Highest recommended voltages for 24/7

[hedge]

What is the highest voltage you guys would recommend running an Intel 2.4c and 2 sticks of HyperX 3500? Right now the cpu is at stock voltages and the ram has 3.15v (need to mode the psu for more). How much higher could I go without instantly killing them?

[hedge]

Bump...

Anytime fellas, don't be shy. ;)

[hollywood]

CPU: no higher than 1.6v for 24/7

Memory: You're near max...keep it there

AGP: 1.8v is ok

[Rio71]

1,70v for the cpu should be ok.
my rigs running 24/7 seti, all northwood p4/c4 @ 1.70v.

jepp, the mem is near max. but you have bh-5 go higher.

[speedstream5621]

As for the CPU...he has and AMD. They themselves have stated that it is safe to increase the voltage .5v.

It is very different for Intels, however, with their SNDS.:p:

I don't know how much juice I would pump through the RAM though...2.9v is good enough for me.:eek:

[Rio71]

SNDS was only for the B0 stepping. :stick:
:D

[speedstream5621]

Originally posted by Rio71
SNDS was only for the B0 stepping. :stick:
:D
What was the max voltage for it anyway? I want to say 1.8v, but for some reason that doesn't sound right...:confused:

[neyoung]

Originally posted by Rio71

jepp, the mem is near max. but you have bh-5 go higher.

Those hyperX sticks are not for sure bh-5. I have 4 sticks of hyperx and none of them are bh-5 :(

[STEvil]

3.15 for ram will do you fine for 24/7 use.

2.0v is fine for AMD's.

[Semper Fi]

I have my 2.4c @ 1.725v, my temps @ full load are 41c, does power mater as long as temps stay low?

Ram is also at 3.2, too much for 24/7?

[xDUCK]

Originally posted by Semper Fi
I have my 2.4c @ 1.725v, my temps @ full load are 41c, does power mater as long as temps stay low?

Ram is also at 3.2, too much for 24/7?

Voltages matter even if your temps are low - for example, even if you're on LN2, putting 2.5v through a CPU (especially a P4) will kill it. No questions asked. :) That having been said, 1.725v should be fine (as long as SNDS is no longer a concern with P4's - can anybody with more Intel experience that I confirm this? [I'm more of an AMD guy myself :)]).

That RAM is running at pretty much the safe limit. Some people run their ram much higher 24/7 (Macci and his 3.6v HyperX 3000 comes to mind), but it's definately not safe for the vast majority of memory. Of course, these suckers carry lifetime warranties (unless you bought generic RAM :stick: ), so it's not THAT big of a concern. ;)

[Semper Fi]

Not tryin to hijack this thread, but whats a safe agp voltage on a 9800pro with watercooling and a sidepanel 190cfm delta :D

[saaya]

didnt some people loose their 9800s with 1.6-1.7v agp voltage?

[Creative]

Ive run my agp at 1.65 on every board for the past 4-5 months :thumbsup:

Ive also been running my P4Cs at 1.70 for mosly 24/7 and even 1.80 for benching....:)

[mrzee]

I run my AMD setup at 1.9vc, 3.2vdimm,2.0 nb,1.8vagp.
Been this way for over a year, never a problem.
9700pro core@1.85, mem@3.3, of course the whole
setup is water cooled.
XP1700 running at 11.5x227@2615mhz, NF7-S w/
vdimm and vdd mods, 2x256 Hyper X 3500.
One of the first things I do with a new setup is
run the vagp up to the maximum.
I've been through 5 9800pros on my main rig,
all due to extreme volt mods , excessive cooling
and hard benching.Thank God for Comp USA's
return policy.:D

[mrzee]

Originally posted by hedge
What is the highest voltage you guys would recommend running an Intel 2.4c and 2 sticks of HyperX 3500? Right now the cpu is at stock voltages and the ram has 3.15v (need to mode the psu for more). How much higher could I go without instantly killing them?

If you're on air, like Hollywood says, 1.6vc max for everyday use.
Hyper X loves voltage, unless you have a vdimm mod on your board you can't give it too much. I have ran my Hyper X as high as
3.3v for days at a time without any problems.

Semper Fi, run your vagp at the max, no worries, it won't hurt your 9800pro.

[Narbo]

The Intel datasheet for the the P4 states no more then 1.7V should ever reach any pin of the CPU. For 24/7 air I would suggest following thier guideline and keep it below 1.7V. After setting a few systems up for friends and my own I find that 95% of the P4C overclock potential can be reached at below 1.6V. Sure you can get another 100-200Mhz from that last little bit but the heat and power dissapation really start to skyrocket.

