• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

E6420 Overclock Thread

3712.jpg
 
I have been working with OCing some of these new CPU's and I like to use SiSandra cpu arithmatic benchmark to compare where the OC is actually sitting at. I have done 3 E4300 cpu's and 1 E4400 cpu thus far and have a E6420 on the way from newegg right now.

After reading this thread I must say I am kinda disapointed in what I may receive. I would like to see your SiSandra scores in Arithmatic CPU Benchmark to see if the 4Meg cache makes any difference.

Here is an example of a OC I did on one of my E4300 CPU's.

3262ghzE4300.jpg


3DMark06 same CPU at 3.331ghz
benchtest2.jpg



E4300 C2D @ 3.331ghz 1.50vcore
E4400 C2D @ 3.500ghz 1.65vcore
EVGA 680i SLI 122-CK-NF68-T1
2 gigs OCZ SLI Ready 1066 mhz ram
1 EVGA 8800GTS 640 Meg ACS3
Thermaltake Armor Water cooled case
2 10,000 RPM Raptor 150gig HD
Dell 24" LCD 1920x1280 resolution
 
I have been working with OCing some of these new CPU's and I like to use SiSandra cpu arithmatic benchmark to compare where the OC is actually sitting at. I have done 3 E4300 cpu's and 1 E4400 cpu thus far and have a E6420 on the way from newegg right now.

After reading this thread I must say I am kinda disapointed in what I may receive. I would like to see your SiSandra scores in Arithmatic CPU Benchmark to see if the 4Meg cache makes any difference.

Here is an example of a OC I did on one of my E4300 CPU's.


E4300 C2D @ 3.331ghz 1.50vcore
E4400 C2D @ 3.500ghz 1.65vcore
EVGA 680i SLI 122-CK-NF68-T1
2 gigs OCZ SLI Ready 1066 mhz ram
1 EVGA 8800GTS 640 Meg ACS3
Thermaltake Armor Water cooled case
2 10,000 RPM Raptor 150gig HD
Dell 24" LCD 1920x1280 resolution



I'm still running orthos on 3.7 currently. What I can tell you is that applications that use alot of math, ie re-encoding, super pi show larger performance gains from the cache. Gaming doesn't seem to benefit at all. I would guess that Sandra Arithmetic, if designed to use the larger cache will show gains.


Don't expect the larger cache to be a noticable difference unless you are really crunching numbers and even then its probably a slight gain. I can tell you this, my E6300 @ 3.6 would run 1M Super Pi just over 16 seconds. The 6420 and 6600s I have can do it at 14.08, same speed. 2 seconds faster in 1M, I'd imagine a few mins in 32M.
 
stable.jpg


Was perfectly happy with this until I saw that last one... now I feel like I'm doing something wrong. Heh.
 
I've never seen a system that was 4hour+ stable on Orthos, that couldn't take a run through 3dMark.

On the other hand, I've been able to completely run through a full trial assault on 3dMark, and crash with Orthos in under an hour.

I think at some point, all of these requirements are overkill.
I can boot at almost 3600Mhz on air.
I can run Super Pi at 3540Mhz.
I can run 3dMark at 3500Mhz.
I can run Orthos 4+ hours at 3460Mhz.
I can run 24/7 at 3400Ghz.......

We can go on for days.
Stick with a simple, stress-testing method, and be done with it.

.


What's the matter, got a system that isn't completely stable when you throw more things at it to test it other than a 4 hour stability test?

FFS, how much harder is it after you have your Orthos run or even before your Orthos run to boot up another program to see if your system can make it through it? If your idea is to compile a load of settings that other people with similar systems can use as a benchmark and guide, would it not make sense to attempt to at least achieve complete system stability?

Any OC'er knows that you can run prime or orthos all day without touching anything else, boot up something 3D and there goes your system. But to each their own I guess.
 
User Name: Mr.White
Stability: Is it stable as it failed after 7h of Orthos?

I think the cause was that Core 0 reached 57°C as shown on the TAT screen.
I will try to re-sit the heatsink as I always have 3°C difference (both idle and under load).

Anyways, this CPU needs lots of Volts to reach "high" OCs. It looks common for 6420's from the end of 2006.

