• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

[H] users 7950 Overclock Results

Just got an MSI Twin Frozr 7950..

ASIC quality 90%
Does the voltage in windows differ from the voltage used while gaming?
 
ironforge said:
Just got an MSI Twin Frozr 7950..

ASIC quality 90%
Does the voltage in windows differ from the voltage used while gaming?
Click to expand...


There is usually a slight voltage drop under load during gpu processing.
 
Finally got the chance to mess around with the 7950 950mhz edition that SonDa5 recommended.

Currently on stock hsf, it is running at 1075/1600 @ 1.062v. ASIC 86.6%.

Going to order a water block for it sometime next week..hopefully I can up the core to 1300 24/7.
 
anthongy817 said:
Finally got the chance to mess around with the 7950 950mhz edition that SonDa5 recommended.

Currently on stock hsf, it is running at 1075/1600 @ 1.062v. ASIC 86.6%.

Going to order a water block for it sometime next week..hopefully I can up the core to 1300 24/7.
Click to expand...

1300mhz, not likely.
If you get 1200 I would say its good.
lower asics "tends" to do better on water.
 
SonDa5 said:
There is usually a slight voltage drop under load during gpu processing.
Click to expand...

While in windows the card is drawing .85v with a max in game 1.13v. using hwmonitor.

running at 1000 core, 1375 mem seems alright with stock volts. Temps are in the usually in the 60's after a couple hours of gaming.
 
Last edited:
ironforge said:
Just got an MSI Twin Frozr 7950..

ASIC quality 90%
Does the voltage in windows differ from the voltage used while gaming?
Click to expand...

Yes it does. Your higher ASIC rating than mine means your voltages will be even lowe than mine. I wouldn't be surprised if you can hit 1100/1575 with a low as ~1.060v. That's a harmless voltage, a toothless piranha. Based on my testing, everything up 1.150v is very safe.

Few results. Keep in mind that my computer costs less than $550 without the video card, it was meant to be a low budget HTPC. My i5 2400 and PCIE 2.0 (to a much lesser degree) are obvious bottlenecks.

Fully stock, 880/1250/1031v run 57C load:

http://www.3dmark.com/3dm11/4107388


2a6v6v9.jpg



+20 power control, 1000/1350/1031v 59C load:

http://www.3dmark.com/3dm11/4107415



n2d8ac.jpg



1100/1575/1075v 68C load:

http://www.3dmark.com/3dm11/4107460


hrlv91.jpg



1150/1850/1149v 77C load:

http://www.3dmark.com/3dm11/4107514


3500y7q.jpg




1200/1865/1162v 71C load, fans were set to 100% before and during load:

http://www.3dmark.com/3dm11/4107624


sfh1tw.jpg


1252/1851/1193v 74C load:

http://www.3dmark.com/3dm11/4108168

mrv9lx.jpg
 
anthongy817 said:
Finally got the chance to mess around with the 7950 950mhz edition that SonDa5 recommended.

Currently on stock hsf, it is running at 1075/1600 @ 1.062v. ASIC 86.6%.

Going to order a water block for it sometime next week..hopefully I can up the core to 1300 24/7.
Click to expand...



Are you running stock voltage? My card could do 1075/1600 with stock voltage and stock hsf. If you are able to do the same then there is a good chance your card could reach 1300/1600 with around 1.25v. Wait for the water block to keep everything as cool as possible..
 
I have two Visiontek 7950s in Xfire. At stock (800/1250/1.03v) and auto fan, they barely break 60C. I can run 965/1500 stock volts fully stable, set fans to 65% and GPU2 sits about 65C, GPU1 70C. Tried 1000 core, and it would lock up with Heaven, tried 990, same thing. Tried up to 1.2V and 80% fan and it still locks up at the same point in Heaven everytime. Have a NZXT 850w HALE82 PSU, so its not that. Power is set to +20.
Any ideas?


EDIT: Disabled Xfire, first card can do 1k with stock volts, and 1100Mhz with 1.125v, and mem at 1575. Max that AB normally allows, will extend, and see how much further it goes. Either other card is a dud, or Xfire is having issues.
 
Last edited:
Just got a MSI TF3 7950 (ASIC 88.3%)-I maxed out the sliders at 1150/1575 in Afterburner and got the following result. Does anybody know how to boost the clocking limits in Afterburner? I'm able to go higher using TRIXX, but I greatly prefer Afterburner over it. I'm running an aggressive fan profile in Afterburner that kicks the fan up to 40% @ 50C, 60% @ 55C, and 70% @ 65C. Max temp I've hit in Furmark at 1.2v is 71C.

