9x333, 8x375, or 7x428 on a Q6600 - Which is faster?

[elec999]

Very Nice results. I hope to get a q6600 soon.
Thanks

 

[tranceaddict]

Nice tests, thanks for the results. I have been wondering if there is any benefit in dropping the multiplier to get a higher FSB. By the looks of things, it really does not make too much difference at all.

 

[shirker]

Nice tests, thanks for the results. I have been wondering if there is any benefit in dropping the multiplier to get a higher FSB. By the looks of things, it really does not make too much difference at all.

granted, as with all cases, your mileage will vary. whatever you use your computer might benefit greatly from a higher fsb, or not at all. fortunately, graysky's "benchmarks" were varying enough to cover a few areas of usage

 

[graysky]

Yeah, I'm open to others posting a similar analysis on their own system. Maybe something w/ some games as benchmarks might be interesting particularly since most games have timedemos that can report a fps... anyone up to it?

 

[srtdodge05]

graysky where you been? ive been trying to get intouch with you over at ABX. I just got a Q6600 and im trying to overclock it but not having much luck. Right now im trying to get it stable at 3.0ghz but having trouble finding the right voltage. Stock VID is 1.325v. I tried to lower it but it always crashes anything under 1.3v. I was able to get up to 356fsb but with 1.4V i was hitting 75C.

 

[graysky]

I'm here What are your hardware details? What are your current voltage settings. You might wanna start a new thread so others can post to it as well. Be sure to PM me w/ the url to that thread since it's not really appropriate to post it in this thread.

Lemme know.

 

[graysky]

I just read the FSB1333 Intel Processors & New 2007 CPU Charts article over at TH.com and am happy to see that the testers over there have drawn the same conclusion that I have about fixed final core speeds with higher and higher FSB speeds: faster FSB speeds w/ a C2Q/C2D don't equate to faster real-world benchmarks.

Have a look at page 8 from their article comparing the "old" 1066 MHz FSB to the "new" 1333 MHz FSB chips: average gain <1 %.

 

[jason4207]

There is no E6650 because the processors use only even multipliers, which leaves no room between 2.33 and 2.66 GHz.

This quote from the article left me wondering if the author knows what they are talking about. The e6550, e6750, and e6850 use 7x, 8x, and 9x multis, respectively, as far as I know.

We expect these new CPUs to support significantly higher core clock speeds, which results in higher bandwidth requirements that FSB800 or FSB1066 probably cannot support - especially with quad core processors in mind.

I'd like to see this test re-run w/ a new g0 Q6600 @ 4GHz, and see if a bottleneck can be found.

 

[orion456]

I'd like to see this test re-run w/ a new g0 Q6600 @ 4GHz, and see if a bottleneck can be found.

IF the processor is getting data as fast as it can distribute it to internal channels [ 1:1 ] then the only advantage of a faster bus speed is when memory to memory transfers are made without the help of the CPU. But most memory transfers are from/to much slower devices like a hard drive, so fast memory has no real life application until the CPU has maybe 8 cores that need data faster than the present bus can deliver it AND if you can find programs that can efficiently use so many cores.

I'm sure that the g0 Q6600 will show the bottle neck to be itself most of the time. I believe analysis has shown that you need a multiplier around 13x before the cores begin to exceed the ability of the bus to supply data at [ 1:1 ]. The current 10x multiplier for the Q6700 is still too low.