This topic is locked from further discussion.

Avatar image for deactivated-5a3583b110a97
deactivated-5a3583b110a97

877

Forum Posts

0

Wiki Points

0

Followers

Reviews: 5

User Lists: 0

#1 deactivated-5a3583b110a97
Member since 2005 • 877 Posts

I was reading that the extra cache on the e6600 actually can hold back overclocking performance, for minimal performance benefits. (Corsair System Build Reports, last page http://tools.corsairmemory.com/systembuild/report.aspx?report_id=78237&sid=13) The author of that article seemed to think that you might get better performance with a slower clock speed but a higher cache on the e6600, but that it was sort of trial and error.

 

On sort of a random side note, how important is it to have your cpu and ram running at 1:1 frequencies? Also am I correct that a 1:1 ratio with ddr2 800 ram would mean a cpu fsb of 400, multiplied by 8 for the e6400 and 9 for the e6600 to get the final clock speed?

 

Basically I'm just trying to figure out ultimately what kind of performance I can get from different processors, I've come up a little short on my expected PC budget, so I'm looking for a place to cut costs, but if I can get equal or better performance from a processor thats $100 cheaper, then that would be an obvious choice.

Avatar image for juggernaut8419
juggernaut8419

872

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#2 juggernaut8419
Member since 2003 • 872 Posts
I would recommend the e6400 and save the $100 if you are going to overclock, because it does have the best overclocking potential of that line from what I have read in the past.
Avatar image for 00al2003
00al2003

200

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#3 00al2003
Member since 2003 • 200 Posts
Ive got an e6400 currently at 3.2ghz and memory at 1000mhz (Crucial Ballistix ofcourse) on a GIGABYTE DS3 and can easily be pushed beyond that with better cooling.
Avatar image for AARONRULZ1
AARONRULZ1

6273

Forum Posts

0

Wiki Points

0

Followers

Reviews: 6

User Lists: 0

#4 AARONRULZ1
Member since 2006 • 6273 Posts

I heard that the E6420 is better at OCing than the regular E6400.

http://xbitlabs.com/articles/cpu/display/core2duo-e6420_11.html

Avatar image for SSJBen
SSJBen

7071

Forum Posts

0

Wiki Points

0

Followers

Reviews: 23

User Lists: 0

#5 SSJBen
Member since 2003 • 7071 Posts

Since when larger cache holds back in OCing? That guy must be drunk.

Lets put it this way, running an E6400 at 3.7ghz and competing it with an E6600 at 3.7ghz... the obvious winner will be the E6600. Both basically can reach the same speed. So larger cache doesn't hold you back in OCing.

Yes you can get E6600 performance with an E6400 with a simple OC from stock.

Avatar image for 00al2003
00al2003

200

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#6 00al2003
Member since 2003 • 200 Posts

I heard that the E6420 is better at OCing than the regular E6400.

http://xbitlabs.com/articles/cpu/display/core2duo-e6420_11.html

AARONRULZ1

GooDness DUDE!!! i want to be able to do that. what Motherboard are you using, and what type of cooling, power supply. And was that picture from a e6400 or the e6420 you wr talking about. I know the pic says "E6400" but just making sure.

Avatar image for deactivated-5a3583b110a97
deactivated-5a3583b110a97

877

Forum Posts

0

Wiki Points

0

Followers

Reviews: 5

User Lists: 0

#7 deactivated-5a3583b110a97
Member since 2005 • 877 Posts

I suppose another factor for me is that with the extra $100 I could possibly afford an 8800gtx over the gts assuming my budget crisis gets worked out. What do you all think of that? Does a larger cache actually give you significantly increased performance?

 

Thanks for the feedback up til now by the way 

Avatar image for AARONRULZ1
AARONRULZ1

6273

Forum Posts

0

Wiki Points

0

Followers

Reviews: 6

User Lists: 0

#8 AARONRULZ1
Member since 2006 • 6273 Posts
[QUOTE="AARONRULZ1"]

I heard that the E6420 is better at OCing than the regular E6400.

http://xbitlabs.com/articles/cpu/display/core2duo-e6420_11.html

00al2003

GooDness DUDE!!! i want to be able to do that. what Motherboard are you using, and what type of cooling, power supply. And was that picture from a e6400 or the e6420 you wr talking about. I know the pic says "E6400" but just making sure.

The E6420 isn't out yet,the mobo is a ASUS P5N-E 650i,and you can't tell the processor has extra L2 cache?