1150/1575/1200mv 58C load

UbUo0.png
 
On air -- I have some decent airflow in my case plus ~70-72*F ambient temps usually. I did change the TIM on this card as well which helped a lil bit. :)
 
NVM, I just noticed 73C max in your second screenshot which is more in line with what I was expecting. I need to replace TIM on my card as well-was removing the hsf an easy process? At 1200/1850 I'm hitting just under 80C load in furmark, rather toasty for my liking.
 
Hey guys, been reading these forums and I am REALLY envious of the clocks that were achieved here.
So I decided to try to oc my sapphire oc 950mhz edition card (it came stock overclocked to 950mhz) I noticed although my gpu temp is cool (70c) for a 1150core/1500mem overclock on 1.14v, my vrm temps were off the charts, around 90C to 115C max. At that time, it would either throttle the core up and down from 1150 to a very low 501mhz, or would most of the time, black out the monitor. I tried putting down the clocks even back to stock but it would still keep constantly throttling the clocks and the gpu usage would fluctuate quickly. I noticed the vrm temps would still be uncomfortably high, 80 to 100 even at stock. Is this normal? I assume not since most if not all overclocks ive seen in this thread has upwards to voltages ranging from 1.1v and higher even 1.2. For me, anything above stock voltage of 1.05v temperatures are uncomfortably high and at 1.8v and above that my card would already crash in games and Furmark/OCCT/EVGA oc scanner. Can someone give me vrm 1 and 2 temps w/ sapphire cooler on it? Im just wondering if my graphic card is just a shitty overclocker or sapphire dual-x coolers just are not good at cooling vrm since that vrms are the most influential part to getting high voltages and getting stable high overclocks. i have an asic quality of 87.1% Thanks!
 
Insanitic said:
Hey guys, been reading these forums and I am REALLY envious of the clocks that were achieved here.
So I decided to try to oc my sapphire oc 950mhz edition card (it came stock overclocked to 950mhz) I noticed although my gpu temp is cool (73) for a 1150core/1500mem overclock on 1.14v, my vrm temps were off the charts, around 90C to 115C max. At that time, it would either throttle the core up and down from 1150 to a very low 501mhz, or would most of the time, black out the monitor. I tried putting down the clocks even back to stock but it would still keep constantly throttling the clocks and the gpu usage would fluctuate quickly. I noticed the vrm temps would still be uncomfortably high, 80 to 100 even at stock. Is this normal? I assume not since most if not all overclocks ive seen in this thread has upwards to voltages ranging from 1.1v and higher even 1.2. For me, anything above stock voltage of 1.05v temperatures are uncomfortably high and at 1.8v and above that my card would already crash in games and Furmark/OCCT/EVGA oc scanner. Can someone tell me their vrm temps w/ sapphire cooler on it? Im just wondering if my graphic card is just a shitty overclocker or sapphire dual-x coolers just are not good at cooling vrm since that vrms are the most influential part to getting high voltages and getting stable high overclocks Thanks!
Click to expand...

You are definitely hitting the upper limits of safe VRM temps. Most people have reported the Sapphire as having adequate VRM cooling, so I am surprised to see your temps are so high-about 40-50C higher than what I am experiencing on my MSI TF3. I would suggest ramping up the fan speeds on the card and/or installing more case fans to get air moving over the card-that should help channel heat away from the VRMs.
 
m3ta1head said:
You are definitely hitting the upper limits of safe VRM temps. Most people have reported the Sapphire as having adequate VRM cooling, so I am surprised to see your temps are so high-about 40-50C higher than what I am experiencing on my MSI TF3. I would suggest ramping up the fan speeds on the card and/or installing more case fans to get air moving over the card-that should help channel heat away from the VRMs.
Click to expand...

Thanks for your quick and helpful reply. but i have ramped the fan up to 100% and its still at unnaturally high temperatures. So do you think its mainly cooling? or can it be that the quality of my gpu might be just abysmal? I'm thinking of returning it since I do have coil whine anyway with the card.
So you have 60 to 70 degress VRM overclocked on a TF3? In gaming and over-the-top stress tests like furmark and OCCT? Wow, im envious again lol. With those temps, overvolting will be no problem
 
Insanitic said:
Thanks for your quick and helpful reply. but i have ramped the fan up to 100% and its still at unnaturally high temperatures. So do you think its mainly cooling? or can it be that the quality of my gpu might be just abysmal? I'm thinking of returning it since I do have coil whine anyway with the card.
So you have 60 to 70 degress VRM overclocked on a TF3? In gaming and over-the-top stress tests like furmark and OCCT? Wow, im envious again lol. With those temps, overvolting will be no problem
Click to expand...

