8350 Socket Temps.

[RGone]

Hales bales...forgot shett bout stock/default shett.
RGone...

 

[papadan]

Did you have the Cpu at stock settings if not in order to get the factory settings it needs to be at default.

Oops, didn't know that

 

[RGone]

Oops, didn't know that

Put me out to pasture...I forgot.
R....Gone.

 

[papadan]

OK, here goes again.

 

[Mandrake4565]

Ok cool looks like it's a 1.362 stock VID chip, like Gonesters and C/D's. Though it seems to like a lot less Cpu V up in the 4.7 range.

 

[papadan]

Ok cool looks like it's a 1.362 stock VID chip, like Gonesters and C/D's. Though it seems to like a lot less Cpu V up in the 4.7 range.

OK not knowing anything about P-States, What would someone use this kind of info for?

 

[RGone]

OK not knowing anything about P-States, What would someone use this kind of info for?

Owning one cpu at a time and buying "new" when we do buy...I cannot think now of a thing knowing all the P-States is really useful for.

If you were buying used, then knowing the P-State voltage that the cpu wants at say x20 multiplier would give ''me" some idea of which to buy of three or so that I knew the P-States for.

I thought you "might" have one of the later <1.31 VIDD at 4.0Ghz and they will do 4.8Ghz on air without a problem. Work good at 4.8ghz but not any higher no matter how much voltage you shoot thru them. You do n0t seem to have what I was thinking you might have.

I will say though you seem to have a VERY good cpu. Wish I had one like it under 'my' water block. Hehehe.
RGone...sterzzzzzzzzzzzzzzzzzz

 

[caddi daddi]

I run a 75 cfm, 120 mm fan on my vrms and when it's going to be working for a long time it's two.
I run one on the backside as exaust, it moves very little air, if you move to much on the backside it just makes the sensor read cooler than it really is.

 

[papadan]

Owning one cpu at a time and buying "new" when we do buy...I cannot think now of a thing knowing all the P-States is really useful for.

If you were buying used, then knowing the P-State voltage that the cpu wants at say x20 multiplier would give ''me" some idea of which to buy of three or so that I knew the P-States for.

I thought you "might" have one of the later <1.31 VIDD at 4.0Ghz and they will do 4.8Ghz on air without a problem. Work good at 4.8ghz but not any higher no matter how much voltage you shoot thru them. You do n0t seem to have what I was thinking you might have.

I will say though you seem to have a VERY good cpu. Wish I had one like it under 'my' water block. Hehehe.
RGone...sterzzzzzzzzzzzzzzzzzz

Thanks RGone, So I guess I won't worry about P-States at all.
Dan

 

[papadan]

I run a 75 cfm, 120 mm fan on my vrms and when it's going to be working for a long time it's two.
I run one on the backside as exaust, it moves very little air, if you move to much on the backside it just makes the sensor read cooler than it really is.

Wow 120mms, Sounds like you have a lot more room in you case than I do. I have small 2" (?mm) on both VRM & socket but they made a big difference in my temps.
Dan

 

[Johan45]

if you move to much on the backside it just makes the sensor read cooler than it really is.

We call that the wind chill north of the border, Oh Baby I can tell you it sure feels real.

Wow 120mms, Sounds like you have a lot more room in you case than I do. I have small 2" (?mm) on both VRM & socket but they made a big difference in my temps.
Dan

Those are 50mm chipset fans or HS fans. The chipset I have is a real screamer when running wide open.

 

[papadan]

We call that the wind chill north of the border, Oh Baby I can tell you it sure feels real.


Those are 50mm chipset fans or HS fans. The chipset I have is a real screamer when running wide open.

So do you agree with too much air back there is no good?


Thanks for the 50mm insight, not too good with metric. Those little guys do put out the air.
Dan

 

[papadan]

Yesterday I tried to do a run with prime at 22x215 & about 2 min. in it started to drop a core & then another one. I started to up the volts & after 4 ticks on the offset it was still dropping cores real quick. Also sometimes it would freeze the sys. I decided that it probably wasn't a Vcore problem. I thought maybe it was the memory not liking that I had it set manually to the XMP-1600 profile. I decided instead of relaxing the mem. timing that I would go up .5 on the multiplier & down to 211 on the FSB. This morning did another run & guess what? It worked great!! I ran prime for about 45 min. & no problems. I am very happy this morning.

 

[RGone]

215 FSB dropped to 211FSB and P95 passes can indicate that CPU_NB voltage was too low for right at 2600Mhz. Would have to up the CPU_NB voltage some.

