Nvidia clockblock: vBIOS (unblocked in 353.00)

I think that decision states with Nvidia being hell bent on locking mobile GPUs with reblocking vBIOS releases, 350.12 requiring vBIOS mod to OC again, atleast we have custom vBIOS which is unrestricted on any grounds for now...unlike that fiasco of mobile OC lock down with the 347.09 & 347.25 even with unlocked vBIOS which was hell...

Not on OC but on other note..regarding win10

Spoiler

TBH, the Spycrosoft / Micro$shaft's Win10 will be full of spyware and limiting user control AFAIK for the current trend of MS updates and news so far....
http://forum.notebookreview.com/threads/windows-10.762434/page-64

Spoiler

I only hope for that DX12 will someway make its way to Win7 (by hackers or any means necessary) - The Best OS for a " Personal Computer"

 

That 'spyware' stuff regarding the Windows Update, that's just a little patch that will give you a notification in Windows Update that Windows 10 is now free to upgrade when Windows 10 is launched. That's my understanding after reading that, I'm OK with that, I'd then just go ahead & update through the normal Windows Update procedure I guess. Do you think display calibration tools that work for Windows 8 & Windows 7 will be compatible with Windows 10 (ie the Spyder4Pro) I have? I did a google on it, but couldn't find anything - is most Windows 8 compatible hardware gonna be compatible with Windows 10 without new drivers for instance? (If I have any further questions I'll ask in Windows 10 thread, don't want to go to far off topic here!)

And fingers crossed that Windows 10 won't lock down GPUs anymore than what we have seen so far!

 

Nvidia Driver 352.86 Win8.1/8/7 64bit (The Witcher 3 Game Ready Driver)

Code:

http://www.geforce.com/drivers/results/85053

Confirmed. OC still blocked.

 

For now it's the same on W10 as on Win7-8.x, GPU is locked only if vBIOS is locked.

 

Hi all,

I've tried both the 347.88 and 347.90 on my 960m with the following results:

347.88 = won't install due to hardware not compatible
347.90 = installs, but screen remains black

Any ideas?

Are these the only non-blocked drivers?

 

Which laptop do you have? A simple .inf mod will help make driver 347.88 install.

 

Didn't the 960M not get support until after those drivers?

 

The 347.90 driver has GTX 960m support, but yes, the 347.88 driver does not have support for the 960m (as far as I am concerned).

 

I knew it was on the border for support, thanks for clearing it up.

 

Acer Nitro VN7-591G

I'm off to Google inf mod...

 

Just don't clean install 347.90, install the driver while you're still have another version of the driver installed.

 

Well, Dell/Alienware has a 347.90 driver that installs fine with the 960m.
http://na.alienwarearena.com/forums...graphics-amplifier-nvidia-titan-x-beta-driver
Given that it has only 1 .inf file, it should work and have no problems.

 

Not necessarily. Different manufacturers GPUs have different hardware IDs.

 

True, I forgot about that.
However, the 347.90 driver was from Microsoft, so it should be universal.

 

Still no luck.

I tried installing 347.90 over the top of a newer driver, but it wouldn't accept it.

I also found my Hardware ID and modified the inf file, which allowed me to install it fine, but it got to the end of the process and just said failed.

Since my laptop is brand new and shipped with a March BIOS, I'm going to assume for the moment that the vBIOS is part of the main BIOS and that it's locked in terms of overclocking.

=(

On the plus point, it's MUCH faster than my old laptop and I'm happy with the price I paid, so it's not all bad. I'm kind of accepting of the fact I can't overclock a laptop...

EDIT - I used to be a serious overclocker with a DIY water-cooled rig and would solder volt mods onto the graphics cards to get more MHz, but those days are over. These days I spend more time in the gym, which is probably a good thing!

=)

 

Try 347.88 using a clean install. If you're on windows 8.1, you need to reboot with driver signature enforcement disabled to get it to install, if you mod the .inf and such.

 

Did someone tip off TechPowerUp?

 

Wasn't me. But nice to see that TechPowerUp noticed this.

