My E4400 runs 3.4Ghz 24/7 with only 1.45V, so don't just assume they aren't good o/c'ers. There are a few guys that are also at 3.3-3.5Ghz with the E4300/4400s at HardOCP and Xtreme. I can easily run Orthos 12hrs stable @ 3.6Ghz @ 1.52V, but I see no need to deal with the extra heat and voltage.
My E4400 runs 3.4Ghz 24/7 with only 1.45V, so don't just assume they aren't good o/c'ers. There are a few guys that are also at 3.3-3.5Ghz with the E4300/4400s at HardOCP and Xtreme. I can easily run Orthos 12hrs stable @ 3.6Ghz @ 1.52V, but I see no need to deal with the extra heat and voltage.
A friend of mine currently (unless it recently has been broken) has the WC for overclocking an E4300 on air. Stock 1.8Ghz up to 3943 Mhz IIRC. thats approx a 119% overclock so yeah, they do clock just as good as conroes 
Originally posted by: Yanagi
A friend of mine currently (unless it recently has been broken) has the WC for overclocking an E4300 on air. Stock 1.8Ghz up to 3943 Mhz IIRC. thats approx a 119% overclock so yeah, they do clock just as good as conroes
Id love to see a link to this, i dont believe it.
lopri
Elite Member
- Jul 27, 2002
- 13,209
- 594
- 126
C2Ds are maxing around 3.0~3.2GHz for production level, or under 'default' voltage. For us enthusiasts, it can be anywhere between 3.4~3.7GHz with extra voltages and better thermal environment. Without a doubt Penryn will take it further up (maybe 3.5~4.0GHz?), but I don't think we'll see an SKU higher than 3.5GHz. A big factor to consider is that it's very likely that higher clocking, cooler running chips will go to quad-core SKUs. So it really depends on how market responds to the quads, and to a lesser extent how well Barcelona performs. Intel's 65nm manufacturing is as mature as can be, and the quality of Conroe chips have improved quite a bit over the year, IMO. (Look how many SKUs they're pumping out, covering the entire market segments - from budget to extreme - with C2Ds) Looking at how well Intel responds to the market, I'm beginning to think it's a huge mistake that AMD decided to design a 'native' quad-core before they didn't have enough expertise on 65nm.
I thought we went through that with netburst?
except that 4ghz didn't feel like 4ghz... :brokenheart:
Highest OC'd C2D was a 6700 OC'd to 5.59GHz on LN.
Highest 4300 OC I've seen is to 4.68GHz on a cascade system....but that OC doesn't get into the top 30 C2D OC's.
HWBot.org has a LOT of the top OC's.....not all, but quite a few.
http://www.hwbot.org//listResu...=13&cpuSubFamilyId=122
Highest 4300 OC I've seen is to 4.68GHz on a cascade system....but that OC doesn't get into the top 30 C2D OC's.
HWBot.org has a LOT of the top OC's.....not all, but quite a few.
http://www.hwbot.org//listResu...=13&cpuSubFamilyId=122
Rafael
Senior member
- May 11, 2001
- 868
- 0
- 0
I agree with n7 and lopri.
I can get my E6400 Allendale to run @ 4Ghz.
Vcore @ 1.55, Thermalright Ultra-120. It is not stable, but I can get
It is not that easy, at all, to reach 4 Ghz stable.
And I would like to add that the performance gain is not that huge, using a Asus P5B Deluxe WiFi, I can't speak for Abit and DFI P965 mobos.
I was not able to get the screen shot of SuperPi 1 Mb @ 4 Ghz, since it was not stable, but I saw the result and the gain was .250 sec.
So maybe with the P35, X38, Penryn we will be able to actually get to 4 Ghz and not end up like JAG87 said.
I can get my E6400 Allendale to run @ 4Ghz.
Vcore @ 1.55, Thermalright Ultra-120. It is not stable, but I can get
It is not that easy, at all, to reach 4 Ghz stable.
And I would like to add that the performance gain is not that huge, using a Asus P5B Deluxe WiFi, I can't speak for Abit and DFI P965 mobos.
I was not able to get the screen shot of SuperPi 1 Mb @ 4 Ghz, since it was not stable, but I saw the result and the gain was .250 sec.
So maybe with the P35, X38, Penryn we will be able to actually get to 4 Ghz and not end up like JAG87 said.
@rafael... 4 ghz vs what speed?
you should see more than a 0.250 second increase... if not, then something is wrong unless you're comparing 3.9 to 4.0...
3.2 to 4.0 should see a dramatic decrease in 1M superpi
you should see more than a 0.250 second increase... if not, then something is wrong unless you're comparing 3.9 to 4.0...