CPU: E6420
OC: 3460Mhz
Stepping: 6 B2
Week: L652
Code: B309
Vcore: 1.5625 (in BIOS)

Cooling: Air
Cooler brand/model: Thermalright Ultra 120

Motherboard: Asus P5K Deluxe WiFi
Bios Version: 0311
Operating System: XP Pro 32

RAM brand: G.Skill
RAM rated speed: 6400 (HZ)
RAM actual speed: 1020 Mhz
Latency: 4-4-4-12

My6420@3400.jpg
 
Mr. White: My cpu was loading around 64, 65 C when I did my 3.4ghz @ 1.5375v I think yours at 57C is not a temperature issue. Some chips just can't hack it at the higher FSB / voltage.

The rules for the overclock thread says at least 4 hours on orthos. So for the sake of being on the database, you should be ok. But would you really want to run it at that speed knowing it could crash on your under extended heavy load?
 
Mr. White: My cpu was loading around 64, 65 C when I did my 3.4ghz @ 1.5375v I think yours at 57C is not a temperature issue. Some chips just can't hack it at the higher FSB / voltage.

The rules for the overclock thread says at least 4 hours on orthos. So for the sake of being on the database, you should be ok. But would you really want to run it at that speed knowing it could crash on your under extended heavy load?

Well, I'll try with just a little more Voltage and see if it can do around 10 hours. Off course the question is really whether Orthos failed after 7 hours as it could just have failed after 1 hour? If so, I will certainly lower the FSB a little. However, I rarely (read never) use my computer for 7 hours in a row at full load.

Maybe it was not a CPU temp issue, but a NB temp problem. With my current room temp at 26-27°C the NB idles at 37°C and reaches around 41-42°C under load. What’s your opinion?
 
Hiyas,
I posted in the Official Database sticky thread with my overclock. I was able to get a 3.48Ghz overclock at 1.45v with CPU VTT at 1.30v (+0.10v) and the MCH at 1.30v (+0.05v).
I tried 3.656 Ghz but it failed Orthos at about 90 min, I think that I need to lap my cpu to get speeds that high, as temp was pretty high (orthos failed after ac was turned off). I think ill stick to 3.48 for everyday operations (after I lap I will retest though). I then tried to get my FSB to max but it seems it wont go all that high, 480-490 unstable and not much more :(
I was running CPU VTT at +0.20v, MCH at +0.20v, ICH at +0.05v, and ICHIO at +0.05v to get 487 FSB, but it still failed orthos. I am kinda wary of running the MB at higher voltages as it is passively cooled. My case has good cooling however. What can I push the voltages to safely?

Mox
 
Well, I'll try with just a little more Voltage and see if it can do around 10 hours. Off course the question is really whether Orthos failed after 7 hours as it could just have failed after 1 hour? If so, I will certainly lower the FSB a little. However, I rarely (read never) use my computer for 7 hours in a row at full load.

Maybe it was not a CPU temp issue, but a NB temp problem. With my current room temp at 26-27°C the NB idles at 37°C and reaches around 41-42°C under load. What’s your opinion?

That's completely up to you. Is it 'stable' enough to pass a 4 hour session of Orthos to enter into the 'stable' settings database? Sure.

Not too sure I would be happy with current settings if I knew my system 'could' crash as shown by your screenshot.
 
How are you guys reading NB temperatures? I'm running a DS3 rev 1.3 and using Everest Ultimate Edition - is it the "motherboard" sensor?

Just curious. :p
 
How are you guys reading NB temperatures? I'm running a DS3 rev 1.3 and using Everest Ultimate Edition - is it the "motherboard" sensor?

Just curious. :p

Hi. On the P5K, I assumed the MB/System sensor was the one of the NB. I may be wrong though. Anyone knows for sure?
 
That's completely up to you. Is it 'stable' enough to pass a 4 hour session of Orthos to enter into the 'stable' settings database? Sure.

Not too sure I would be happy with current settings if I knew my system 'could' crash as shown by your screenshot.

Well, rest assured that I am not completely happy :p
I'll give a try with a slightly higher CPU Voltage. If that doesn't help, I'll just lower the FSB by 5 and settle there until a new bios comes out and try again. With the summer coming I don't want to push the volts too high :)
 
What's the matter, got a system that isn't completely stable when you throw more things at it to test it other than a 4 hour stability test?