If you've already got the fan running at 100%, surely airflow can't be the problem...I would pop off the heatsink and check the mount on the VRM sink. It may not be making proper contact.

And yeah, my full load temps in a Heaven loop @ 1200mv/1200mhz core/1850mhz mem are 74C (GPU), 71C (VRM1) and 61C (VRM2). Temps in Furmark are slightly higher. I have little obstruction in my case and the GPU gets plenty of air moving over it. Pic of internals:

b4i6z.jpg
 
m3ta1head said:
If you've already got the fan running at 100%, surely airflow can't be the problem...I would pop off the heatsink and check the mount on the VRM sink. It may not be making proper contact.

And yeah, my full load temps in a Heaven loop @ 1200mv/1200mhz core/1850mhz mem are 74C (GPU), 71C (VRM1) and 61C (VRM2). Temps in Furmark are slightly higher. I have little obstruction in my case and the GPU gets plenty of air moving over it. Pic of internals:

b4i6z.jpg
Click to expand...

Smexy rig buddy, love the leds
ya, looking at the majority of people's results of VRM temperature in this thread, I think I'll return it and pray I get something better in the silicon lottery. Maybe I'll check how well the heatsinks are seated on the vrms, just worried that it would cause the store I bought it from to reject it.
So just to confirm, do u have coil whine in yours? and when you loop the Heaven benchmark, does it ever stutter and drop in performance drastically after a decent amount of time? (10-15 mins in the loop). Mine does throttle in performance and i conclude from that the temperature mustve gotten too high.
 
Insanitic said:
So just to confirm, do u have coil whine in yours? and when you loop the Heaven benchmark, does it ever stutter and drop in performance drastically after a decent amount of time? (10-15 mins in the loop). Mine does throttle in performance and i conclude from that the temperature mustve gotten too high.
Click to expand...

No coil whine, no throttling. It definitely sounds like your card is faulty if it's exhibiting all of those issues-I would exchange it for sure.
 
I see, I will then.
In your opinion though, if you had a choice, would u pick the sapphire over the TF3? ive heard people say aftermarket coolers have crappy vrm cooling but superb core cooling, but that doesnt seem that is the case with your TF3
 
Insanitic said:
I see, I will then.
In your opinion though, if you had a choice, would u pick the sapphire over the TF3? ive heard people say aftermarket coolers have crappy vrm cooling but superb core cooling, but that doesnt seem that is the case with your TF3
Click to expand...

The general consensus is that the Sapphire VaporX is supposed to have the best VRM hsf. I believe the best stock cooler out of the bunch is the Gigabyte Windforce (but it's considerably more expensive). The Asus DCII and MSI TFIII are right behind in cooling capability. The MSI uses a single plate to remove heat from both the memory and VRMs-most people consider this a compromise. In my experience VRM cooling on the MSI has been more than adequate, but I think this is highly variable depending on ambient temps, case airflow, fan profile, etc.
 
m3ta1head said:
The general consensus is that the Sapphire VaporX is supposed to have the best VRM hsf. I believe the best stock cooler out of the bunch is the Gigabyte Windforce (but it's considerably more expensive). The Asus DCII and MSI TFIII are right behind in cooling capability. The MSI uses a single plate to remove heat from both the memory and VRMs-most people consider this a compromise. In my experience VRM cooling on the MSI has been more than adequate, but I think this is highly variable depending on ambient temps, case airflow, fan profile, etc.
Click to expand...

Hmm you are absolutely right with airflow. I changed my fan config by putting up a 140mm corsair fan originally from the bottom of my case(HAF922), to just to the right of the graphics card. It made the vrm temps stay around decent temperatures even if I overvolted it though: around mid 70s to 80s and made my gpu core cooler @1160mhz/1500mem @1.17. But in EVGA oc scanner, it still went off the roof 90 plus. Do you think tools like furmark, occt and OC scanner are healthy and realistic?
 