3.2 to 4.0 should see a dramatic decrease in 1M superpi
Originally posted by: Acanthus
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
A friend of mine currently (unless it recently has been broken) has the WC for overclocking an E4300 on air. Stock 1.8Ghz up to 3943 Mhz IIRC. thats approx a 119% overclock so yeah, they do clock just as good as conroes
Id love to see a link to this, i dont believe it.
And link you shall get
Originally posted by: Yanagi
<div class="FTQUOTE"><begin quote>Originally posted by: Acanthus
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
A friend of mine currently (unless it recently has been broken) has the WC for overclocking an E4300 on air. Stock 1.8Ghz up to 3943 Mhz IIRC. thats approx a 119% overclock so yeah, they do clock just as good as conroes
Id love to see a link to this, i dont believe it.</end quote></div>
And link you shall get
Dear lord, 1.7 volts!
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
<div class="FTQUOTE"><begin quote>Originally posted by: Acanthus
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
A friend of mine currently (unless it recently has been broken) has the WC for overclocking an E4300 on air. Stock 1.8Ghz up to 3943 Mhz IIRC. thats approx a 119% overclock so yeah, they do clock just as good as conroes</end quote></div>
Id love to see a link to this, i dont believe it.</end quote></div>
And link you shall get
</end quote></div>
And its a skewed record.
"The cooling was Noctua NH-U12 + Delta FFB1212EHE at around +5 degrees centigrade room temperature."
Its air cooling in a 40 degree room at 1.7 volts.
Itd not only be a dead cpu in hours, but air cooling on the frigging north pole isnt exactly normal air cooling
Also, theres no proof that it was stable, even under these retarded conditions.
<div class="FTQUOTE"><begin quote>Originally posted by: Acanthus
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
A friend of mine currently (unless it recently has been broken) has the WC for overclocking an E4300 on air. Stock 1.8Ghz up to 3943 Mhz IIRC. thats approx a 119% overclock so yeah, they do clock just as good as conroes
</end quote></div>Id love to see a link to this, i dont believe it.</end quote></div>
And link you shall get
</end quote></div>
And its a skewed record.
"The cooling was Noctua NH-U12 + Delta FFB1212EHE at around +5 degrees centigrade room temperature."
Its air cooling in a 40 degree room at 1.7 volts.
Itd not only be a dead cpu in hours, but air cooling on the frigging north pole isnt exactly normal air cooling
Also, theres no proof that it was stable, even under these retarded conditions.
Originally posted by: Acanthus
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
<div class="FTQUOTE"><begin quote>Originally posted by: Acanthus
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
A friend of mine currently (unless it recently has been broken) has the WC for overclocking an E4300 on air. Stock 1.8Ghz up to 3943 Mhz IIRC. thats approx a 119% overclock so yeah, they do clock just as good as conroes
Id love to see a link to this, i dont believe it.</end quote></div>
And link you shall get
</end quote></div>
And its a skewed record.
"The cooling was Noctua NH-U12 + Delta FFB1212EHE at around +5 degrees centigrade room temperature."
Its air cooling in a 40 degree room at 1.7 volts.
Itd not only be a dead cpu in hours, but air cooling on the frigging north pole isnt exactly normal air cooling
Also, theres no proof that it was stable, even under these retarded conditions.
I'd never said you should be stable. Most WRs done are not on stable systems either. Just enough stable to post CPU-Z validation.
I can agree with the +5 centigrade room temp, but hey, it is still air
Originally posted by: Yanagi
<div class="FTQUOTE"><begin quote>Originally posted by: Acanthus
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
<div class="FTQUOTE"><begin quote>Originally posted by: Acanthus
<div class="FTQUOTE"><begin quote>Originally posted by: Yanagi
A friend of mine currently (unless it recently has been broken) has the WC for overclocking an E4300 on air. Stock 1.8Ghz up to 3943 Mhz IIRC. thats approx a 119% overclock so yeah, they do clock just as good as conroes
Id love to see a link to this, i dont believe it.</end quote></div>
And link you shall get
</end quote></div>
And its a skewed record.
"The cooling was Noctua NH-U12 + Delta FFB1212EHE at around +5 degrees centigrade room temperature."
Its air cooling in a 40 degree room at 1.7 volts.
Itd not only be a dead cpu in hours, but air cooling on the frigging north pole isnt exactly normal air cooling
Also, theres no proof that it was stable, even under these retarded conditions.
</end quote></div>
I'd never said you should be stable. Most WRs done are not on stable systems either. Just enough stable to post CPU-Z validation.