 

Yeah, good to see. But still the laptop hater / destktop fanbois are spouting off in the comments there.

 

Yep, ignorance is the easiest reaction...

 

We have replicated this issue and are fixing this in an upcoming driver for next week. We’ll be sure and let you know when this gets posted. Sorry for the inconvenience.

Regards,
Manuel Guzman
NVIDIA Forums Rep

 

I'll believe it when I see it.

 

That's good that you're gonna remove the overclock restriction again, maybe test the drivers next time to make sure they don't have the block before you release them.

 

Right, like that 880M fix that you promised us and then disappeared. Forgive me for being cynical...

 

We just posted a driver that fixes this issue:

https://forums.geforce.com/default/topic/834157/geforce-drivers/announcing-hotfix-driver-353-00/

 

He's right.

 

I hope Nvidia and NBR will continue closely communicate for the sake of Notebook's gaming performance .Here are many enthusiasts who spend part of their life exploring all around laptops and helping their owners... huge number of which.

Wishlist:
1. Aftermarket mobile GPU right from Nvidia Kinda strange that 3-4 year's mobile GPU cost 200$+ just because of small market. You have no idea how many would like to buy 770M and 970M for cheaper than ebay prices.

 

Or the driver fix for 970. Oh wait nvm, the segmented memory was an intended feature.

But seriously, actions speak louder than words. Fix it and then announce it, not the other way around.

 

Now also on Swedish http://www.sweclockers.com/nyhet/20535-nvidia-clock-blockar-geforce-gtx-900m-serien-pa-nytt

pidge We just posted a driver that fixes this issue:
https://forums.geforce.com/default/topic/834157/geforce-drivers/announcing-hotfix-driver-353-00/

Works this hotfix 350.00 driver also with gtx780m although this is not blocked for overclocking?

 

Good job!

How about the 880m fix we're still waiting for a reply on?

 

Okay. Good job. Now, please just keep giving us control and good products. We're honestly not asking for a whole lot... the same kind of attention desktop cards get for our laptops, non-gimped cards and good driver and OC support. We already pay a whole lot more than should ever be necessary for the hardware, even.

Also, do you mind PLEASE making a driver that doesn't require us to reboot to disable/enable SLI?

 

Impressive Nvidia how likelylistening for customers, great to join this group. I tried this driver and overclocking works with 950M now. Unfortunately the undervolting feature still does not seems to working in Nvidia Inspector, which would be more important to me than overclocking. Seeing the factory 1.093V is enough for +135MHz GPU core speed, than for defaults speed would be enough about 1V.

 

I can't remember if I explained this to you or to another person, but undervolting is impossible because the card itself consistently undervolts itself. Even under heavy load, it constantly lowers and raises its voltage across a very wide range of values. Undervolting the GPU in any way would INSTANTLY cause the lower voltage ranges to go into extremely dangerous territory. Even the stock voltage ranges have issues, as explained by n=1 HERE. In fact, if you had forced voltage to be a high (stable) constant then undervolted, you'd experience much less benefit than leaving it how it is.

 

THX! What took 2 months?

Win10 version please!

I am glad the decision was made to ignore the lock-bit once again (until now only 347.88-92 ignored it).
Hope this gets merged into the next WHQL release and we will never see this block again.

 

Be sure to post results here also: Announcing Hotfix driver 353.00 | GeForce Forums and express appreciation if it works correctly. Hopefully, this is the driver we have all been waiting for.

Testing to follow...


Edit: NOPE! Still a crap driver for my 780M SLI setup. Massive throttling and horrible FPS with moderate overclock. Didn't bother testing stock since stock is unimportant if overclocking is throttled. Don't care if DSR works if it throttles. Testing with 980M SLI on my Clevo P570WM sometime tomorrow.

 

Thanks, I start to getting it. I was thinking it is like an Intel CPU, every state has a certain voltage and adjusting offset voltage will do the same trick for each P-state. It is obviously lot more difficult with this constantly voltage changes for each state and hard to understand. How could I track down if using +135MHz will result generally higher voltages than running at base clock? Checking the voltage in Hwinfo64 or Inspector they report always 1.093V under load even with overclocking or without. I'd like to manage out the best efficiency of this GPU and if using the same voltages for overclock than that is what I need.