Avatar image for --JpR--
--JpR--

394

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#9 --JpR--
Member since 2004 • 394 Posts
how do you get to the CPU-Z?
Avatar image for BeavermanA
BeavermanA

2652

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#10 BeavermanA
Member since 2003 • 2652 Posts
You should wait regardless of what you choose. Either for the E6420, or the E6600 to drop $100. The extra cache offers like 2-3% improvement on average, with up to 10% with some apps. Not spectacular, but if you wait less than two weeks you'll get it for the same price, and the benefits may become more evident as time passes. I'd definitely get the E6420 + 8800GTX though if you had to cut the E6600 due to budget, the 6420 will easily accomodate the 8800 once oc'd.
Avatar image for juggernaut8419
juggernaut8419

872

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#11 juggernaut8419
Member since 2003 • 872 Posts

how do you get to the CPU-Z?--JpR--

 Download it from here.

Avatar image for Baselerd
Baselerd

5104

Forum Posts

0

Wiki Points

0

Followers

Reviews: 14

User Lists: 0

#12 Baselerd
Member since 2003 • 5104 Posts

Since when larger cache holds back in OCing? That guy must be drunk.

Lets put it this way, running an E6400 at 3.7ghz and competing it with an E6600 at 3.7ghz... the obvious winner will be the E6600. Both basically can reach the same speed. So larger cache doesn't hold you back in OCing.

Yes you can get E6600 performance with an E6400 with a simple OC from stock.

SSJBen

Technically larger cache should hold back OCing, since more current flows through the extra cache, generating more heat. However, it shouldn't make enough of a difference to back off to an e6400 if the e6600 is in your budget. 

Avatar image for Makari
Makari

15250

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#13 Makari
Member since 2003 • 15250 Posts
[QUOTE="SSJBen"]

Since when larger cache holds back in OCing? That guy must be drunk.

Lets put it this way, running an E6400 at 3.7ghz and competing it with an E6600 at 3.7ghz... the obvious winner will be the E6600. Both basically can reach the same speed. So larger cache doesn't hold you back in OCing.

Yes you can get E6600 performance with an E6400 with a simple OC from stock.

Baselerd

Technically larger cache should hold back OCing, since more current flows through the extra cache, generating more heat. However, it shouldn't make enough of a difference to back off to an e6400 if the e6600 is in your budget.

it's tended to hold it back before, however slightly - thinking of the 512k cache s754 a64s vs the 1mb cache early on, at least.
Avatar image for LordEC911
LordEC911

9972

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#14 LordEC911
Member since 2004 • 9972 Posts
[QUOTE="SSJBen"]

Since when larger cache holds back in OCing? That guy must be drunk.

Lets put it this way, running an E6400 at 3.7ghz and competing it with an E6600 at 3.7ghz... the obvious winner will be the E6600. Both basically can reach the same speed. So larger cache doesn't hold you back in OCing.

Yes you can get E6600 performance with an E6400 with a simple OC from stock.

Baselerd

Technically larger cache should hold back OCing, since more current flows through the extra cache, generating more heat. However, it shouldn't make enough of a difference to back off to an e6400 if the e6600 is in your budget.

More current isn't flowing through the extra cache though...

I also believe that more cache = greater overclocking. I made that statement in a thread in the OCU and someone stated that it would be more likely to cause instability since there are more transistors but that is also wrong, since the transistors of the cache aren't going to be "clocked" any higher. 

Avatar image for Makari
Makari

15250

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#15 Makari
Member since 2003 • 15250 Posts
[QUOTE="Baselerd"][QUOTE="SSJBen"]

Since when larger cache holds back in OCing? That guy must be drunk.

Lets put it this way, running an E6400 at 3.7ghz and competing it with an E6600 at 3.7ghz... the obvious winner will be the E6600. Both basically can reach the same speed. So larger cache doesn't hold you back in OCing.

Yes you can get E6600 performance with an E6400 with a simple OC from stock.

LordEC911

Technically larger cache should hold back OCing, since more current flows through the extra cache, generating more heat. However, it shouldn't make enough of a difference to back off to an e6400 if the e6600 is in your budget.

More current isn't flowing through the extra cache though...

