• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Does overclocking really replace a $1000 dollar chip?

K

known12345

Junior Member
Ofcourse a extreme edition processor will overclock much higher then a q6600 but can the q6600 ever surpass a stock qx9650? I mean lets says we have two comp buyers; one who knows nothing about comps and another who has overclocked many computers. If the one who knew nothing went out and purchased a $1000 (in this case a qx9650), and the one who knew alot about comps purchased a q6600 for around $300 and overclocked it, will his overclocked q6600 ever surpass the speed of the cpu (qx9650)of the person who knew nothing about comps?

I know that overclocking a q6600 (or any other chip for that matter) will make it run much faster but does it run fast enough to replace a $1000 dollar chip? I mean lets say we clock a q6600 at 3.0 ghz with an L2 cache of 8 and we compare its speed (in doing applications) to that of a qx9650 at 3.0 ghz and an L2 cache of 12, does the q6600 perform similar to the qx9650? I mean ofcourse the qx9650 will be faster then the q6600 because it has a higher L2 cache but what if the q6600 was clocked at 3.2 or 3.4 ghz, will it then run faster then a qx9650? If it doesn't run faster, how high would you have to overclock a q6600 to surpass a qx9650 at stock settings in terms of speed?

If this is a noobie question, I am sorry for asking. I mean I know that comparing cpu isn't only about comparing clock speeds but comparing L2 cache as well. I read an article over at tomshardware that talked about the importance of Cache size but they tended to focus on older chips with less cache size. Quads these days, on the other hand, use much larger L2 caches and I am wondering if the 4 cache difference between the q6600 and the qx9650 makes a significant difference between the two in terms of speed if they have similar clock speeds. I mean will the 4 L2 cache difference even make a qx9650 at 3.0 ghz surpass a q6600 overclocked to 3.6 ghz?
 
Well, a Q6600 @ 3.2GHz will match a stock QX9650. Since most Q6600s can be overclocked to around 3.6GHz, I'd say yes, it has the potential to surpass a stock QX9650, easily.

Of course if you overclock the QX9650 to 4GHz+ it's not much of a contest. 😛

 
It will not *exactly* match a higher end processor (cache and multiplier differences), but overclocking a lower-end chip will yield approximately the same performance as a higher-end chip, barring cache differences. For example, an E6420 (at stock, a 2.13ghz chip) will match an E6850 (at stock, a 3ghz chip) once overclocked - both caches are equal; the only difference is FSB speed and multiplier, which make hardly any difference.
 
There are some architecture differences between the 65nm and 45nm Core2 chips. A 3.0Ghz 45nm chip would be slightly faster than a 65nm chip even with the same cache size. That said, you should be able to overclock it to be faster, probably.
 
If you were comparing the Q6600 to the QX6850, Id say yes the Q6600 can easily surpass the 1000$ processor. These processors are essentially the same processor, just binned differently.

The QX9650 however is a different process, 45nm and 12MB cache. They cannot be compared as much. If you look at the Q9450 and Q9550 that will be out in a month or so, you can compare those to the QX9650 more easily.

What I'm getting to, to access the maximum potential of a Quad core chip, you have to look at its multiplier. All (or at least 99% of them) quad core chip will run out of steam at 475MHz FSB. The Extreme parts will give you the unlocked multiplier and thats what you will be the extra 600$. Will you need more than 3.2GHz of quad core power?

As for the importance of cache... I would tell you this. You will not see a significan difference between 8MB and 12MB, 2% more or less. The Q9000 series processor have SSE4 instructions which will help you in media encoding that support SSE4. Thats where you will see huge performace gains. Thats why you would wait for a Q9000 over the Q6000 Series.
 
As others here have alluded to, it really depends on your metric of success when comparing two processors.

If your metric of success is system-level (at the wall) power consumption, then no way an overclocked Q6600 is going to beat a QX9650.

If your metric of success is the performance (time to completion) of an application that is L2 cache-size dependant then there is the obvious chances that the QX9650 will outperform...unless you continue to move the goal-posts of what you consider an overclocked Q6600 so as to always ensure you "win" your argument 😉

If your metric of success is performance/price or performance/time (speed) of applications that fit inside 8MB L2 cache then you stand good chance to have an overclocked Q6600 match or best a QX9650 at stock.

Who forks out the cash to purchase a multiplier unlocked CPU but does not take advantage of the very feature they paid so dearly for?
 
Generally yes but eventually the market catches up with you. For example my E6400 has been at 3.2GHz for almost a year and a half. I was at 3.2GHz even before Intel was at 3.0GHz. I was at 3.2GHz for $220, Intel was at 2.93GHz for $1000. Intel's chip was 4MB cache while mine was 2MB. But for the apps I was running, mainly video editing, the extra cache did basically nothing to increase performance. Plus I had a few hundred MHz on them. As time has passed I have been somewhat caught in the price/performance metric but am still doing reasonably well.

