QX6850 Extreme Restoration Blog

[agent_x007]

QX6850 needs good VRM to go high on clock
FSB is easy, decrease the multiplier and up the voltage on NB.
I used QX6700 (that's B3 revision of Kentsfield, G0 should be better than that).
Just FYI : I tested/measured just now how much power goes from PSU into my UD3Ps 8-pin EPS power plug (when this CPU is OC'ed to 3,6GHz, Vcore is at 1,34-1,39V). Tested on "Standard" IntelBurnTest.
@storm-chaser Would you like to guess amps value (that is power delivered by PSU to VRM) ?

 

[storm-chaser]

QX6850 needs good VRM to go high on clock
FSB is easy, decrease the multiplier and up the voltage on NB.
I used QX6700 (that's B3 revision of Kentsfield, G0 should be better than that).
Just FYI : I tested/measured just now how much power goes from PSU into my UD3Ps 8-pin EPS power plug (when this CPU is OC'ed to 3,6GHz, Vcore is at 1,34-1,39V). Tested on "Standard" IntelBurnTest.
@storm-chaser Would you like to guess amps value (that is power delivered by PSU to VRM) ?

I would guess we are in the neighborhood of 7-12 amps delivered by the PSU? I really have no idea I'm just going by feel here....

BTW- I did a little overclocking of my own tonight

FSB 500 --

And 480 FSB with 9x multi (by this CPU overclocks like a dream):

Should I go for the magical 50% overclock? LOL do you even have to ask???

 

[hat]

4.4 on that old chip would be an impressive result if stable...

 

[storm-chaser]

4.4 on that old chip would be an impressive result if stable...

This old chip has been in my possession since 2008 came out of a trashed worstation computer. And yes I'm running AIDA64 stress tests on it now. So far there are "no limits" devott laratt!! I'll be happy when I have more time to work on it tomorrow. CPU has never ever been over clocked this is my first go around vs the e8400. Think of all the untapped potential out there!

 

[AlwaysHope]

This old chip has been in my possession since 2008 came out of a trashed worstation computer. And yes I'm running AIDA64 stress tests on it now. So far there are "no limits" devott laratt!! I'll be happy when I have more time to work on it tomorrow. CPU has never ever been over clocked this is my first go around vs the e8400. Think of all the untapped potential out there!

Yes, those 45nm dual cores were easy to OC the living daylights out of them. But in today's world, I think dual cores are past their prime. Imo, quads as a minimum from now on.
BUT get the FSB OC potential from those duals & make the quads do it for real under sustained heavy load for daily use. That would be the ultimate from this decade old tech.

 

[agent_x007]

I would guess we are in the neighborhood of 7-12 amps delivered by the PSU? I really have no idea I'm just going by feel here....

Add 1 before that "7", and switch around "1" and "2" on that "12".
17-21A on 12V rail at 3,6GHz and 1,34-1,38V (LLC enabled).
My EP45-UD3P shuts itself down if EPS 8-pin goes over 20A
On the other hand...

 

[storm-chaser]

Well we are getting closer to the 50% threshold.... but that 4.7 number by agentX_0007 is going to be hard to beat especially since I'm out of multiplier

AND good to know those voltage and amp #s how did you measure your amps was it a calculation from voltage or did you actually use a multimeter or something?

Edging up on the number

Achieved the magical 50% overclock mark (that's a record for me across both platforms) and 14Gb/s on the memory read. Very impressed with what this little E8400 can do!

 

[Mr.Scott]

Kick it man.

https://valid.x86.fr/zvwunc

 

[storm-chaser]

Well done, Mr. Scott.

I think I've got my work cut out for me to play with big boys.

In any event this E8400 swap has inspired me to test / benchmark / overclock all of my other C2D chips seeing as how I have quite a surplus at the moment And also my motherboard is proving to be quite the requirable workhorse, good investment for sure.. Over the next couple weeks I will be benching the following
CPUs:
-Intel Core 2 Duo e7500
-Intel Core 2 Duo e4200
-Intel core Pentium D 960
-A couple other intels just cant come to mind right now to the fullest extent.

 

[agent_x007]

@storm-chaser Clamp meter :

 

[Mr.Scott]

Well done, Mr. Scott.

I think I've got my work cut out for me to play with big boys.

In any event this E8400 swap has inspired me to test / benchmark / overclock all of my other C2D chips seeing as how I have quite a surplus at the moment And also my motherboard is proving to be quite the requirable workhorse, good investment for sure.. Over the next couple weeks I will be benching the following
CPUs:
-Intel Core 2 Duo e7500
-Intel Core 2 Duo e4200
-Intel core Pentium D 960
-A couple other intels just cant come to mind right now to the fullest extent.

You just need a cooling upgrade.
Message me at home. I have some CPU's I can send you.

 

[agent_x007]

https://valid.x86.fr/9n6hea

That is it for DDR2 2x2GB
Board really cannot reach high FSB clocks with 4x2GB RAM configuration.
Maybe GTL NB would change that (I pushed 1,56V into NB with a E8600, and still no-go zone... unplug memory - works fine), but I didn't wanted to play around too much.
On to DDR3...

PS. Just noticed - Intel Processors : "Expert"... interesting.

 

[AlwaysHope]

Seeing this thread is about older Intel cpus & OC fun.....
Has anyone seen this review about i7-980X against Ryzen 3 2200G ?

I have a few criticisms of that article.

 

[storm-chaser]

Seeing this thread is about older Intel cpus & OC fun.....
Has anyone seen this review about i7-980X against Ryzen 3 2200G ?

I have a few criticisms of that article.

