• We've upgraded our forums. Please post any issues/requests in this thread.

My new E8400 OC 490X8=3920 @ 1.35V

dadi_oh

New Member
Joined
Jul 18, 2007
Messages
809 (0.21/day)
Likes
77
Location
Almonte, Canada
System Name Sonny Boy
Processor Q9550 3.8GHz 450X8.5 @ 1.3 Vcore
Motherboard ASUS P5Q-PRO
Cooling OCZ Vendetta 2 - lapped and MX2
Memory 2 X 2GB OCZ Reaper PC2-8500 1080MHz 5-5-5-15-2T 2.1V
Video Card(s) 5870 Eyefinity + GT 240 Physx
Storage WD Caviar Black 640
Display(s) 3X AOC 23" 1920X1080 = 5760X1080
Case HAF922 - 3X 200mm fans
Audio Device(s) Onboard - whatever
Power Supply Ultra LSP 750W
Software Windows 7 Professional 64 bit
#1
Picked up a new retail E8400 and installed it last night. Batch was Q811A115. VID is 1.1125 as reported by Coretemp.

Installed it in my P5K-E motherboard with Swiftech H20-220 watercooling. Details are in my system specs.

So my resting point at the moment is 3920MHz @ 490X8 with 1.352V load. Max temp is 47C on both cores running Prime95 small FFT's. Ran successfully overnight Prime95. Ran 3 hours cycling through 3DMark06 so video is stable @ 490MHz FSB.

After reading about CPU degradation I am nervous about exceeding 1.365V since I don't want to kill my CPU just to get a few hundred MHz out of it. Even at 1.376V I was not Prime95 stable at the magic 4000MHz mark. I think that I am on the part of the slope where it will take significantly more voltage to get stable >4GHz so diminishing returns for me. The 4GHz barrier is more of a psychological barrier than a real one. Being 80MHz short of it doesn't really bother me.

A few observations:

Although Coretemp reports a VID of 1.1125V my P5K-E seems to set the Vcore to 1.2000V at default clocks. Is that odd?

I notice that "auto" adjusts the Vcore up as you overclock the FSB. I guess they program some sort of curve into the BIOS to account for overclocking?

So watcha think? Is this about average for this CPU? Good enough... or should I set it to (little finger at corner of mouth) 1 MILLION Volts... :D
 

Attachments

Joined
Nov 3, 2007
Messages
1,567 (0.42/day)
Likes
611
Processor i5-6600K / i7-6700K / E5-1620 v3 / E5-2650L v4 / E5-2630L v3 / E5-2530 v4
Motherboard Asus Maximus IX APEX / Asrock EPC612D4I Mini
Cooling Noctua / Aquacomputer / EKWB / Corsair
Memory Corsair DP DDR4 2x16GB 3200 / Crucial LVP 2133 ECC 8x8GB
Video Card(s) R9 NANO / VEGA 64 / VEGA 56
Storage nvme's
Display(s) Acer XR382CQK
Case CaseLabs / Lian Li
Audio Device(s) Creative SB X7
Power Supply Corsair AX1500i / Corsair AX1200i / Seasonic (A handful)
Mouse Logitech
Keyboard Logitech
Software Windows 7 ULT / Windows 10 Pro / unRAID
#2
Well 1.3625 Is intels Highest recommended voltage. Whats your cooling like? Seems your water is holding up. If its good, then The extra voltage wont hurt. However, Overclocking usually does degrade form teh potential lifespan of the chip. So, its your call. But guys are running these at 4ghz all day
 

dadi_oh

New Member
Joined
Jul 18, 2007
Messages
809 (0.21/day)
Likes
77
Location
Almonte, Canada
System Name Sonny Boy
Processor Q9550 3.8GHz 450X8.5 @ 1.3 Vcore
Motherboard ASUS P5Q-PRO
Cooling OCZ Vendetta 2 - lapped and MX2
Memory 2 X 2GB OCZ Reaper PC2-8500 1080MHz 5-5-5-15-2T 2.1V
Video Card(s) 5870 Eyefinity + GT 240 Physx
Storage WD Caviar Black 640
Display(s) 3X AOC 23" 1920X1080 = 5760X1080
Case HAF922 - 3X 200mm fans
Audio Device(s) Onboard - whatever
Power Supply Ultra LSP 750W
Software Windows 7 Professional 64 bit
#3
Well 1.3625 Is intels Highest recommended voltage. Whats your cooling like? Seems your water is holding up. If its good, then The extra voltage wont hurt. However, Overclocking usually does degrade form teh potential lifespan of the chip. So, its your call. But guys are running these at 4ghz all day
I think the rules have changed with the 45nm chips. It used to be that thermal would kill you before voltage so you had some warning signs. With the new technologies used with the 45nm process (new metals and processes) it seems voltage can kill you even when you are cool.

I am watercooling and hitting about 47C on the cores under Prime95 small FFT's so thermal is not an issue at all. Voltage might not be an issue either but, as I said, I will err on teh safer side until there is more data on the 45nm side. There have been enough reports of dead E8xxx chips over 1.4V to make me think twice. Maybe I'm being paranoid... :twitch:
 

dadi_oh

New Member
Joined
Jul 18, 2007
Messages
809 (0.21/day)
Likes
77
Location
Almonte, Canada
System Name Sonny Boy
Processor Q9550 3.8GHz 450X8.5 @ 1.3 Vcore
Motherboard ASUS P5Q-PRO
Cooling OCZ Vendetta 2 - lapped and MX2
Memory 2 X 2GB OCZ Reaper PC2-8500 1080MHz 5-5-5-15-2T 2.1V
Video Card(s) 5870 Eyefinity + GT 240 Physx
Storage WD Caviar Black 640
Display(s) 3X AOC 23" 1920X1080 = 5760X1080
Case HAF922 - 3X 200mm fans
Audio Device(s) Onboard - whatever
Power Supply Ultra LSP 750W
Software Windows 7 Professional 64 bit
#4
Can you do me favour and see how low you can get your temps. So basically, what id do is set it at 200x6 with 0.8v and measure at idle. Also max out all fans.

If your diodes arent stuck, you should be at ambient temps for those speeds.
I did the Realtemp calibration procedure when I was first setting it up. I think one of my diodes may be stuck at the lower end. I don't have the numbers in front of me right now but with the system set to 266X6 and voltage at 1.1V (lowest setting in my BIOS) Core 0 was at about 22C and Core 1 was at 29C. Room abient was about 22C. I tried a correction factor of -2 on Core 1 and now it is close to Core 0 on the idle reading (although the distance to Tj max is still uncorrected.. which makes sense).

Under load both cores are exactly the same temp so this sounds like my Core 1 is sticking when going to idle?

If it is only the idle temps that are affected then I'm not too worried about it. Load temps are what I care about.

By adjusting the idle correction factor I assume it does not affect the load temps for Core 1?