e2180 and IP35-E OC'ing
- Thread starter soccerballtux
- Start date
You mean your vCore when running at stock speed is 1.2875v? I assume you got the number from CPUz? If so, then IP35E's vDroop caused the actual BIOS vCore to be about 1.30v. So that's pretty much the same stock vCore I have on my E4500. The BIOS vCore, or VID, is the official numbers we go by. CPU-z always show results after vDroop, this is inevitable due to power draw of processor. Do a thread search on my previous thread on factory vCore on E2xxx and E4xxx.
You should be able to hit 2.9 right away at standard 1.30v, and may be even 3.0Ghz Orthos stable. I have it stable at 3.2Ghz @ 1.38v, and then 3.3Ghz @ 1.42v. You can keep pushing about 80-100mhz per 0.04vCore until you hit 3.4 or so and I think it'll be alot harder to get it to 3.5Ghz unless you have good cooling. It'll hit 3.6Ghz max at around 1.65vCore, but that's really heating it up.
You should be able to hit 2.9 right away at standard 1.30v, and may be even 3.0Ghz Orthos stable. I have it stable at 3.2Ghz @ 1.38v, and then 3.3Ghz @ 1.42v. You can keep pushing about 80-100mhz per 0.04vCore until you hit 3.4 or so and I think it'll be alot harder to get it to 3.5Ghz unless you have good cooling. It'll hit 3.6Ghz max at around 1.65vCore, but that's really heating it up.
CPU-z reports 1.248 at full load :/
Strange.
But my temps are crazy I think: Coretemp is reporting 57C max, however Speedfan (beta 4.34v37) is reporting 42C max. I thought the Speedfan 15C too low was fixed?
Strange.
But my temps are crazy I think: Coretemp is reporting 57C max, however Speedfan (beta 4.34v37) is reporting 42C max. I thought the Speedfan 15C too low was fixed?
BlueAcolyte
Platinum Member
- Nov 19, 2007
- 2,793
- 2
- 0
The latest beta is right, what coretemp are you using? 0.95 works for me. The VCore drop is supposed to happen; it's called VDroop. Anandtech did an article on it some time ago.
the ip35-e I have also has enormous vdroops. I think it's the way it's built. 1.345 BIOS setting becomes 1.288v at load.
So does this mean I can push higher voltages in the mobo safely? Or that I need to keep them down?
It's designed as a safety feature so the CPU doesn't get overloaded between idle and load. Think of it as buffer zone for CPU voltage. Theoretically you can push the vCore higher if the vDroop causes the actual voltage (as reported by CPU-z) to be lower than what it appears in BIOS, but it's not that big of a difference, we're talking only 0.05v max. Considering Intel spec says 1.35v VID max for E2xxx and E4xxx processors, I wouldn't go higher than setting BIOS VID 1.48v for normal everyday use, that's about Intel spec +10%. With 1.48v you should get 3.3Ghz stable, but might heat it up some.
By the way, the latest CoreTemp 0.96 is accurate, many other temp readers are 15c too low due to Intel adjusted Tjunction to 100c for M0 chips. This also means that Intel felt their M0 chips are more tolerant to heat, which is the reason they raised the max by 15c.
By the way, the latest CoreTemp 0.96 is accurate, many other temp readers are 15c too low due to Intel adjusted Tjunction to 100c for M0 chips. This also means that Intel felt their M0 chips are more tolerant to heat, which is the reason they raised the max by 15c.
Also - if you want dead on voltage readings don't trust that software or any software for that matter. It will give you a very good idea as to what it is but its usually off by some degree or another. A DMM is about the only way to get a for sure reading from the contact points on the mobo (don't try with out research)
My board used to have a fair bit of vDroop until I did a pencil mod to stabilize it and now its all but 100% gone. But as stated above basically every board has some amount of vDroop.
My board used to have a fair bit of vDroop until I did a pencil mod to stabilize it and now its all but 100% gone. But as stated above basically every board has some amount of vDroop.
Could the BIOS temp be wrong? Because at 6x mult, CT reports a higher idle temperature than in the mobo BIOS when mult is at 10x. Which means if CT is right, my processor would get to 105C before it gets shut off by the motherboard (default cutoff at 90C)?
AbitEQ and Speedfan Beta37 both correspond with BIOS readings. I'm having trouble accepting that CT is correct, especially seeing as I have a great case (CM690) and a Thermalright Ultra120 with a 120mm fan blowing straight towards the fan in the back of the case blowing out. The air is just not that hot coming out either.
On a side note, I can get to 3.4Ghz on 1.48v (BIOS) but Orthos Small brings temps to 70C (85C CT but that just can't be right). I guess if I'll be getting a quad when Penryn goes EOL I might as well run it fast as I can; I just don't like jumping 0.1v just for an extra 200Mhz.
AbitEQ and Speedfan Beta37 both correspond with BIOS readings. I'm having trouble accepting that CT is correct, especially seeing as I have a great case (CM690) and a Thermalright Ultra120 with a 120mm fan blowing straight towards the fan in the back of the case blowing out. The air is just not that hot coming out either.
On a side note, I can get to 3.4Ghz on 1.48v (BIOS) but Orthos Small brings temps to 70C (85C CT but that just can't be right). I guess if I'll be getting a quad when Penryn goes EOL I might as well run it fast as I can; I just don't like jumping 0.1v just for an extra 200Mhz.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,570
- 10,199
- 126
BIOS readings are TCase, not Tjunction. Tjunction is usually 10-15C hotter than Tcase. If AbitEQ and SpeedFan give temps for the core (Tjunction) that are equal to the BIOS (Tcase), then that indicates that they are wrong and off by 15C. Hence Coretemp is, IMHO, correct.