Slowing FSB from 215 to 211 also slowed the ram clock so it could be the ram did not like being pushed past DDR3-1700 where it is now. Or it could be the ram did not like the tighter timings when forced above DDR3-1700, though this I doubt if the ram is 1.5V ram and you have it set to 1.55V.

A number of variables when about 40Mhz additional speed causes P95 Blend failures.
RGone...

 

[Mandrake4565]

215 FSB dropped to 211FSB and P95 passes can indicate that CPU_NB voltage was too low for right at 2600Mhz. Would have to up the CPU_NB voltage some.

Slowing FSB from 215 to 211 also slowed the ram clock so it could be the ram did not like being pushed past DDR3-1700 where it is now. Or it could be the ram did not like the tighter timings when forced above DDR3-1700, though this I doubt if the ram is 1.5V ram and you have it set to 1.55V.

A number of variables when about 40Mhz additional speed causes P95 Blend failures.
RGone...

On this point, Dan since you're trying to play around and see what you can do with this chip. Try doing this, work on the memory now, drop the cpu multi to something like 19/19.5 and the nb freq so it's in the 2300 range. Then start bumping the FSB, this will help you see how far you can push the ram at those timings. If you lower the multiplier and Nb Freq and you still have issues at 215 then try dropping the Ram Mhz or loosen the timings.

 

[papadan]

215 FSB dropped to 211FSB and P95 passes can indicate that CPU_NB voltage was too low for right at 2600Mhz. Would have to up the CPU_NB voltage some. I have the NB set to 1.26v, is that too low?

Slowing FSB from 215 to 211 also slowed the ram clock so it could be the ram did not like being pushed past DDR3-1700 where it is now. Or it could be the ram did not like the tighter timings when forced above DDR3-1700, though this I doubt if the ram is 1.5V ram and you have it set to 1.55V. The ram is set to 1.55v. my idea was the tight timing, I'm going to loosen the timing & run it at 22x215 again & see what happens.

A number of variables when about 40Mhz additional speed causes P95 Blend failures. Thanks for all the input
RGone...

 

[papadan]

On this point, Dan since you're trying to play around and see what you can do with this chip. Try doing this, work on the memory now, drop the cpu multi to something like 19/19.5 and the nb freq so it's in the 2300 range. Then start bumping the FSB, this will help you see how far you can push the ram at those timings. If you lower the multiplier and Nb Freq and you still have issues at 215 then try dropping the Ram Mhz or loosen the timings.

Thanks Mandrake, will give it a try!

 

[Mandrake4565]

Just letting you into the bizarre would in my brain that's all Dan. Through frustration I have learned to try and figure out what my Ram will do on X Cpu and take notes of everything. My reasoning is this, if you know the sticks will do x mhz at x timings on X Cpu. When ocing with the FSB you'll know if you have issues and you're above what you "know" X Cpu likes the ram at, then you have a better idea of what is the issue. Furthermore, "EVERY" cpu will like and allow the ram to run differently!!!!!!!! What you start acquiring parts like the bunch of us you'll see what I mean. I have 5 AMD cpu's 2 athlons, 2 phenom II's and the 8350 as well as a 2500k and a 4770k that I play with. I also have at least 6 different pairs of ram and thankfully the clowns in the rodeo told me to take notes in the beginning of my overclocking foray. From these notes I can see and have now been testing my different sticks of ram, that some of my cpu's will allow tighter timings at X mhz and some will allow higher Mhz but not tighter timings. Anyhow keep playing it's fun stuff!

 

[papadan]

Just letting you into the bizarre would in my brain that's all Dan. Through frustration I have learned to try and figure out what my Ram will do on X Cpu and take notes of everything. My reasoning is this, if you know the sticks will do x mhz at x timings on X Cpu. When ocing with the FSB you'll know if you have issues and you're above what you "know" X Cpu likes the ram at, then you have a better idea of what is the issue. Furthermore, "EVERY" cpu will like and allow the ram to run differently!!!!!!!! What you start acquiring parts like the bunch of us you'll see what I mean. I have 5 AMD cpu's 2 athlons, 2 phenom II's and the 8350 as well as a 2500k and a 4770k that I play with. I also have at least 6 different pairs of ram and thankfully the clowns in the rodeo told me to take notes in the beginning of my overclocking foray. From these notes I can see and have now been testing my different sticks of ram, that some of my cpu's will allow tighter timings at X mhz and some will allow higher Mhz but not tighter timings. Anyhow keep playing it's fun stuff!

All great wisdom, I'm glad I'm not as complicated as you & some of the other guys at this point. I have enough to process with just one sys.

 

[Mandrake4565]

All great wisdom, I'm glad I'm not as complicated as you & some of the other guys at this point.

Once you get the bug it's hard to get away, this is your first warning.