FFS, how much harder is it after you have your Orthos run or even before your Orthos run to boot up another program to see if your system can make it through it? If your idea is to compile a load of settings that other people with similar systems can use as a benchmark and guide, would it not make sense to attempt to at least achieve complete system stability?

Any OC'er knows that you can run prime or orthos all day without touching anything else, boot up something 3D and there goes your system. But to each their own I guess.

You missed the part where I said it passes your suggested test and failed on the one recommended by the thread starter.

I'm not usually up for a pissing contest, but since I will be spending 64.2769 hours trying to fulfill the system test, stress, time-limit, program variations that you require in order to finally deem a system, "Stable"......I'll have a little extra time on my hands to get into it with you if that's what you prefer. :cool:
 
Please try to stay on topic!

A processor is considered stable enough for this thread if it passes 4 hour orthos. End of story. Stop bickering please so we can get back to the subject at hand.

I will be adding all the posts so far either late tonight or tomorrow (whenever I finish my final project for a database course I just took).

Sorry, I cant help myself. I found that my 3D screensaver actually makes it more likely that I pass orthos. I have to disable it so that I get a legit test for my stable overclock claims in this forum. This is because the temperatures drop on my cpu while it waits on my ATI x1650. If a system is less stable on 3D apps, you might want to take a look at your graphics card.

This forum will continue to focus on help with overclocking the E6420. So far we have found that it certainly requires higher voltage than most, and it seems that it requires a significant voltage ramp up as you approach its fsb limit.

It seems that 3650Mhz is around the best you can hope for although 3700 is? may be in reach.

Some processors may only reach around 3200Mhz, which still acheives similar mathematical performance as some of the highest E4x00 series overclocks to date.

3500Mhz turns in sub 15 second superpi times on good ram timings.

Please contribute information like the above examples. Thank you!
 
You missed the part where I said it passes your suggested test and failed on the one recommended by the thread starter.

I'm not usually up for a pissing contest, but since I will be spending 64.2769 hours trying to fulfill the system test, stress, time-limit, program variations that you require in order to finally deem a system, "Stable"......I'll have a little extra time on my hands to get into it with you if that's what you prefer. :cool:

I agree man. All this testing for stablilty is sometimes ridiculous. I don't have(nor want to spend) the time on running all these benchmarks just to prove stability. I'd much prefer spending my time encoding or playing a game then watching Orthos run for 10 hours. I didn't spend money to run benchmarks.
 
I agree man. All this testing for stablilty is sometimes ridiculous. I don't have(nor want to spend) the time on running all these benchmarks just to prove stability. I'd much prefer spending my time encoding or playing a game then watching Orthos run for 10 hours. I didn't spend money to run benchmarks.

Amen, Brother.....;)

And now, back on topic.
Here is what I've been able to spit out with 3500'ish and tight RAM timings, for those that would like to compare / compete:

14.jpg

.
I'm still trying to get a stable (4+ Hours on Orthos :p ) overclock at higher than 3500Mhz.
I made it just over 3 hours with 3500 / 1200, but alas....she crashed and burned on me.

......back to the drawing board.
I think I'll run 3dMark at 3540 / 1250 to kill some time while I figure out how to get a stable overclock above 3500.......
.
.
 
I know this sounds crazy but have you tried underclocking your ram to 400mhz using 1T instead of 2T and even lower timmings.
 
I know this sounds crazy but have you tried underclocking your ram to 400mhz using 1T instead of 2T and even lower timmings.

You want me to go 400Mhz and lower than 4-3-3-4 in order to get 1T?

I can't imagine how it would make up for the difference in raw speed that it loses by dropping 700Mhz......

But I might try just for the hell of it. :D
 
Yah the ram speed IMO doesnt make that much of a difference as far as performance its only a factor in where you want to OC at 1:1 etc. Some people say that sync the ram with the FSB gives best performance which I tend to agree with, however I have seen some VERY high OC's 5ghz+ where they have their ram at 400mhz lowest possible latency and 1T timing and they seem to swear by it.
 
