Wolfdale and Yorkfield watch thread and performance thread

[Tatty_Two]

680i handle memory better? i'd say so... my 1066mhz sticks run up to 1290mhz on my 680i and only 1120mhz on my x38 and about the same on my p35.

You need the right BIOS for memory with the DQ6, my home PC is knackered at the moment but if you are not running BIOS F6 you will not be getting the best out of either memory or CPU, this BIOS is the best balanced of the lot and prety much bug free, that cannot be said for the rest of them, trust me on this one, I tried 5 of them! But TBH things may work different for the Wolfdale.

Unfortunatly, the best BIOS of them all is pre 45nm support but the F6 should be better than the F7.

 

[trt740]

Tatty Xeon 3320 ?

You need the right BIOS for memory with the DQ6, my home PC is knackered at the moment but if you are not running BIOS F6 you will not be getting the best out of either memory or CPU, this BIOS is the best balanced of the lot and prety much bug free, that cannot be said for the rest of them, trust me on this one, I tried 5 of them! But TBH things may work different for the Wolfdale.

Unfortunatly, the best BIOS of them all is pre 45nm support but the F6 should be better than the F7.

well when that bad boy gonna be up

 

[giorgos th.]

the quest for the best 24/7 config continues....(warm room,low rpm fans etc etc)

 

[trt740]

Did you also have cores thats had different

the quest for the best 24/7 config continues....(warm room,low rpm fans etc etc)

temps

 

[trt740]

thanx..
the only two things i`m concerned are the ~10C difference between the 2 cores (probably a monitoring thing) and the minor issues with the multi and the mobo...
with clear multi f.e. 8-9 i had problems,with .5 f.e. 8.5-9.5 everything was good..

this is what I mean

 

[giorgos th.]

no not in this one.

 

[trog100]

i have settled for 4 gig at roughly 1.3 vcore.. 450 x 9.. memory at 900.. i just like nice round figures.. he he he

trog

 

[trt740]

no not in this one.

I ment that you did have a chip with a similar problem to mine correct?

 

[giorgos th.]

yeap,but no performance issue..

 

[trt740]

yeap,but no performance issue..

Well this one performs very well just won't overclock aswell as I exspected but it will do over 4.0ghz .

 

[trog100]

coretemp really isnt reading these bloody wolfdale chip temps correctly.. i am now running the sandra cpu burn in wizzard in the background while browsing and doing whatever..

why people are taking so much notice of what its telling em i have no idea..

looking at the taskbar i see one core going between 62 and 68 C.. the other one sits there showing 60 C and never moves..

at idle one core show 45 C-ish and moves about.. the other one is locked at 49 C and again never moves.. if i load the cores the one locked at 49 jumps to 60 then never moves..

my bloody bios wont even read the chip temp so why should coretemp be able to..

my bios is locked at 4 C.. others have reported all sorts of silly figures.. some as high as 129 C..

something is odd about how these chips work as regarding the temp sensor..

i dont know what coretemp reads to produce its figures but it should not be relied upon to even be vaguely accurate..

it might be hard for some to not know what temp their chip is running at.. but i recon thats the reality of the situation..

trog

 

[trt740]

coretemp really isnt reading these bloody wolfdale chip temps correctly.. i am now running the sandra cpu burn in wizzard in the background while browsing and doing whatever..

why people are taking so much notice of what its telling em i have no idea..

looking at the taskbar i see one core going between 62 and 68 C.. the other one sits there showing 60 C and never moves..

at idle one core show 45 C-ish and moves about.. the other one is locked at 49 C and again never moves.. if i load the cores the one locked at 49 jumps to 60 then never moves..

my bloody bios wont even read the chip temp so why should coretemp be able to..

my bios is locked at 4 C.. others have reported all sorts of silly figures.. some as high as 129 C..

something is odd about how these chips work as regarding the temp sensor..

i dont know what coretemp reads to produce its figures but it should not be relied upon to even be vaguely accurate..

it might be hard for some to not know what temp their chip is running at.. but i recon thats the reality of the situation..

trog

Your right trog my core temps reads my cpu's at 40c idle not sure why it hit me so late that not possible my Quad core xeon didn't idle at those temps. that way too high and the prior chip was reading 25C ish. Right now check this out. This is at idle with a Zerotherm 120 at 80 percent fan

I think that Asus probe temp is right and core temps is reading this second chip wrong and I wonder why. The prior chip was matching probes temps. I wonder why chip one would read fine but chip two doesn't

 

[Fitseries3]

You need the right BIOS for memory with the DQ6, my home PC is knackered at the moment but if you are not running BIOS F6 you will not be getting the best out of either memory or CPU, this BIOS is the best balanced of the lot and prety much bug free, that cannot be said for the rest of them, trust me on this one, I tried 5 of them! But TBH things may work different for the Wolfdale.

Unfortunatly, the best BIOS of them all is pre 45nm support but the F6 should be better than the F7.

ok... i sold the dq6.. can anyone help me with a maximus formula? i've got it to 4.32ghz but it stops there.

