This CPU is the latest thing, can't believe it. :eek:
Enjoy it, maybe you'll never get such an exemplar. :up:
Printable View
This CPU is the latest thing, can't believe it. :eek:
Enjoy it, maybe you'll never get such an exemplar. :up:
:D Happy new year to all XS members !!!:)
Here we have 5 hours prime test with 2893 loops @ 4GHz @ 1.208 volts...:p: :
http://img69.imageshack.us/img69/3701/4ghzprimeeb6.jpg
Temps in the right down corner are correct, read with coretemp 0.95.6. - room temp at 20°C.
Think it's stable..............:yepp:
:up:
You've had your time playing with that toy, now wrap it up and send it to me. :D
Did you max it out yet?
L739A664 or L739A765 with
version # E23438-002 1.25Vmax
are crazy clockers for air, water & phase......:)
:D Nope - 10.5 x 450 at 1.488 volts is not maxed out - next step is to get this 32M stable and then I'll make some 4.8GHz testing with my special "ghetto cooling" (on water with open window:p: ).
Won't give the chip above 1.5 at room temp - just for the safety.;)
:up:
prime95 gets hotter than cpu tester. :)
anyways still a very nice cpu, i dont like mine.
I still try to find the 24/7 sweet spot, but i doubt i can get anything higher than 3,6ghz with acceptable temps. :(
Good going guys. :)
Hornet331: What was your stock voltage/cooling/ambient/fan setup/system setup (i.e. case) for the above ss?
Thanks.
Case TJ09 (fully stocked with fans), Cooler IFX-14 (with HR-10 cooler).
Room temp is ~24°C (maybe a bit higher).
CPU is @ stock means 333x9 (3ghz) 1,25V (drops to 1,20 under load, damn gigabyte board...)
my qx9650 is really bad, the cpu has a VID of 1,1375... i know other 9650s that have a VID of 1,050. :rofl:
i can get above 4ghz but that sucker burns (70°C + in core temp :ROTF: )
if it would be cooler in my room (20°C) im sure you would see quite a bit lower temps.
Excellent info. Thanks. :toast:
Yeah VID and starting voltage will decide what temps/oc/power you get in the end. Early ES guys had 1.1V stock volts and 1.0xx stock VID and were getting high clocks@low volts.
That 70C looks incorrect or its reading TJunction because by 64.5C Tcase (IHS) your core is automatically throttling and modulating 30-50% as a safe guard for damage. QX9650 is designed to only operate safely at a maximum of 64.5C before TCC kicks in and drops the clocks and the voltages down because its exceeded maximum safe temperature and PROCHOT gets asserted. Thats the limit before you get a damaged processor. ;) (see Intel QX9650 specs). Not sure what is happening with Core Temp there.
Also have you measured the VCC using a DMM when you set it in BIOS and after bootup?
under load the voltage is ~1,39-1,40V (bios 1,45 in windows idel 1,424)
But its only spi stable, i think would need another 0,2V to get it prime stabel but that means core temp nearly at 80°C. :p:
heres a pic of a QX9650 with VID lower than 1,1V :)
http://img293.imageshack.us/img293/7...p096ph5.th.jpg
You have to remember that the Core temp is not Tcase temp. Coretemp is the hottest spot on the die while Tcase is measured on the center of the IHS. There is no way to measure tcase (apart from modifing your IHS).
Afaik Tjuncion temp is the temp where the cpu should beginn to throttle, i think the cpu will stat with throtteling when the temp reaches nearly 100°C (its not the whole cpu, just the hottest part that reaches this temp).
i haven't measured VCC with a DMM right now, first i need to know where to measure it. :D
our chips are close
L739A637
http://i107.photobucket.com/albums/m.../prime142v.jpg
I didn't know this. I was trying to find out what its reporting and it definitely looks like its reporting Tjunction, which is the hottest part of the core and is allowed up to 105℃ for QX9650. ;)
Tjunction temp is where the CPU will shut down.
~20℃ below Tjunction is where it'll begin skipping clock cycles and throttling (85℃).
OK. Still thanks. :DQuote:
i haven't measured VCC with a DMM right now, first i need to know where to measure it. :D
I have a qx9650 with 1.125 VID.
I took it back to stock 3ghz and read 1.15v, under load 1.17v
But it takes 1.41v for 4.0, 1.43v for 4.1 and 4.7v for 4.2, 4.9 for 4.25ghz. each can run prime 2-3 hrs ft or blend. So not a low volt for 4.0 ghz, but seems to be scaling reasonably.
4.4ghz at 1.59 runs prime ok, but temps high so I dont do for long.
batch L729A646
Happy new years Camo,
Have more results for us yet?
:D Thanks mate - more will come soon. Played too much with the DDR3 ref voltage and somthing killed my WinXP (SATA!) - hangs before startup screen appears, have to reinstall now:( .
Will use a bit more moderate timings in future, I found out a certain setting of sharp timings that passes SP32M fine even cause crashes when I use LAN for Internet. Browser shuts down or system hangs, so I think the MAXIMUS EXTREME is a bit weak on the LAN rail............:yepp:
The chip is easy to handle, but board & ram require a lot of work.;)
EDIT: Really unlucky circumstances - last nite we had -8°C here - would have been perfect ambient for my "ghetto cooling" experiments (open window)..........:(
:D Now 1M faster - just relaxed the sub timings a bit:
http://img505.imageshack.us/img505/5...25ghz1mcy0.jpg
Now broke 9.8sec, but still hunting for more stability:(
you cant get it stable, even if your running it at 1.48? with that vcore you looking at a possible 4.7G 24/7 rig?! :)
is it your board, ram or cpu that's limiting you? if it's the cpu give it 1.5V and see what you'll get, it won't kill to try ;)
i hate you and i hate all the other guys with some nice silicone :D