And then sometimes you'll get $900 of the way to $1000 with $76 as in the case of my E2160 at 3.0 on stock cooling.
 
Originally posted by: harpoon84
Well, a Q6600 @ 3.2GHz will match a stock QX9650. Since most Q6600s can be overclocked to around 3.6GHz, I'd say yes, it has the potential to surpass a stock QX9650, easily.

Of course if you overclock the QX9650 to 4GHz+ it's not much of a contest. 😛
Click to expand...

^ only way this Topic could will make sense..

nice 1 harpoon. 😀
 
I think when intel makes a lower clocked version of qx9650 (meaning, same cache size/process just clocked lower) laster this year, in most likeliness, it can be overclocked to or surpass the stock qx9650 . That chip in the hands of a overclocker will be faster. For instance, my friend got a E6600 running stock at 2.4ghz. I overclocked a E2140 M0 to 3.2Ghz. After several benches, I found that even in games where E6600 has an edge with extra cache, my overclocked E2140@3.2 is faster by far especially considering my RAM is overclocked to DDR800 and timings tightened as well. He's just doing DDR667 stock. Anyways, there you have it. I think overclocker in general has an edge vs. regular folks in that we pay less and get more outta the CPU//gcard//RAM etc.
 
Originally posted by: Aflac
It will not *exactly* match a higher end processor (cache and multiplier differences), but overclocking a lower-end chip will yield approximately the same performance as a higher-end chip, barring cache differences. For example, an E6420 (at stock, a 2.13ghz chip) will match an E6850 (at stock, a 3ghz chip) once overclocked - both caches are equal; the only difference is FSB speed and multiplier, which make hardly any difference.
Click to expand...


IMO, an identical CPU (same architecture/cache - ex. E6600 and X6800) with the higher FSB will outperform the higher clocked CPU with the higher multiplier.

Another words - higher FSB means better performance with the same CPU Clock Speed.

Overclocked E6600 will outperform stock X6800, or I should say: a system with the overclocked E6600 will outperform the system with stock X6800.

 
Well, all of the CPUs are cut from the same die. Some chips have less cache, disabled features, or a lower default FSB. The $1000 chips are generally cut from the centre of it and generally clock higher on lower volts. That said, most CPUs made on the same process in the same week should generally overclock to within a few hundred Mhz of each other.

So: You pay $800 extra for perhaps 200-300Mhz, which in turn is maybe 10% extra performance. If there is a cache differential, it can have a further impact, usually in the range of 5% to 30%, depending on the size of the cache and the particular application.

This is, however, an extreme example. Generally a top-end CPU will cost around $500, with the exception of some extremely overpriced chips that come along now and again. 😉
 
Originally posted by: SickBeast
Well, all of the CPUs are cut from the same die. Some chips have less cache, disabled features, or a lower default FSB. The $1000 chips are generally cut from the centre of it and generally clock higher on lower volts. That said, most CPUs made on the same process in the same week should generally overclock to within a few hundred Mhz of each other.
Click to expand...

Are you restricting that claim to CPUs that were sold at the same rated speed?
 
Originally posted by: CTho9305
Originally posted by: SickBeast
Well, all of the CPUs are cut from the same die. Some chips have less cache, disabled features, or a lower default FSB. The $1000 chips are generally cut from the centre of it and generally clock higher on lower volts. That said, most CPUs made on the same process in the same week should generally overclock to within a few hundred Mhz of each other.
Click to expand...

Are you restricting that claim to CPUs that were sold at the same rated speed?
Click to expand...

Normally i would, but the binning lately seems to be a big trash can labeled "CPU" instead of actual bins 😛

As long as its the same physical cpu (e.g. Kentsfield) on the same stepping, they should all overclock to similar speeds.

Even Xeon vs its equivalent Part seems to be very very similar.

Edit: Spelling.
 
The answer is simply, no.

Unlocked multipliers are always nice to have. Does that mean you need to spend the extra money, probably not.
 
Overclocking means different things to different people. To me, it means a cheap CPU giving me roughly the same overall performance as an expensive CPU. I think of it as tapping underutilized performance. 😀

Originally posted by: harpoon84
Since most Q6600s can be overclocked to around 3.6GHz
Click to expand...

Doesn't that depend on stepping? I'd hesitate to make a blanket statement like that.
 
Originally posted by: Zap
Overclocking means different things to different people. To me, it means a cheap CPU giving me roughly the same overall performance as an expensive CPU. I think of it as tapping underutilized performance. 😀

Originally posted by: harpoon84
Since most Q6600s can be overclocked to around 3.6GHz
Click to expand...

Doesn't that depend on stepping? I'd hesitate to make a blanket statement like that.
Click to expand...

Unless youre buying used, its extremely unlikely youll get a B3 stepping now.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top