I can agree with the +5 centigrade room temp, but hey, it is still air
Well then intel wins with their 10ghz ALUs theyve had in the lab for years
They dont have to be stable, as long as they can execute a 1MB program that reads their clockspeed.
Rafael
Senior member
- May 11, 2001
- 868
- 0
- 0
Shimmishim,
Compared to 3.5 Ghz - 500 x 7. One is stable that I can use 24/7 not much of a vcore bump.
I will post a few results from my tests here:
SuperPi 1.5 - 1 Mb
Config - Mem 400 Mhz - CAS 4-4-3-8 = 17.860s
Config - Mem 400 Mhz - CAS 3-3-3-8 = 18.250s
Config - Mem 402 Mhz - CAS 4-4-3-8 = 18.171s
Config - Mem 402 Mhz - CAS 3-3-3-8 = 18.094s
Config - Mem 458 Mhz - CAS 4-4-3-8 = 17.875s (multi 7)
Config - Mem 501 Mhz - CAS 4-4-4-12 = 16.343s (multi 7)
I think it is due to the NB Strap of the Asus P5B Deluxe WiFi.
I know that I have a good C2D chip here with me. I just did not have enough time to play around. But hopefuly in a few weeks I will have 10 days off work. And I will do another tests.
Let me know about these results.
I will be later tonight here.
TIA Shimmishim!
Rafael
you're showing the mem but not the mult so it's hard to tell what speed you're testing...
I can show you 3.2 vs. 4.0 or 3.5 vs. 4.0 and you will definitely see a dramatic decrease in 1M pi and 32M pi.
I can also show you 4.3 ghz pi if you'd like...
every mhz counts especially if you're a bencher.
and yes the strap issue is something that definitely effects the scores on the P5B
there is a HUGE drop in bandwidth going from 400 to 401 mhz FSB since the strap changes from 1066 to 1333.
if you compare a pi run at 3.6 9x400 vs. 3.61 9x401 you will actually see a dramatic change due to the strap change.
4 ghz > 3.5 ghz
Rafael
Senior member
- May 11, 2001
- 868
- 0
- 0
Shimmishim,
Sorry, I got the wrong results posted here. And I did a mistake also it was not compared to 3.5 Ghz, it was 4 Ghz compared to 3.8 Ghz.
Those results were from my initial test to see how the NB strap act like.
Here is the full results I got:
SuperPi 1.5 XS
Config - FSB/Memory 458 Mhz - CAS 4-4-3-8 = 17.875s (multi 7)
Config - FSB/Memory 501 Mhz - CAS 4-4-4-12 = 16.343s (multi 7)
Config - FSB/Memory 400 Mhz - CAS 4-4-4-8 = 17.750s (multi 8)
Config - FSB/Memory 400 Mhz - CAS 4-4-3-8 = 17.860s (multi 8)
Config - FSB/Memory 400 Mhz - CAS 3-3-3-8 = 18.250s (multi 8)
Config - FSB/Memory 402 Mhz - CAS 4-4-3-8 = 18.171s (multi 8)
Config - FSB/Memory 402 Mhz - CAS 3-3-3-8 = 18.094s (multi 8)
Config - FSB/Memory 425 Mhz - CAS 4-4-3-8 = 17.204s (multi 8)
Config - FSB/Memory 450 Mhz - CAS 4-4-3-8 = 16.359s (multi 8)
Config - FSB/Memory 475 Mhz - CAS 4-4-3-10 = 15.390s (multi 8)
Config - FSB/Memory 488 Mhz - CAS 5-5-5-12 = 15.234s (multi 8)
Config - FSB 400 Mhz - Memory 500 Mhz - CAS 5-5-5-12 = 17.469s (multi 8)
I don't have the screen of 4 Ghz running SuperPi since it was not stable, but I remember the score 15.163s,
Thanx!
Rafael
Sorry, I got the wrong results posted here. And I did a mistake also it was not compared to 3.5 Ghz, it was 4 Ghz compared to 3.8 Ghz.
Those results were from my initial test to see how the NB strap act like.