Yah the ram speed IMO doesnt make that much of a difference as far as performance its only a factor in where you want to OC at 1:1 etc. Some people say that sync the ram with the FSB gives best performance which I tend to agree with, however I have seen some VERY high OC's 5ghz+ where they have their ram at 400mhz lowest possible latency and 1T timing and they seem to swear by it.

Interesting.

I don't think the issue here is the RAM though......I think it's my MoBo.
She craps out before she ever hits 450Mhz, so with an 8x Multi, I'm not even able to try for 3600Mhz. :(
 
I'm baaaaaackk....just posted this up in the Intel OC thread.

I'm taking back the E6420 crown...for now...moving to a D-Tek Fusion this week, went from P5B Deluxe to DS3 to push higher.

User Name: dnottis
Stability: 6 Hours Orthos

CPU: E6420
OC: 3680 Mhz
Stepping: 6 B2
Week: L709
Code: A171
Vcore: 1.550 (BIOS), 1.49v actual

Cooling: Water
Cooler brand/model: Thermochill PA120.2, Swiftech Apogee, D5 Pump, MCW60, 2x 75CFM 120s @ 12v

Motherboard: Gigabyte DS3 rev.1, C1 chipset
Bios Version: F11 Bios
Operating System: Vista x64, XP SP2

RAM brand: OCZ 2GB
RAM rated speed: 8500
RAM actual speed: 920Mhz
Latency: 5-5-5-15

3680.jpg

Your only running 1.49v of actual vcore to hit 3680? If so, you should have no problem getting that thing 3700+++ stable.

I have to use about 1.58v after vdroop to get 3680 to be stable. Im curious to see what results you can get using about 1.55-1.58v.

Nice work!
 
Your only running 1.49v of actual vcore to hit 3680? If so, you should have no problem getting that thing 3700+++ stable.

I have to use about 1.58v after vdroop to get 3680 to be stable. Im curious to see what results you can get using about 1.55-1.58v.

Nice work!


I'm trying like hell man. I can get about 3 hours, then it will BSOD and reboot. The P5B was doing the same thing. I've got some new tubing coming, D-Tek Fusion block, backplates and some new 90cfm fans. I will break 3.7 dammit!!!

Gonna retube / refill system, put in new CPU block and add some more airflow to my PA120.2. Time to retire the original apogee and flush out 1 year old fluid...we'll see what happens. Currently I'm pushing 1.54v actual @ 3.7 and its still rebooting on me. Load temps are about 53 deg c.
 
If your load temps are 53C do you really think a different water block and tubes is going to make a difference? 65C I beleive is the Intel ceiling on the C2D although after all the CPU's I have done I am lead to beleive its 90C.

Why arent you guys going higher with your Core Voltage 1.6 vcore is pretty common when trying to find your ceiling for OCing shoot several of my CPUs were at 1.65 (With non-conductive liquid cooling).

C2D are really some mind boggling CPUs for OCing, cant wait to get back home this weekend and play with my new E6420 and contribute some real hands on knowledge about them.
 
If your load temps are 53C do you really think a different water block and tubes is going to make a difference? 65C I beleive is the Intel ceiling on the C2D although after all the CPU's I have done I am lead to beleive its 90C.

Why arent you guys going higher with your Core Voltage 1.6 vcore is pretty common when trying to find your ceiling for OCing shoot several of my CPUs were at 1.65 (With non-conductive liquid cooling).

C2D are really some mind boggling CPUs for OCing, cant wait to get back home this weekend and play with my new E6420 and contribute some real hands on knowledge about them.

haha no, I'm changing the tubing cause its all clouded :) and the liquid is over a year old now. The block may give me a few degrees - I have someone offering me $40 for an original Apogee, how can I go wrong lol.

However the chip seems to crash once it starts hitting 53 / 54 when loaded tho. After orthos runs about 2 1/2 - 3 hours temps start to creep over 53 and thats when it always reboots on me. I'm rebuilding the entire loop anyways, has nothing to do with this - hoping some fresh liquid and a new block help, but I'm not doing it to hit 3.7 LOL.
 
e6420 batch #L704A885 @ 3440 @1.45V (1.425V after vdrop) with a syth. ninja rev B (w/ as5) on a gigabyte p35c-ds3r.

being held back by ram, ordered some d9 based pc6400 firestix. (on sale!!! :D )

compuh7.jpg

when the ram arrives, I am going to do some vmods to this board. going for 3.6ghz+
 
finally managed to test mine with orthos.