I like this quote from the article:
"That said the single core performance is rather weak and shockingly even at 4.4 GHz is well down on what the 2200G offers. In fact the single thread performance of the 980X was 8% slower than the 2200G and 17% slower once the APU is overclocked. Still the 2200G’s 4-threads can’t compete with the old 12-thread CPU in multi-threaded workloads and is up to 37% slower once both CPUs are overclocked."

And I agree that for most users, the single core performance is what you should be after so that would one point for the AMD chip. For business and high workload applications, the Intel wins hands down.

Here is an interesting comparison of the "alleged flagship" Core i9-9900K versus the Ryzen 2700X. Seems to soundly beat the AMD chip into submission but these are just rumors at this point and nothing is concrete.

https://www.techradar.com/news/inte...enchmark-leaves-amd-ryzen-7-2700x-in-the-dust

 

[AlwaysHope]

Storm Chaser, if techspot are going to compare 8 yr old 1366 extreme edition cpu with Zen+ 2200G why didn't they use high end DDR3 in those tests..... I mean my Gigabyte 1156 system from that era (2009) boasts DDR3 2200 capability yet they used DDR3 1600 only... what are they trying to do? dump it down a bit...lol...
If we keep in mind that 1366 system was designed for win7 & not win10 like AM4 - 2200G is that factor must be taken into account imo.

 

[agent_x007]

...I have six 2GB sticks of DDR3-1600 memory and that’s the best stuff I have available for this test.

Direct quote from article. If they had 3x4GB setup 1866MHz shoudn't be a problem. But with 6x2GB, and those are on a crapier side...
They simly went with what they have.
I slso think they didn't optimise UnCore frequency like they should have.
Also my R15 score :

 

[AlwaysHope]

Direct quote from article. If they had 3x4GB setup 1866MHz shoudn't be a problem. But with 6x2GB, and those are on a crapier side...
They simly went with what they have.
I slso think they didn't optimise UnCore frequency like they should have.
Also my R15 score :

That's the way to do it ^^^ , but lame excuse author of that techspot review supplied... in other words he couldn't be bothered optimising & it's easy to do but anyway...life goes on!

 

[storm-chaser]

Tonight I had a little OC fun with my Core 2 Duo E7500:
Lite OC:

Working on 27 % OC:

35% OC and topping out on CPU max clock ( right around 3.90Ghz)

I can get to FSB 395 but the system is not stable. I settled for 46% OC on the FSB:

 

[MrGenius]

So far there are "no limits" devott laratt!!

Devon's a monstrosity.

 

[agent_x007]

@storm-chaser I'm not sure what strap/perf. level or FSB Voltage you set. But below 1600MHz FSB on E7x00 is simply sad

 

[storm-chaser]

Must be a bad CPU or perhaps my FSB settings. Term voltage is set to 1.70 and LLC is enabled. What was really slowing me down, honestly, was the CPU topping out around 4ghz. Strap should be set to what? I'm currently on auto. Also any other FSB settings I should be aware of for this CPU?

 

[agent_x007]

1) PLL is 1,5V (you can go up to 1,65V without issues).
2) FSB Termination at ~1,45V is the max. safe volage, over it you will degrade/kill your CPU (also, with 1600MHz FSB, you shoudn't need more than 1,25V)
3) Strap is dependant on DRAM frequency you want to use. 266MHz and 333MHz straps are fine in most cases. 200MHz is better performance, but needs A LOT of NB voltage to be stable with high capacity RAM at high speeds.
I highly recommend increasing tRFC timing for DDR3 to over "80" range for ~1500MHz (and if you want to push even higher frequencies).
4) Higher RAM Frequency doesn't always equall better performance.
If tRFC doesn't help, try manually setting AI Transaction Booster to 8-10.

 

[storm-chaser]

1) PLL is 1,5V (you can go up to 1,65V without issues).
2) FSB Termination at ~1,45V is the max. safe volage, over it you will degrade/kill your CPU (also, with 1600MHz FSB, you shoudn't need more than 1,25V)
3) Strap is dependant on DRAM frequency you want to use. 266MHz and 333MHz straps are fine in most cases. 200MHz is better performance, but needs A LOT of NB voltage to be stable with high capacity RAM at high speeds.
I highly recommend increasing tRFC timing for DDR3 to over "80" range for ~1500MHz (and if you want to push even higher frequencies).
4) Higher RAM Frequency doesn't always equall better performance.
If tRFC doesn't help, try manually setting AI Transaction Booster to 8-10.

Thanks. I will update you on my progress. I may have time later today to take another look at it. I also have another E7500 to swap in incase the CPU itself is holding me back but it will take me a couple days to harvest the chip and bring it back to my place. As for TRFC I think I was running 60 on my e8400 setup without issue @ 1600Mhz -- but I will try bumping that up to 82 as well.

-Tried PLL @ 1.5v, FSB term @ 1.45v, FSB strap @ 266Mhz, tRFC @ 80 and even manually tried to set AI Booster. Still no go. I get a NO POST scenario with the FSB at anything over 396.

I know the motherboard is good because it just overclocked my e8400 like a champ. So I'm gonna try swapping to my other E7500 and see if that makes a difference. I know the memory will run at these timings no problem, so I should'nt be running into anything there. I have a feeling this CPU is just not up to the task. In terms of grade it's probably a C- as it has scratches on the heat spreader. Not sure what the history is on it.

 

[agent_x007]

Are you sure you are not trying to boot at 11 multi ?
If so, you will need ~1,45V as Vcore and ~1,6V PLL as well.
Something to think about :

I did use only memory 2 sticks.

 

[storm-chaser]

Im only using two memory sticks as well. Also note, I am using the 9X multiplier since I just want to test max FSB and not be limited by the CPU. I think it's just a bad CPU.

I have successfully harvested the other e7500 and I will try that tomorrow hopefully.