 

On a different note... What is that widget and how does it work on windows 8... Is that HWinfo?

 

BOOKS ARE AWESOME

 

I'm not sure, but checking the vBIOS voltage tables would provide a good idea. You'd likely need extremely rapid monitoring software to notice the voltage changes. Maybe if you set HWiNFO64 to show 10x a second you might notice the fluctuations? Overvolting works just like it always did. Add an offset to the voltage range. So if the ranges are say... 1v to 1.093v for clockspeed 1000MHz, and you overvolted by 20mV, it'd go like 1.020v to 1.113v under that load. Understand?

I LOVE BOOKS

 

I'll check the driver a bit later Mr. Fox, and see if DSR does indeed work. 350.12 has not been a crap overclocker for me. I tagged you in another post on the AW forums; uncertain if that user has the same M18x R2 or if he/she is using an AW18, assuming that might make any difference.

 

No, that is an AIDA64 sensor panel... I am using HWiNFO64 output through RivaTuner Statistic Server on-screen display in the benchmark screen shot.

Yes, it works with Windows 8. In fact, that screen shot is Windows 8 desktop.

I've already expressed my displeasure with the crippled driver at GeForce Forum. I hope it at least works right with my 980M SLI setup... if it does, I guess having the driver work properly on only one out of my four systems is better than a poke in the eye with a sharp stick.

I'll check your other post. I have mentioned it before, but it is possible this throttle is being caused by my special, extremely aggressive, vBIOS mod for 780M. I have not tested it to see if it is OK using the normal @svl7 / @johnksss public vBIOS mod. It might be OK with that one, or using a stock vBIOS. I haven't tested yet because I don't intend to tone things down just to accommodate cancerous drivers. I'll just keep using older NVIDIA's drivers that are not garbage. I might test just out of curiosity. I'd like to know why, and that possibility is the only one that comes to mind.

Edit: FOR THE RECORD, I just checked and it does not throttle at stock clocks. So, behavior is same at 350.12 and 352.86 for me and throttles like a son of a female canine overclocked. Worthless driver. Don't care about DSR if I cannot overclock.

Going back to my 345.20 DT CUDA driver mod... again.

 

Nah, still a pile of crap throttle-fest driver. I haven't tested it with 980M SLI yet, but I am going back to 345.20 again on my systems with GTX 780M. 350.12 and 352.86 are both absolutely worthless for me on my 780M systems.

NVIDIA GeForce Driver 353.00 Discussion Thread

 

Just tried it & it looks stunning !!! Hope you don't mind that I kindda stole your format cuz I am too lazy to make one of my own



I get some latency spikes every like 5-10 seconds though ... Am I doing something wrong ?

EDIT: The DPC latency spikes are from Low Level MSR Operations sensor ... Disabling that eliminates the spikes & the processor monitoring as well

 

I am AFK till tomorrow night. I will give it a try then.

 

This driver is another pile of garbage, no overclocking, all stock, stock settings in nVIDIA control panel.......

345.20:



353.00:

 

Hotfix driver 350.00 is bad for processor. @Matrix Leader6 you can see the broken score in gt 1 and physics score especially. Gt2 and gt3 are better in 350.00 than in driver 345.20

 

You'd still get the boost in performance with Witcher 3 that was offered by 352.86, correct?

 

Performance on the 353.00 compared to 347.90 in games was essentially the same with my OC 860m.
In benchmarking, though, as Mr. Fox said, it throttled like crazy with 353.00.
Ehh.... this driver isn't too bad for me then.

 

I usually don't update too often, but I'll indeed have to check later. When I check I will report here, but be warned: I am in a gaming binge XD

 

I never update my drivers unless there is an error, too much headache. But thanks to you guys I can figure it out whether it is a faulty one or not. I think it is a habit back from my old AMD days but it's a good habit