CPU: e6420, L702C555, pack date 04/13/07
HSF: Thermalright XP90C with Panaflo M1a 92mm fan at 1950rpm (very quiet but not silent)
MB: Asus P5K Deluxe
RAM: 2gb Crucial Ballistix Pc2-6400 44412 - at 615Mhz 55515:)
GFX: PCI Radeon 7000 64Mb (x1950Xt on its way).
PSU: Corsair HX620W
Drives: 2x320gb SATA2, 2x160gb SATA2, 2x optical
Case: Chieftec full tower

Currently running all Volts on minimum in BIOS, CPU: 1.3125V, DDR: 2.2V:)

screenies:


orthos:



once id done all the orthos'ing
[img=http://img184.imageshack.us/img184/8493/410x8stockvoltsddr21230kx3.th.jpg]

EDIT: wrong screenies above:
http://www.imagehosting.com/out.php/i766813_e6420overclock.JPG



will bring back some more benchies once the new graphics card arrives!

i have got to say - this is the best money i have ever spent. although it was very difficult installing XP on the RAID volume (had to disable USB then it worked a dream, must be a driver issue), overclocking is very easy. The temps are very stable too. Not wanting to do anything drastic just yet - very warm in my loft-conversion in the daytime!
 
Updated entry.

I now have 2 entries, if that's allowed.

The first one was the uber-RAM speed version, this one is just for a nice, solid, stable overclocked CPU speed.

User Name: Evoic
Stability: 5+ hours - Small FFTs Orthos Beta

CPU: E6420 (Model L651B325)
OC: 3500Mhz
Stepping: 6 B2
Week: L651
Code: B325
Vcore: 1.53v (in BIOS) 1.49v (Actual)

Cooling: Air
Cooler brand/model: Scythe Mine Rev. B

Motherboard: eVGA 680i LT
Bios Version: P04
Operating System: XP Pro 32

RAM brand: PNY Value Ram
RAM rated speed: PC-2 5300 / 667Mhz
RAM actual speed: 673Mhz
Latency: 4-4-4-12

CPUZ Validation

3500-Submission.jpg
 
I just posted this in the database thread...swapped blocks and some new fans.


User Name: dnottis
Stability: 4 Hours Orthos

CPU: E6420
OC: 3712 Mhz
Stepping: 6 B2
Week: L709
Code: A171
Vcore: 1.6 (BIOS), 1.54v actual

Cooling: Water
Cooler brand/model: Thermochill PA120.2, D-tek FuZion, D5 Pump, MCW60

Motherboard: Gigabyte DS3 rev.1, C1 chipset
Bios Version: F11 Bios
Operating System: XP SP2

RAM brand: Crucial Ballistix 2GB
RAM rated speed: 8000
RAM actual speed: 928Mhz
Latency: 5-5-5-15

3712.jpg


IMG_4231.jpg
 
So, after checking the time-stamps, I see that I was able to enjoy my new higher stable-speed post for exactly 5 minutes before dnottis came along and kicked me in the nuts with his water-cooled rig again.

:p :p :p
 
Well like I said got my new E6420 this weekend and here goes my first test will run orthos tonight to see how it does but how about a 3.8ghz superpi score of 13.625 with some 85$ GSkill ram hehe waiting on my OCZ ram to get back from my RMA. Anywhoo here is the screenie more to come.

piscore.jpg
 
Request that your 3.8Ghz be posted as a Max unstable overclock and then get a stable overclock (4 hours + of Dual Prime or Orthos) and have that one added to the main list.

Unless of course you end up running that 3.8Ghz for 4 hours +......

In which case, :eek: :eek: :eek:

p.s. Nice chip, but even nicer board ! I can't get my 680i LT over 440'ish.
 
Beat ya... 13.5 ;) ...and Orthos stable @ 3.712 Ghz. Haven't tried to push higher for Orthos.