[hedge]

Thanks guys. This is just the information I need for when my maze4 shows up. I believe that this chip has something to it. My hyperx may be bh-5, it may not though. Right now it won't go over 240 with 3.15v because it will fail in prime95 and memtest. I am still burning them in to see if there is more.

[FlyingHamster]

Hi guys,

I use all air cooling with my setup, so what do u think would be a safe agp voltage to set my 9800 to run at 24/7? should I just leave it at 1.5 or increase it? does increasing allow a better OC basically? Oh, and by how many increments in my NF7-S bios does it allow me to increase it?

Thnx in advance..

[Semper Fi]

Now if I could just work up the balls to vmod the memory on my NF7-S, we'd be in business :D

[hatemi]

I wouldn`t recomend uppig the VAGP too much! I`m pretty sure I fried two 9000PROs at 1.8v on my EP-4PDA2+

[mrzee]

Raising vagp gives better stability when overclocking.
For what it's worth, I run all my rigs at max vagp 24/7.
As I said earlier, I've killed 5 9800 pro's on my benching rig,
due to extreme vmods, excessive cooling and hard benching,
but my other 2 rigs, both overclocked, have run at max vagp for over a year, no problems.

[Silversink]

Originally posted by Rio71
SNDS was only for the B0 stepping. :stick:
:D

Show me proof of that ;)

[FlyingHamster]

Originally posted by mrzee
Raising vagp gives better stability when overclocking.
For what it's worth, I run all my rigs at max vagp 24/7.
As I said earlier, I've killed 5 9800 pro's on my benching rig,
due extreme vmods, excessive cooling and hard benching,
but my other 2 rigs, both overclocked, have run at max vagp for over a year, no problems.

wut vagp are u running them at 24/7 now?

[mrzee]

I have two NF7-S boards, 1 rev.1, 1 rev.2,
both are running at 1.8vagp, which is the
maximum on an NF7-S.

The IC7 Max 3 is running at 1.6, which is
the max also.

[FlyingHamster]

oh ok. but u also have a prommie too, lol. wut do u recommend for me, since I only have air cooling (28 C case temp).

[mrzee]

Prommie runs on my Intel. Every AMD system I've built, and I've built dozens, if they are overclocked they run at MAX vagp.
IMO, the idea that high vagp burns cards is a myth. I have
never seen it happen in the three years that I have been overclocking.
You should run your vagp at whatever you feel comfortable with,
if it were me I'd be running at 1.8v. Try different settings, you
probably won't notice any difference, but higher settings will give you more stability when overclocked, especially when benching.

[BBThumper]

.......maybe you guys can help me. Whenever I raise my vcore past 1.6v my rig won't post :confused: . Cooling is fine PSU rails are great so I don't get why the p4 won't boot over 1.6v. Any ideas?..............

[hedge]

I know that certain AMD chips have problems at high voltages like 1.9v, but I have never heard of an intel chip doing that.

[N00B pro]

Wow wtf are u guys talking about???? 1.6 vcore on amd cpu is considered max safe???? Rofl........are u guys kidding me>>> I run my 1700xp 12x200fsb ,2400mhz at 1.8vcore and my temp never hit 50c on air. So who ever said 1.6 is max safe temp is 100% too paranoid. These amd cpu's can tolarate even 2vcore without damagin it as long as you use thermarright slk 900 heatsinks....1.6vcore max safe rofl, what a newbish advice. Get rid of that stock heatsink u got buddy. lol

[BBThumper]

Originally posted by N00B pro
Wow wtf are u guys talking about???? 1.6 vcore on amd cpu is considered max safe???? Rofl........are u guys kidding me>>> I run my 1700xp 12x200fsb ,2400mhz at 1.8vcore and my temp never hit 50c on air. So who ever said 1.6 is max safe temp is 100% too paranoid. These amd cpu's can tolarate even 2vcore without damagin it as long as you use thermarright slk 900 heatsinks....1.6vcore max safe rofl, what a newbish advice. Get rid of that stock heatsink u got buddy. lol

........did you read this P4 thread or did you just put your foot directly in your mouth :D ................