Here is the full results I got:
SuperPi 1.5 XS
Config - FSB/Memory 458 Mhz - CAS 4-4-3-8 = 17.875s (multi 7)
Config - FSB/Memory 501 Mhz - CAS 4-4-4-12 = 16.343s (multi 7)
Config - FSB/Memory 400 Mhz - CAS 4-4-4-8 = 17.750s (multi 8)
Config - FSB/Memory 400 Mhz - CAS 4-4-3-8 = 17.860s (multi 8)
Config - FSB/Memory 400 Mhz - CAS 3-3-3-8 = 18.250s (multi 8)
Config - FSB/Memory 402 Mhz - CAS 4-4-3-8 = 18.171s (multi 8)
Config - FSB/Memory 402 Mhz - CAS 3-3-3-8 = 18.094s (multi 8)
Config - FSB/Memory 425 Mhz - CAS 4-4-3-8 = 17.204s (multi 8)
Config - FSB/Memory 450 Mhz - CAS 4-4-3-8 = 16.359s (multi 8)
Config - FSB/Memory 475 Mhz - CAS 4-4-3-10 = 15.390s (multi 8)
Config - FSB/Memory 488 Mhz - CAS 5-5-5-12 = 15.234s (multi 8)
Config - FSB 400 Mhz - Memory 500 Mhz - CAS 5-5-5-12 = 17.469s (multi 8)
I don't have the screen of 4 Ghz running SuperPi since it was not stable, but I remember the score 15.163s,
Thanx!
Rafael
Rafael
Senior member
- May 11, 2001
- 868
- 0
- 0
Yeah my initial goal of those tests were to see if it was worth spending the PLUS on memory that you can run with tight timings and how much Mhz you need to add in the FSB to catch the same result with tight timings.
But the 965P does not get much advantage of it, as you can see that CAS 3 has a worse result than the CAS 4. The new P35 takes more advantage of it.
Probably my next move will be to 2 more GB of good memory and a P35 mobo.
But the 965P does not get much advantage of it, as you can see that CAS 3 has a worse result than the CAS 4. The new P35 takes more advantage of it.
Probably my next move will be to 2 more GB of good memory and a P35 mobo.
@Rafael
your results are unique to the asus p5b series boards.
4-4-4 > 4-4-3 > 3-3-3
which should not be the case at all.
3-3-3 should be the fastest.
Not quite sure what it is about the p5b but that's how it is.
Now if you were to get a 975x board or a different brand 965, you'd see vastly different results.
also, you must be using a 2MB chip with your times.
at 3.6, i can pull off sub 14 seconds at 1M
at 4.0, i can get in the 12.5 to 12.6 range
that's 1.5 seconds with a 400 mhz gain and that's pretty significant.
your results are unique to the asus p5b series boards.
4-4-4 > 4-4-3 > 3-3-3
which should not be the case at all.
3-3-3 should be the fastest.
Not quite sure what it is about the p5b but that's how it is.
Now if you were to get a 975x board or a different brand 965, you'd see vastly different results.
also, you must be using a 2MB chip with your times.
at 3.6, i can pull off sub 14 seconds at 1M
at 4.0, i can get in the 12.5 to 12.6 range
that's 1.5 seconds with a 400 mhz gain and that's pretty significant.
Rafael
Senior member
- May 11, 2001
- 868
- 0
- 0
Yeah Shimmishim,
P5B is the worst of the mobos based on the P965. It does OC very good in FSB. I really can reach 515 Mhz in the FSB. The problem that P5B does not have a good performance is the NB strap. While the DFI and Abit have the NB strap @ 1066 Mhz over 400 Mhz FSB, the P5B NB strap jumps to 1333 Mhz. And when you go above 500 Mhz FSB it goes to another NB strap.
And from what I have seen the latencies seem to be a problem with the P965 chipset, I don't know about the second revision of the chipset.
So the final result is that Abit and DFI P965 mobos have a better result than the Asus P5B.
Getting back on the track of the 4 Ghz, if you are going to aim this high, I think you will want to take all the advantage of your invested money. So the best bet right now are the mobos based on the P35. Or wait for the X38.
Rafael
P5B is the worst of the mobos based on the P965. It does OC very good in FSB. I really can reach 515 Mhz in the FSB. The problem that P5B does not have a good performance is the NB strap. While the DFI and Abit have the NB strap @ 1066 Mhz over 400 Mhz FSB, the P5B NB strap jumps to 1333 Mhz. And when you go above 500 Mhz FSB it goes to another NB strap.
And from what I have seen the latencies seem to be a problem with the P965 chipset, I don't know about the second revision of the chipset.
So the final result is that Abit and DFI P965 mobos have a better result than the Asus P5B.
Getting back on the track of the 4 Ghz, if you are going to aim this high, I think you will want to take all the advantage of your invested money. So the best bet right now are the mobos based on the P35. Or wait for the X38.
Rafael
TRENDING THREADS
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 10K
-
Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)- Started by Vattila
- Replies: 13K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 7K
-
Facebook
Twitter