3792.jpg
 
thats really nice, it crashed in orthos after 2hrs so gotta tweak it more when i get time. I cant wait to get my ram back from OCZ and see what my superpi is at 3-3-3-8-1T timmings. They have been out of stock for a week and unable to replace it waiting on their RM Manager to call me back im hoping I can get him to give me 2 gigs of their Flex water cooled DDR2 which would be really nice.

Dnott what are all your voltage's at I set my ram and core volts and let the system auto the rest I have a feeling thats where I need tweak it to get it stable because my Temps are VERY low compared to the E4400 and E4300 at 1.65+ Vcore at full load i hit up to 80C with them and only hit 65C with the E6420.
 
thats really nice, it crashed in orthos after 2hrs so gotta tweak it more when i get time. I cant wait to get my ram back from OCZ and see what my superpi is at 3-3-3-8-1T timmings. They have been out of stock for a week and unable to replace it waiting on their RM Manager to call me back im hoping I can get him to give me 2 gigs of their Flex water cooled DDR2 which would be really nice.

Dnott what are all your voltage's at I set my ram and core volts and let the system auto the rest I have a feeling thats where I need tweak it to get it stable because my Temps are VERY low compared to the E4400 and E4300 at 1.65+ Vcore at full load i hit up to 80C with them and only hit 65C with the E6420.


Memory is at 2.1v, Vcore 1.6v with droop 1.54v actual. +.2 to all others, +.1 to PCI-e.
 
thats really nice, it crashed in orthos after 2hrs so gotta tweak it more when i get time. I cant wait to get my ram back from OCZ and see what my superpi is at 3-3-3-8-1T timmings. They have been out of stock for a week and unable to replace it waiting on their RM Manager to call me back im hoping I can get him to give me 2 gigs of their Flex water cooled DDR2 which would be really nice.

Dnott what are all your voltage's at I set my ram and core volts and let the system auto the rest I have a feeling thats where I need tweak it to get it stable because my Temps are VERY low compared to the E4400 and E4300 at 1.65+ Vcore at full load i hit up to 80C with them and only hit 65C with the E6420.

65c is quite toasty.
You want to always make sure you remain within about 15-20c below T-Junction at a minimum.

Also, I'm dealing with an OCZ RMA on my beloved Reapers.
Tell Jimmy Houn that Jeff says hello. :p
 
Updated entry.

I now have 2 entries, if that's allowed.

The first one was the uber-RAM speed version, this one is just for a nice, solid, stable overclocked CPU speed.

User Name: Evoic
Stability: 5+ hours - Small FFTs Orthos Beta

CPU: E6420 (Model L651B325)
OC: 3500Mhz
Stepping: 6 B2
Week: L651
Code: B325
Vcore: 1.53v (in BIOS) 1.49v (Actual)

Cooling: Air
Cooler brand/model: Scythe Mine Rev. B

Motherboard: eVGA 680i LT
Bios Version: P04
Operating System: XP Pro 32

RAM brand: PNY Value Ram
RAM rated speed: PC-2 5300 / 667Mhz
RAM actual speed: 673Mhz
Latency: 4-4-4-12

CPUZ Validation

Congrats Evoic, I know how long you've been working on that elusive 3.5Ghz

I'm sorry to everyone that I haven't posted overclock information yet. I ran into issues at work an had to spend a lot of time in the office that I didn't expect to need to put in, I'm currently on vacation until this Sunday. Still anxiously waiting on my overclock to show up in the official thread's database.

Keep the insane stuff coming, and please, if you make a new record, list everything you're willing to tell us about how you managed your overclock. Especially focus on volts, how high, where, and of course, the board/bios you're using.

I think that I have decided to do my volt mod, and also lap and apply that fancy thermal paste I found. I want 3.7 before I post my final overclock on the official thread. Also, has anyone beat 3820Mhz in unstable boots? I have no screenshot, but I managed a 13.250 second 1M SuperPi Mod time at 3820Mhz with 1:1 RAM. Perhaps I'll add an additional spreadsheet listing superpi times... I figure why not?

Anyhow, keep it coming, it's nice to see 3.7Ghz stable is possible on this chip, and I'm glad to see how active everyone has been in this thread.

I'll see if I can bear using this hunk of junk I'm posting from to put the first spreadsheet together tomorrow.
 
Back
Top