[bmg]

Originally posted by BBThumper
.......maybe you guys can help me. Whenever I raise my vcore past 1.6v my rig won't post :confused: . Cooling is fine PSU rails are great so I don't get why the p4 won't boot over 1.6v. Any ideas?.............. I notice that you're using the P4C800-E mb. What I've found with the two P4C800-E boards I've had is that the 1.625V setting actually gives lower voltage than 1.60V. The 1.65V setting is close to 1.60V, but might still be slightly lower. You need a 1.70-1.725V bios setting to be noticably better than at 1.60V. Under load the 1.70V setting only increases the cpu voltage about .03V relative to the 1.60V setting (you might do a little better than that depending on the frequency your processor is running at). If your processor is pushed to the max at 1.60V then you'll definitely have problems with the 1.625V setting, and likely the 1.65V setting. At 1.60V and below the idle voltage will be around .06V higher than the setting. Above 1.60V the idle voltage is close to the bios setting. This all applies to the Rev 1.02 boards. Rev 2.0 might be different.

At a 1.70V bios setting I get an idle voltage of right at 1.70V. This is pushing what I'm comfortable with 24/7 (my prommy cooled cpu temperature also drops to a reported temp of close to -40C). I run a folding screensaver that kicks in after 5 minutes, which pulls the voltage down to 1.60v or a bit lower (and increases the cpu temperature to around -23C), so I'm very comfortable with the 1.70V bios setting.

[FlyingHamster]

Originally posted by mrzee
Prommie runs on my Intel. Every AMD system I've built, and I've built dozens, if they are overclocked they run at MAX vagp.
IMO, the idea that high vagp burns cards is a myth. I have
never seen it happen in the three years that I have been overclocking.
You should run your vagp at whatever you feel comfortable with,
if it were me I'd be running at 1.8v. Try different settings, you
probably won't notice any difference, but higher settings will give you more stability when overclocked, especially when benching.

ok, thnx. Im still curious as to why ppl are saying that they fried their 9800's due to raising the vagp. if it wasnt that, then what else could have fried it.. thats my question.

anyways, so when I raise the vagp, its going to raise my case temps as my card will get hotter right?

[BBThumper]

Originally posted by bmg
I notice that you're using the P4C800-E mb. What I've found with the two P4C800-E boards I've had is that the 1.625V setting actually gives lower voltage than 1.60V. The 1.65V setting is close to 1.60V, but might still be slightly lower. You need a 1.70-1.725V bios setting to be noticably better than at 1.60V. Under load the 1.70V setting only increases the cpu voltage about .03V relative to the 1.60V setting (you might do a little better than that depending on the frequency your processor is running at). If your processor is pushed to the max at 1.60V then you'll definitely have problems with the 1.625V setting, and likely the 1.65V setting. At 1.60V and below the idle voltage will be around .06V higher than the setting. Above 1.60V the idle voltage is close to the bios setting. This all applies to the Rev 1.02 boards. Rev 2.0 might be different.

At a 1.70V bios setting I get an idle voltage of right at 1.70V. This is pushing what I'm comfortable with 24/7 (my prommy cooled cpu temperature also drops to a reported temp of close to -40C). I run a folding screensaver that kicks in after 5 minutes, which pulls the voltage down to 1.60v or a bit lower (and increases the cpu temperature to around -23C), so I'm very comfortable with the 1.70V bios setting.

........I'll try higher but I still don't get why it won't post at 1.62v or 1.65v but I haven't tried any higher than that yet. Thought maybe I missed a bios setting or something.........

[mrzee]

Originally posted by BBThumper
........did you read this P4 thread or did you just put your foot directly in your mouth :D ................
Foot in mouth, definately.
Start on page one, read whole thread.
No problem, you're right 1.6vc is very conservative on AMD.
This did start out as a P4 thread though.

[mrzee]

Originally posted by FlyingHamster
ok, thnx. Im still curious as to why ppl are saying that they fried their 9800's due to raising the vagp. if it wasnt that, then what else could have fried it.. thats my question.

anyways, so when I raise the vagp, its going to raise my case temps as my card will get hotter right?

I doubt you will see any rise in temps from raising the vagp.
I never have (although I'm watercooled).
Your card won't run any hotter either.

[BBThumper]

.........ok got it to post at 1.7v - thk you. Now to get it to pass prime :( ..............

[bmg]

Originally posted by BBThumper
........I'll try higher but I still don't get why it won't post at 1.62v or 1.65v but I haven't tried any higher than that yet. Thought maybe I missed a bios setting or something......... Because the actual cpu voltage at a bios setting of 1.625V (and usually 1.65V) is actually lower than a setting of 1.60V. The cpu supply voltage controller actually only handles up to 1.60V. Asus is apparently adding some kind of offset to get 1.625V and above. A bios setting of 1.60V and below overvolts under light load. Above that the idle voltage doesn't overvolt, it's close to the setting.

[BBThumper]

......at 1.7v I am at 1.63v load and 1.71v idle. Would be a great mobo except for this problem. Then again we don't have divider problems :) .................thk for the help guys.........