I also believe that more cache = greater overclocking. I made that statement in a thread in the OCU and someone stated that it would be more likely to cause instability since there are more transistors but that is also wrong, since the transistors of the cache aren't going to be "clocked" any higher.

they're there, generating more heat at least. i really don't know the specifics of why, but i tend to deal in the practical applications of computer hardware as opposed to the theory, and it's pretty much always worked out that way for me.. when you have two processors of otherwise the same architecture side by side, the one with less cache tends to be able to clock higher. maybe it's similar to single-core vs dual-core or something - the opty 14x's could make it up to higher clocks relatively easily compared to their dual-core brothers. TBH i don't have any idea how the e6400 vs e6600 is going right now with that, but i wouldn't be surprised if it followed the pattern set in the past. the difference probably isn't so large in performance in the end, either way.
Avatar image for LordEC911
LordEC911

9972

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#16 LordEC911
Member since 2004 • 9972 Posts

they're there, generating more heat at least. i really don't know the specifics of why, but i tend to deal in the practical applications of computer hardware as opposed to the theory, and it's pretty much always worked out that way for me.. when you have two processors of otherwise the same architecture side by side, the one with less cache tends to be able to clock higher. maybe it's similar to single-core vs dual-core or something - the opty 14x's could make it up to higher clocks relatively easily compared to their dual-core brothers. TBH i don't have any idea how the e6400 vs e6600 is going right now with that, but i wouldn't be surprised if it followed the pattern set in the past. the difference probably isn't so large in performance in the end, either way.Makari

The single vs dualcore can't be compared here, for that is due to one core lacking and keeping the other one back.

Overclocking, without separate volts/clocks for each core, linearly gets harder with each additional core.

Avatar image for Makari
Makari

15250

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#17 Makari
Member since 2003 • 15250 Posts

[QUOTE="Makari"]they're there, generating more heat at least. i really don't know the specifics of why, but i tend to deal in the practical applications of computer hardware as opposed to the theory, and it's pretty much always worked out that way for me.. when you have two processors of otherwise the same architecture side by side, the one with less cache tends to be able to clock higher. maybe it's similar to single-core vs dual-core or something - the opty 14x's could make it up to higher clocks relatively easily compared to their dual-core brothers. TBH i don't have any idea how the e6400 vs e6600 is going right now with that, but i wouldn't be surprised if it followed the pattern set in the past. the difference probably isn't so large in performance in the end, either way.LordEC911

The single vs dualcore can't be compared here, for that is due to one core lacking and keeping the other one back.

Overclocking, without separate volts/clocks for each core, linearly gets harder with each additional core.

i dunno.. even on consistently high quality silicon - say, a CCBBE 0613 Opteron dual-core - the chips seem to run into a wall at a lower point than the single-core brethren that no amount of cooling or voltage will affect. it is one of too many variables to really look at though, true. :)
Avatar image for LordEC911
LordEC911

9972

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#18 LordEC911
Member since 2004 • 9972 Posts
[QUOTE="LordEC911"]

[QUOTE="Makari"]they're there, generating more heat at least. i really don't know the specifics of why, but i tend to deal in the practical applications of computer hardware as opposed to the theory, and it's pretty much always worked out that way for me.. when you have two processors of otherwise the same architecture side by side, the one with less cache tends to be able to clock higher. maybe it's similar to single-core vs dual-core or something - the opty 14x's could make it up to higher clocks relatively easily compared to their dual-core brothers. TBH i don't have any idea how the e6400 vs e6600 is going right now with that, but i wouldn't be surprised if it followed the pattern set in the past. the difference probably isn't so large in performance in the end, either way.Makari

The single vs dualcore can't be compared here, for that is due to one core lacking and keeping the other one back.

Overclocking, without separate volts/clocks for each core, linearly gets harder with each additional core.

i dunno.. even on consistently high quality silicon - say, a CCBBE 0613 Opteron dual-core - the chips seem to run into a wall at a lower point than the single-core brethren that no amount of cooling or voltage will affect. it is one of too many variables to really look at though, true. :)

One core is going to be lacking and holding the other core back, since they are currently linked together.  No matter the quality of the silicon, one will still be lacking, in comparison to the other. That is just the way it is with any manufacturing process.

Avatar image for Makari
Makari

15250

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#19 Makari
Member since 2003 • 15250 Posts
One core is going to be lacking and holding the other core back, since they are currently linked together. No matter the quality of the silicon, one will still be lacking, in comparison to the other. That is just the way it is with any manufacturing process.LordEC911
what i'm referring to is that the 'one of the cores lacking' is pretty much ALWAYS at a noticably lower point than any of the single-core cores, where the best duals are in the realm of the worst singles. if it really were a case of 'one core is just worse,' i think there would be a lot more variation. it's just too consistent!
Avatar image for LordEC911
LordEC911

9972

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#20 LordEC911
Member since 2004 • 9972 Posts

[QUOTE="LordEC911"]One core is going to be lacking and holding the other core back, since they are currently linked together. No matter the quality of the silicon, one will still be lacking, in comparison to the other. That is just the way it is with any manufacturing process.Makari
what i'm referring to is that the 'one of the cores lacking' is pretty much ALWAYS at a noticably lower point than any of the single-core cores, where the best duals are in the realm of the worst singles. if it really were a case of 'one core is just worse,' i think there would be a lot more variation. it's just too consistent!