Screw it I'm going to cook this for all it's worth. WTF did I get a good cooler for anyhow if my OC temps are 75 lol. I simply don't believe it.
Update:
Per http://forums.abit-usa.com/showthread.php?t=128969
It appears SPEEDFAN b37 has the CORRECT temps and that CT 0.95.4 does as well. CT 0.96 is 15C too high, however.
Per http://forums.abit-usa.com/showthread.php?t=128969
It appears SPEEDFAN b37 has the CORRECT temps and that CT 0.95.4 does as well. CT 0.96 is 15C too high, however.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,570
- 10,199
- 126
There's a lot of contradictory nonsense in that thread.
First he says:
Then he says:
CT 0.96 is correct, IMHO - assuming that these chips are in fact 100C TJmax, and Intel's specified manner of reading the DTS is correct.
First he says:
So if the TJmax is read (incorrectly) as 85C, then temps need to be adjusted +15C. Okay.
Then he says:
The readings that were 15C too low, are actually the correct ones? And the ones with the correct offset are wrong? WTF?
CT 0.96 is correct, IMHO - assuming that these chips are in fact 100C TJmax, and Intel's specified manner of reading the DTS is correct.
CT is right, soccerball tux. I'll betcha you're ultra 120 isnt seated well, especially since you said your cooking upwards of 70-80 degrees, and the sink blades don't feel warm to the touch. try lapping the base of the ultra 120, mark it first. somebody here (i dont have link) just lapped his TRultra 120, and he idles at 29-30 now which was down almost 18 degrees from his previous seating of the cooler. i bet if you're lucky and you wanna "cook it" you could get 3.6 out of it, just lap & reseat the cooler with some as5. use a sharpie to mark the 120's base and see where your marks rub off. see if the cooler base is convex, concave, or just indirectly not flat.
reinstall it, and lube it up good with some as5 and you're on your way to 3.6ghz +
edit: heres one that was concave
http://forums.anandtech.com/me...=2140942&enterthread=y
reinstall it, and lube it up good with some as5 and you're on your way to 3.6ghz +
edit: heres one that was concave
http://forums.anandtech.com/me...=2140942&enterthread=y
Ok I've lapped it; ended with 2000 and a mirror finish.
No improvement to temps.
Maybe I should lap my CPU too, however lapping the Ultra should have made a difference.
Orthos still (wrong) at 71C.
I'm going to try 3.4Ghz again and see if I still hit "85".
No improvement to temps.
Maybe I should lap my CPU too, however lapping the Ultra should have made a difference.
Orthos still (wrong) at 71C.
I'm going to try 3.4Ghz again and see if I still hit "85".
wouldnt lap the cpu...
did you do a razor test? marker test? on the sink base?
you can do the cpu if you know what you're doing, but i wouldn't personally. ohh, make sure you mount the thing correctly, it has 2 or 3 fins at the bottom that point the airflow downwards towards your power phase circuitry, and also make sure your fan is running at sufficient rpms, and that its exhausting air from the mem modules towards the rear of the case, and then in turn is exhausted by the rear 120mm case fan. I'm assuming you have good case air-flow from the front, and decent cable management and ambient temps.
what type of thermal compound did you use?
did you do a razor test? marker test? on the sink base?
you can do the cpu if you know what you're doing, but i wouldn't personally. ohh, make sure you mount the thing correctly, it has 2 or 3 fins at the bottom that point the airflow downwards towards your power phase circuitry, and also make sure your fan is running at sufficient rpms, and that its exhausting air from the mem modules towards the rear of the case, and then in turn is exhausted by the rear 120mm case fan. I'm assuming you have good case air-flow from the front, and decent cable management and ambient temps.
what type of thermal compound did you use?
AS5;
Fan RPM is 1100; perhaps might get a Scythe higher CFM higher RPM fan and see how that helps.
Temps are...just like they were before at 3.4Ghz. 85C/70C depending on if you think CT 0.96 is right.
Btw, for stressing to test for stability should I be doing Orthos small or large? Large is 2x 1mB; seeing as my cache is only 1MB total I'm thinking Small would be more stressful since the cache isn't getting a break while fetching from RAM. Temps seem to reflect this by about 3-4C.
Oh, and ya the sink blades def do feel warm; just not...almost boiling water warm.
This is only 1.448V load after Vdroop.
As long as I can get ~2+ years out of this (right now 72C for orthos stable at 3.4Ghz) I'll be happy.
Fan RPM is 1100; perhaps might get a Scythe higher CFM higher RPM fan and see how that helps.
Temps are...just like they were before at 3.4Ghz. 85C/70C depending on if you think CT 0.96 is right.
Btw, for stressing to test for stability should I be doing Orthos small or large? Large is 2x 1mB; seeing as my cache is only 1MB total I'm thinking Small would be more stressful since the cache isn't getting a break while fetching from RAM. Temps seem to reflect this by about 3-4C.
Oh, and ya the sink blades def do feel warm; just not...almost boiling water warm.
This is only 1.448V load after Vdroop.
As long as I can get ~2+ years out of this (right now 72C for orthos stable at 3.4Ghz) I'll be happy.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,570
- 10,199
- 126
Except TAT hasn't been updated to deal with the new steppings. I have no idea about the accuracy of Everest.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 23K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 18K
-
-
-
Facebook
Twitter