games I'd be impressed if you could ever tell the difference.

 

Alright, that's what I figured.

I have another question then, don't feel like making an entirely new topic.

the i7 2920xm (mobile cpu) vs i7-960 (desktop cpu)? I would assume that since the 960 is stock clocked higher and a desktop cpu that it would at least slightly out perform the 2920xm, however, I'm not sure hence me asking the question.

 

Apparently the notebook CPU shoud perform a little better. I wouldn't count on you noticing much difference though, assuming the test scores scale linearly then it's only about 15% faster, not exactly ground breaking stuff. You shouldn't be disappointed with it though.

PassMark Intel vs AMD CPU Benchmarks - High End

 

If the root of your question is do desktop components perform better than their laptop counterparts, then the answer would be a resounding yes.

As a general rule, an equivalent laptop CPU would perform one step below it's desktop counterpart for all the obvious reasons.

 

Actually RAM speeds CAN bottleneck you in gaming. They jus tdon't often.

When Dolphin, the Wii emulator, is set to use a RAMCache (as opposed to using VRAM) the timings and speed will effect your overall FPS though not greatly and in the test I saw they had full FPS one way or another.

 

Sandy Bridge has narrowed the gap between desktop and laptop CPUs. Also, the 960 is a revision of a processor from 2008 with a minor increase in clock speed. The 2920XM will win.

Regarding RAM as a bottleneck: this is only the case with the new "Llano" processors. They stuff a fairly decent GPU together with the CPU on the same die, but it is forced to use system memory. Thus, you have the classic case of a bandwidth-constrained GPU and faster RAM helps. With other processors, it does not matter.

 

Well everyone pretty much covered it about the RAM not being a bottleneck. As for the 960 vs 2920xm, Sandy Bridge is faster at the same clock speeds so the two should be really close. You can also easily OC the hell out of your 2920xm with the right chipset if you feel like it.

 

Thanks, what about one of the slower mobile processors vs the 960. I heard the 2920xm is the only overclockable mobile cpu, and I don't feel like dropping an extra $900 for it, and that the 960 is overclockable.

 

I think it was mentioned that a 2720qm (SB) would be able to outperform the XM of the previous generation (or was it a desktop one?).

Either way, getting a cpu above 2720qm is not really practical because the prices do NOT justify the (ludicrous) costs, and furthermore, they are overcharging for minor increases in speed (100 to 200Mhz per core).