.........1 last question. ICH Delayed Trans=??? is this the same as PCI Delayed Trans? Sorry to get off topic...........

[st0nedpenguin]

****EDITTED****
-KnightElite

[BBThumper]

.......well it looks like I just have a chip that won't do over 280fsb no matter what volts :( . This is when I miss my AMD Chips. They were only like 90$ to replace but $160 is alot harder to justify........

[ns_ripper]

raising the agp voltage will help sh*t.

I see many people who think that it does help. Well, depending on wich voltage is fed to the card's onboard voltage regulators, you can raise them a bit to make sure that the voltage regulator is receiving the voltage it need's.

But raising the agp voltage with more then 0.1V can't make a difference. The only thing it does would be heating the onboard voltage regulators of the card some more. the voltage regulators will pull down the voltage to it's desired voltage level.

when you also vmod you're card it's a different story.

[N00B pro]

****EDITTED****

Post was reported for offending someone. Please try to avoid insulting other members, and using offensive language. That goes for st0nedpenguin as well.
-KnightElite

Well after reading this thread again, many amd users also posted their voltages on their amd rigs so that's what got me confused and thinking that people were saying 1.6v is highest on amd cpu, stupid confusion.

[st0nedpenguin]

****EDITTED****

/me runs from the admin hand of lockage...

Better run fast.... :cool: - KnightElite

[bigballaga]

just because the TEMPS dont get high,, DOES NOT mean that it is safe to run a lot of voltage into the chip !!!!!!!

[ns_ripper]

Originally posted by bigballaga
just because the TEMPS dont get high,, DOES NOT mean that it is safe to run a lot of voltage into the chip !!!!!!!

I'll agree 100% with that statement :)

chips don't have to be hot for electromigration to happen. Heat only encourages it to happen faster

[neyoung]

what exactly is electromigration?

[STEvil]

Physical movement of materials resulting in disconnected bridges and such.

[neyoung]

Tthat doesn't sound good. So basicly if you run alot of voltage through your cpu for example, the voltage can move the materials in the cpu around and connect bridges and disconnect bridges here and there, which screws up the path of electricity and makes the cpu unstable or not work at all.

[ns_ripper]

Originally posted by neyoung
Tthat doesn't sound good. So basicly if you run alot of voltage through your cpu for example, the voltage can move the materials in the cpu around and connect bridges and disconnect bridges here and there, which screws up the path of electricity and makes the cpu unstable or not work at all.

well, everything is physics.

when something heats up, it grows. in a cpu core there are a few milions of transistors and stuff. You have to know that there are certain hotspots in cpu (parts that get hotter then others). So these parts can grow a bit bigger (big is relative when you are talking about nm processes :D ) so you can imagine that when a few transistors grow, they haven't got enough room and they will literally visit their neighbour transistors.

it's only a small part of the things that do occur and it's may be explained a bit childish but I think that everyone can catch my drift. ;)

[BBThumper]

....this might not be easy to answer but why is the volts more dangerous on p4 then a amd chip?..........

[ns_ripper]

Originally posted by BBThumper
....this might not be easy to answer but why is the volts more dangerous on p4 then a amd chip?..........

no one can answer you that. We just know it out of experiance.

It would take a chipdesigner to tell us why, and even then ...
no outsider can know how intels chips are designed. it's kept secret. but I guess that it has something to do whith the transistors for sure. and that can depend on many many physical and even chemical things. for example;

wich materials are being used for polution to create the semi-conductance, ...

no one can tell you.

[FlyingHamster]

Originally posted by ns_ripper

when you also vmod you're card it's a different story.

is that hard to do? got any guides?

[KnightElite]

Originally posted by ns_ripper
no one can answer you that. We just know it out of experiance.

It would take a chipdesigner to tell us why, and even then ...
no outsider can know how intels chips are designed. it's kept secret. but I guess that it has something to do whith the transistors for sure. and that can depend on many many physical and even chemical things. for example;

wich materials are being used for polution to create the semi-conductance, ...

no one can tell you.

As he says, it's pretty hard to say exactly why, but it is likely in the actual design of the gate on the MOS transistors. AMD appears to make theirs more robust than Intel's, so when overvolting Intel CPUs, it is easier for electrons to tunnel through the gate to the main transistor channel and cause a short. This is purely speculation, however, and I have no data to back it up ;).

[st0nedpenguin]

Originally posted by st0nedpenguin
****EDITTED****

/me runs from the admin hand of lockage...

Better run fast.... :cool: - KnightElite

LMAO!!!

:D

[BBThumper]

......thks for the info :) ................