Last edited:
Insanitic said:
Hmm you are absolutely right with airflow. I changed my fan config by putting up a 140mm corsair fan originally from the bottom of my case(HAF922), to just to the right of the graphics card. It made the vrm temps stay around decent temperatures even if I overvolted it though: around mid 70s to 80s and made my gpu core cooler @1160mhz/1500mem @1.17. But in EVGA oc scanner, it still went off the roof 90 plus. Do you think tools like furmark, occt and OC scanner are healthy and realistic?
Click to expand...

no. why do people still use furmark or occt. just run real apps like Heaven 3.0, 3D Mark 11, BF3, Crysis 2. these are what you should look at for stability, core and VRM temps etc. The effect of good case cooling and air flow is not known to many. Its in fact the most important factor other than the cooling on the card.
 
raghu78 said:
no. why do people still use furmark or occt. just run real apps like Heaven 3.0, 3D Mark 11, BF3, Crysis 2. these are what you should look at for stability, core and VRM temps etc. The effect of good case cooling and air flow is not known to many. Its in fact the most important factor other than the cooling on the card.
Click to expand...

UPDATE after running heaven for about 15 mins, the benchmark started to stutter and decrease in fps. I checked and I saw that the vrms are at 91c? Isnt that still fine? I really think. Sapphire's dual x cooler is shit at cooling vrms now. This is all at 1160core/1500mem @ 1.18 v
 
Pretty bad coil whine in my MSI 7950 that I received a couple days ago. Already have a return setup with Amazon but I really don't like playing the lottery with video cards that are this expensive. I honestly have had more issues with ati than nvidia but the 7950s are such a good value. Not sure what to do :confused:
 
t4t3r said:
Pretty bad coil whine in my MSI 7950 that I received a couple days ago. Already have a return setup with Amazon but I really don't like playing the lottery with video cards that are this expensive. I honestly have had more issues with ati than nvidia but the 7950s are such a good value. Not sure what to do :confused:
Click to expand...

You are not alone buddy, the 7950s are an incredible value, that is, the ones without any heat problems, coil whine, etc.
 
Insanitic said:
UPDATE after running heaven for about 15 mins, the benchmark started to stutter and decrease in fps. I checked and I saw that the vrms are at 91c? Isnt that still fine? I really think. Sapphire's dual x cooler is shit at cooling vrms now. This is all at 1160core/1500mem @ 1.18 v
Click to expand...

Couple of things. Make sure the card is getting good air flow. so clear up cable clutter if there is any. How many case fans are you running ? Also why don't you run at 1100 Mhz. see if a lower voltage can handle those speeds . you will have a significant reduction in temps with lower voltage. But since your card has coil whine i think you should consider getting a replacement or return to the store for refund and get the Gigabyte HD 7950. it has the best cooler.
 
Thanks for all the help raghu and m3ta1head, this forum is really helpful!
Im still wondering what the VRM temps are for people that have the Sapphire OC cards. Anyone?
 
I've got a 7950 coming today (the Sapphire that's on sale right now at Newegg for $290).

Is there anything I should know before OCing this bad boy?
 
Ok, i found something about why my sapphire 7950 oc is throttling after a few minutes of heaven benchmark @ stock speeds. Oddly enough, when I moved my mouse at the time it was fluctuating from 10 fps to 40 fps in the benchmark, the frame rate shoots up and stays at a constant fps right after I move my mouse. It goes back to normal!??! Can anyone explain this? I checked GPUZ at the time it was doing the fluctuating and the gpu usage was jumping from 64 to 100 and back to 64 over and over again??!! Im on 12.11 beta drivers
 
Last edited:
Okay, I'm going to test OCing on my Gigabyte 7950 WF. I haven't changed anything yet, but ran Heaven once to see what I get.

FXQxe.png


lNUo4.png


Should I be worried that my ASIC % is at 63.7 though?
 
Updated Bios on the video card to F3, and turned on auto-OCing on the motherboard for the CPU, now have the following:

l0VGy.png


3KKqH.png


ILxnX.png


Updated the card to a Core of 1003 and Memory of 1575 and ran Heaven again, this time with a score of around 1961 (I lost the first edit of this). I'm going to do more testing before I change anything else. Card is currently idling at 50C with 33% fan speed.

3DMark 11 (basic) gives me P7854. I think I'm going to test now in actual gaming. Heat never got above 65C and fan was never over 50%.
 
Last edited:
You must log in or register to reply here.
Back
Top