How so?

It is mainly due to the added heat of the near double number of transistors...

Which singlecore vs dualcore are you talking about? 

Avatar image for Makari
Makari

15250

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#21 Makari
Member since 2003 • 15250 Posts

[QUOTE="Makari"][QUOTE="LordEC911"]One core is going to be lacking and holding the other core back, since they are currently linked together. No matter the quality of the silicon, one will still be lacking, in comparison to the other. That is just the way it is with any manufacturing process.LordEC911

what i'm referring to is that the 'one of the cores lacking' is pretty much ALWAYS at a noticably lower point than any of the single-core cores, where the best duals are in the realm of the worst singles. if it really were a case of 'one core is just worse,' i think there would be a lot more variation. it's just too consistent!

How so?

It is mainly due to the added heat of the near double number of transistors...

Which singlecore vs dualcore are you talking about?

opty vs opty, 90nm.. they're the only really good examples of similar architectures being modified around, same with 1mb cache vs. 512k cache. i don't think the c2d has a desktop single-core equivalent right now, does it?
Avatar image for jjjhsmith
jjjhsmith

799

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#22 jjjhsmith
Member since 2003 • 799 Posts

Since when larger cache holds back in OCing? That guy must be drunk.

Lets put it this way, running an E6400 at 3.7ghz and competing it with an E6600 at 3.7ghz... the obvious winner will be the E6600. Both basically can reach the same speed. So larger cache doesn't hold you back in OCing.

Yes you can get E6600 performance with an E6400 with a simple OC from stock.

SSJBen

 

Well...maybe.  But the E6600 has a larger multiplier, of 9x instead of the E6400s 8x.  That means at 3.7 ghz, the FSB of the E6600 would be 411 mhz, while the E6400 would be 462 mhz.  The faster FSB might be enough to overcome the approximately 3% performance gain that the E6600 has from the doubled cache size.  

Avatar image for LordEC911
LordEC911

9972

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#23 LordEC911
Member since 2004 • 9972 Posts
[QUOTE="SSJBen"]

Since when larger cache holds back in OCing? That guy must be drunk.

Lets put it this way, running an E6400 at 3.7ghz and competing it with an E6600 at 3.7ghz... the obvious winner will be the E6600. Both basically can reach the same speed. So larger cache doesn't hold you back in OCing.

Yes you can get E6600 performance with an E6400 with a simple OC from stock.

jjjhsmith

 

Well...maybe. But the E6600 has a larger multiplier, of 9x instead of the E6400s 8x. That means at 3.7 ghz, the FSB of the E6600 would be 411 mhz, while the E6400 would be 462 mhz. The faster FSB might be enough to overcome the approximately 3% performance gain that the E6600 has from the doubled cache size.

Good luck finding an E6400 that can clock that high... 

Avatar image for Baselerd
Baselerd

5104

Forum Posts

0

Wiki Points

0

Followers

Reviews: 14

User Lists: 0

#24 Baselerd
Member since 2003 • 5104 Posts
Back to the extra cache vs. half cache. Think about this. You have current flowing through more cache area on the chip, thus more heat is generated. Simple really. On the e6400, the extra cache is still there, just deactivated (cut with a laser even if my memory serves right.) That would technically cool the processor, giving it a higher heat generated per heat capacity ratio, thus dispersing the heat across the die more. However, this difference is so small it's negligible.
Avatar image for LordEC911
LordEC911

9972

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#25 LordEC911
Member since 2004 • 9972 Posts

Back to the extra cache vs. half cache. Think about this. You have current flowing through more cache area on the chip, thus more heat is generated. Simple really. On the e6400, the extra cache is still there, just deactivated (cut with a laser even if my memory serves right.) That would technically cool the processor, giving it a higher heat generated per heat capacity ratio, thus dispersing the heat across the die more. However, this difference is so small it's negligible.Baselerd

Also, it doesn't exactly work into the equation when you take the heat/temps out of the picture with a phase changer. The E6600 is still going to overclock higher on suicide runs then the E6400, about 800mhz higher.