 

[trt740]

Xeon quad 105 watt chip 65nm under full load compared to a 45nm chip with speed step

Your right trog my core temps reads my cpu's at 40c idle not sure why it hit me so late that not possible my Quad core xeon didn't idle at those temps. that way too high and the prior chip was reading 25C ish. Right now check this out. This is at idle with a Zerotherm 120 at 80 percent fan

I think that Asus probe temp is right and core temps is reading this second chip wrong and I wonder why. The prior chip was matching probes temps. I wonder why chip one would read fine but chip two doesn't

There is no way under load that the Xeon 105 watt chip 65nm could run almost as cool as a 45nm chip at idle. No way only 6 degree difference

 

[trt740]

chip one overclocked idle/ chip two not overclocked idle

same system same cooler both E8400

how is this chip cooler

 

[DOM]

^^ Barbie lol

I would RMA it seems you'll have trouble with it later

 

[Wile E]

Yeah, I'd say there is definitely something amiss with the Wolfdale temp readouts. Kinda reminds me of the goofy readouts CoreTemp woould give me on my Brisbane. Core Temp just doesn't seem to like some chips.

 

[trog100]

one other thing i have noticed about coretemp.. its not actually reading both cores.. it shows a figure for both cores.. but when i load both cores i see.. say 65 C with one core the other kinda similar but fixed..

when i only load one core i should still see 65 C on the working core and way way less on the none working core.. least my logic tells me i should..

what do see is about 55 C on the fully working core and a fixed something similar on the none working core..

i think it might be correct or somewhere near correct at high temps just on one core the other read out means nothing..

at lower temps it just loses the plot and show a temp way higher than it really is..

if it really was reading two cores separately with one fully working and the other at idle the temps should be different for both cores.. they are not.. they are always roughly the same.. just lower with one core working than they are with two..

trog

 

[trt740]

^^ Barbie lol

I would RMA it seems you'll have trouble with it later

yes once again I have 4 small children 3 are girls under 12

 

[trog100]

the dude who wrote coretemp seems to think the problem is the chip itself.. hmmm..

backed up perhaps by this.. bit of plagiarism on my part but who cares.. one has to be on the ball.. he he

http://www.guru3d.com/newsitem.php?id=6317

trog

 

[Fitseries3]

i agree... every bit of software i've used to monitor temps has show a 10-15 degree Celcius difference between the core and the cpu and one core is at least 5c higher than the other and seems to hang around that temp no mater what. really weird.... only the cpu temp changes on my e8400.

 

[trog100]

do we send em all back.. he he he

the problem is a bit worse for me cos it buggers up my nice abit guru fan control.. if poor old guru thinks my cpu temp never alters form 4 C he cant do much fan controling can he.. he he

trog

 

[Franklinwallbrown]

i have settled for 4 gig at roughly 1.3 vcore.. 450 x 9.. memory at 900.. i just like nice round figures.. he he he

trog

I like to sqeeze those extra bits out! +.14 FTW! lol

 

[trt740]

it's not the chip it's the programs

do we send em all back.. he he he

the problem is a bit worse for me cos it buggers up my nice abit guru fan control.. if poor old guru thinks my cpu temp never alters form 4 C he cant do much fan controling can he.. he he

trog

read this on this thread http://www.xtremesystems.org/forums/showthread.php?t=173999&page=8

It seems many E8400's have core temp sensors which report incorrect low temperature readings. These false readings are being reported in all software monitoring programs.

I had a E8400 which reported idle temps of 17C in one core and 34C in the other. Under load, the 34C core readings were frozen until it exceeded 34C.
I exchanged that processor a new one, albeit from the exact same batch, but the new cpu more correctly has core temps of 38C/36C (4.0Ghz with 1.34v).

Programs like CoreTemp read the diode right in the cpu, so software updates or bios upgrades will have no effect on these false readings.

Some E8400 purchasers are being told to just ignore these false core readings and instead rely on the "cpu temp" reading, rather than core temps.

 

[trt740]

it's not the chip it's the programs

do we send em all back.. he he he

the problem is a bit worse for me cos it buggers up my nice abit guru fan control.. if poor old guru thinks my cpu temp never alters form 4 C he cant do much fan controling can he.. he he

trog

read this on this thread http://www.xtremesystems.org/forums/showthread.php?t=173999&page=8

It seems many E8400's have core temp sensors which report incorrect low temperature readings. These false readings are being reported in all software monitoring programs.

I had a E8400 which reported idle temps of 17C in one core and 34C in the other. Under load, the 34C core readings were frozen until it exceeded 34C.
I exchanged that processor a new one, albeit from the exact same batch, but the new cpu more correctly has core temps of 38C/36C (4.0Ghz with 1.34v).

Programs like CoreTemp read the diode right in the cpu, so software updates or bios upgrades will have no effect on these false readings.

Some E8400 purchasers are being told to just ignore these false core readings and instead rely on the "cpu temp" reading, rather than core temps.



another intresting statement

As someone posted in another thread, Intel designed these core sensors to read extremley high core temps to throttle down the cpu when overheated. They apparantly were not designed to be used as thermometers or necessarily be accurate at lower temperatures. While some enterprising software designers have brought out programs to read these core sensors, Intel is probably not obligated to ensure their accuracy at low temps. Intel's own Thermal Analysis Tool does not work with the Wolfdale cpus.