• 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.

960T unlock / overclock

DWolvin

2[H]4U
2FA
Joined
Nov 13, 2002
Messages
3,797
So, my trusty 720 (4core/3.4 Ghz) has been replaced by a 960T, but I don't seem to have gotten lucky this time. If I activate even one more core it fails to boot past the Win7 startup splash, most of the BSoD's are gone way to fast to read. Anyone have unlocking tips for BIOS settings?

The other question is that I was able to bump it up to 3.7 Ghz on a Corsair H50 (push/pull), temps seem stable at about 30C. Where would this be on the bell curve for overclocking a 960? I have to admit that I was almost expecting to hit 4Ghz, but can't really provide any reason for it other than getting a good 720 last time... ;)
 
I searched on some forums since I have the same processors and it seems most people are able to hit 4ghz no problem.

http://www.overclock.net/t/1186778/my-new-960t-4-0ghz-overclock-any-comments-suggestions

This thread had some tips which I followed but ended up settling at 3.7 myself.

Though I have not tried unlocking I will be giving it a shot this weekend since I'll be able to get my new fans and mods in by that time to get temps down a bit more.
 
I'm doing the exact same upgrade as you (720 x4 at 3.4 to 960T) so I'll let you know how mine does when I get it Friday. My 720 has been really unstable all of a sudden - maybe something to do with running it at 1.65V for the past few months.

Motherboard is a GA-MA790X-UD4P and cooler is H100.
 
I just bought a 960t from newegg two weeks ago roughly. Went into bios, unlocked cores, made NO other changes. installed windows, been gaming, benchmarking stable -- system has been running over a week straight no crashes/BSOD though 99% of the time it sits idle.

this was on MSI 890FX-GD70 motherboard.
 
Sits idle?! [H]ard folding needs your computer!
Kidding, but that's about half of why I upgraded...
 
So, my trusty 720 (4core/3.4 Ghz) has been replaced by a 960T, but I don't seem to have gotten lucky this time. If I activate even one more core it fails to boot past the Win7 startup splash, most of the BSoD's are gone way to fast to read. Anyone have unlocking tips for BIOS settings?

The other question is that I was able to bump it up to 3.7 Ghz on a Corsair H50 (push/pull), temps seem stable at about 30C. Where would this be on the bell curve for overclocking a 960? I have to admit that I was almost expecting to hit 4Ghz, but can't really provide any reason for it other than getting a good 720 last time... ;)


Keep in mind that some 960T will truly have one or two defective cores. Have you tried to overclock as just a quad?
 
I'm doing the exact same upgrade as you (720 x4 at 3.4 to 960T) so I'll let you know how mine does when I get it Friday. My 720 has been really unstable all of a sudden - maybe something to do with running it at 1.65V for the past few months.

Motherboard is a GA-MA790X-UD4P and cooler is H100.

that is some crazy voltage. Yikes.

i was looking at this upgrade for my 720 unlocked and honestly, i'm not sure if it's worth it because i think it's only worth it if the 960T unlocks. If it doesn't unlock, i kinda think the 960T as a 4 core is overpriced.
 
Yep- the 3.7 stable is on 4 core. It's not the bump I was hoping for but I'm going to keep playing around.
 
you should be able to hit 4.0

what voltage do you have it set to ?
 
I got one of these POST at 4.7, stable @ 4.4 24/7 stable at 4.2 @ 1.38V

::EDIT::

Sabertooth 990FX will not unlock the cores on this for love of money or voltage.... :(
 
I appear to have gotten a pretty 'meh' 960T... can't unlock any cores and needs 1.5V to do 4.2.

At 1.5V temps start creeping up to 60 unless I set my H100 to its fastest speed, and that's a little loud for my tastes.
 
i have mine on DD TDX + Fesser 1 on it's own 3x120mm rad :D might explain the low tempz and high oc :p, but seriously i got half lucky with my 960T
 
Mine was able to hit 3.8, with 6 cores enabled, at all stock voltages, under a corsair H60. I still haven't been able to take the time to tweak it any higher. Stable under 8h of prime as well as IBT. Sitting at those settings right now actually.

I'll post back here with some screens, this weekend.

I'm working with an asus 880g MATX board and 2x2gb of ripjaws at 1600mhz.

I think my current goal is going to be 4.0 if I can keep the temps under control. From what i've read, it should be possible at or around 1.4 vcore. Still getting the hang of these AMD chips. My last o/c endeavor was with a 920 under a 3x120 rad... A little different game here.
 
I appear to have gotten a pretty 'meh' 960T... can't unlock any cores and needs 1.5V to do 4.2.

At 1.5V temps start creeping up to 60 unless I set my H100 to its fastest speed, and that's a little loud for my tastes.

Anything over 1.4 is just destructively pursuing the last bit of clock on 960T, if you are on air. Mine is on a $35 cooler, running 6 cores at 3.75, 1.28V. The processor was bought right after they hit Newegg.
 
Mine was able to hit 3.8, with 6 cores enabled, at all stock voltages, under a corsair H60. I still haven't been able to take the time to tweak it any higher. Stable under 8h of prime as well as IBT. Sitting at those settings right now actually.

I'll post back here with some screens, this weekend.

I'm working with an asus 880g MATX board and 2x2gb of ripjaws at 1600mhz.

I think my current goal is going to be 4.0 if I can keep the temps under control. From what i've read, it should be possible at or around 1.4 vcore. Still getting the hang of these AMD chips. My last o/c endeavor was with a 920 under a 3x120 rad... A little different game here.


All that voltage and heat for 200Mhz? Nah. Just do what I did and run the NB at ~2700, and HT at ~2700. CPU NB only needs 1.22v for that.
 
H100 is not air...

And congrats on your 960T. Not every chip is the same.

Anything over 1.4 is just destructively pursuing the last bit of clock on 960T, if you are on air. Mine is on a $35 cooler, running 6 cores at 3.75, 1.28V. The processor was bought right after they hit Newegg.
 
H100 is not air...

And congrats on your 960T. Not every chip is the same.

The reference to the voltage wasn't referring to heat and the dissipation, so much as the very low voltage requirement these processors exhibit with 4 cores. I am not heat limited on my processor when clocking 4 cores up, but cannot break 3.95 Ghz with any reasonable voltage. Since I can either get 3.9 stable on 4 cores, or 3.75 on 6 I chose the latter. Part of this is that the North bridge can be clocked up a bit more, and overall heat is within 1c. I really believe that the 960T is simply not capable of breaking 4.4 Ghz under most any circumstances. Maybe this is the low latency or other parts of the cache system not being able to get up there. Electromigration is one of the considerations I make in my clock, also, since it goes up as voltage goes up, reducing the life of the chip.

One final thing I consider is that as you get outside the peak reference clock for the CPU line, I would expect that correctable cache and processor unit errors will chip away at the actual throughput.

I see the pentium 4 and the new FX processors as inverse proof of this. They scale well in terms of clock despite not having a superior fab process (in nm) because the cache is simpler, the ops per clock is lower, and latency is higher. This would seem to give leniency to higher clocking. Much like looser memory timings allow for higher pure clock rate. The outcome in memory is a loss in non-sequential or non-bursty throughput for gains in burstable throughput. So, while you have a faster clock rate, the access wait times allow for less transfer than would exist with tighter timings at the higher clock.

Basically, I would rather not push the envelope for a shorter lived CPU that spends a small amount of time correcting errors or waiting on increased latency required for stability
 
Back
Top