*Official* NBR Desktop Overclocker's Lounge [laptop owners welcome, too]

It is highly unlikely that we will be able to use a K|INGP|N vBIOS on anything else. That has never been possible in the past as far as I know. But, that GPU model is always in a class of its own and, if history holds true, all of the rest should be cross-flashable.

It will be interesting to find out if the Lightning really is as good as the marketing hype makes it out to be. There was a ton of hype around the Duke and Trio, and neither one of those are as good as the XC2 or FTW3. With all of the firmware being as crippling as it is, the extra power phases on some of these cards seems like a waste of resources and a gimmick that exists primarily to have an excuse to charge extra for something that is not needed.

 

This was kind of weird, but was easy to fix (after I found the example).

https://forums.evga.com/POST-CODE-4F-m2877499.aspx#2913494

Solution:

 

Ask and you shall receive.

MSI RTX 2080 Ti Lightning vBIOS

https://www.techpowerup.com/vgabios/207628/msi-rtx2080ti-11264-181118

 

Awesome, and thanks! Let's see how it compares to these that I just ran. The specs seem similar to the Aorus Extreme vBIOS, but (hopefully) the behavior will improve a little more. (The runs below are with the Aorus Extreme vBIOS.)

https://hwbot.org/submission/4050736_ | https://www.3dmark.com/spy/5939448

https://hwbot.org/submission/4050728_ | https://www.3dmark.com/spy/5939335

https://hwbot.org/submission/4050753_ | https://www.3dmark.com/fs/18025270

 

Well, shoot. This vBIOS sucks. It doesn't hold voltage high enough to match my overclock with the EVGA XC2, FTW3 and Gigabyte Aorus Extreme firmware. The slightly higher power limit is utterly worthless, just like the Galax HOF vBIOS. But, thank you for sharing it. At least now we know. (I wasn't really expecting awesomeness from MSI, but you never know until you try.)

---

Not to change the subject, but I just got this email and I am scratching my head. Why? Why do people that design computer products have to be so stupid? LOL.. what an idiotic thing to do. They must have way too much free time on their hands. What a worthless feature. But, I suppose the kiddos are going to want it because it's new and will excite their friends on Facebook.

 

Well, the parts for the shunt mod didn't show up as expected today. Hopefully, USPS will deliver them 1 day late tomorrow. Got a bunch new personal high scores tonight and racked up some nice points for the PremaMod team. This is using the Aorus Extreme vBIOS. I'm eager to see how much further the shunt mod will take this GPU.

https://hwbot.org/user/mr._fox/

 

Still waiting for the resistors to show up for the shunt mod. But, making progress for Team @Prema

 

https://hwbot.org/user/mr._fox/

 

All this talk of using a different vBIOS on a GPU got me curious to try the same thing on my GTX 1070. I've got the Zotac GTX 1070 AMP Edition (which has 264W limit), and wanted to try the Zotac AMP Extreme version of this vBIOS (300W limit). I followed the nvflash instructions here ( https://www.reddit.com/r/nvidia/comments/5imp8c/successfully_flashed_zotac_amp_extreme_bios_on_a/), except using 'nvflash64' rather than 'nvflash'. It flashed sucessfully with no hassle. There were some differences, fan control was still wanting to keep the GPU at around 68-70 degC, except that it was reporting a lower % fan speed whilst doing so, therefore I'm assuming this vBIOS gave me a greater max fanspeed. My overclock didn't increase or change even though the Extreme vBIOS has a slightly higher default Mhz shipping frequency. Clockspeed didn't hold so well during extended benchmarks though, I went from 2050Mhz sustained with the original vBIOS down to 2037Mhz on the Extreme vBIOS. Needless to say I reverted back to the old original vBIOS because that has more than enough TDP headroom, but I did want to try a new vBIOS to see if it would work & if there could be any advantages.

 

You should try more. Some of the vBIOS I have tried that "supposedly" are better because of a higher power limit totally suck because the power limit doesn't get utilized properly and there are other aspects of the vBIOS that are poorly optimized compared to others. MSI Lightning and Galax HOF vBIOS both suck for me, but the ones that produce the highest benchmarks scores and the most stable overclocks are the Gigabyte Aorus Extreme (380W) and FTW3 (377W). My stock vBIOS (366W) works better than the Lightning and HOF vBIOSes because it holds the voltage where I lock it, as to the Aorus and FTW3. Those two "better" vBIOS don't hold the voltage worth a damn, LOL.

 

OK, my shunt mod is complete and it has worked with a major caveat: voltage is still the same... TOO LOW. It is not drawing any more watts because it was already getting all of the watts it needed access to based upon the stupid low voltage limit. So, I cannot get a higher overclock with the same voltage. It was a success in that it is functioning correctly by grossly under-reporting power draw and there are no more voltage and power limit reasons showing in HWiNFO64 and GPU-Z. There is a little less fluctuation in the power utilization so it might return slightly higher benchmark scores, but it won't be earth-shattering improvement. Testing that theory is pending. So, I am going to have to do something to get the voltage higher to take this GPU where I want it to go. So far, I have not found any vBIOS to allow more than 1.093V max (all are identical), so we either need a hardware mod or a special vBIOS like the Strix XOC 1080 Ti vBIOS that allows adjustment up to 1.200V.



Here is a side-by-side video showing power draw before and after the mod with the same vBIOS, CPU BIOS settings and same GPU overclock profile. HWiNFO/OSD was reporting about 380W maximum and 120% of TDP during 3DMark 11 Test #1 and now it reports around 250W and 65% of TDP.



Here are some before and after photos... As you can see, the first 0.005 ohm resistors I tried didn't work, so I had to decrease the resistance using 0.008 ohm resistors.













@Johnksss @Prema @Khenglish @Rage Set

 

Still waiting for the new version of CLU to show up. I'm ready when it arrives.

 

Thing is I've got Micron VRAM on my GPU, and I have to use a 'special' improved vBIOS that wasn't officially released from Zotac (even though they created them) in order to get a couple of hundred extra Mhz on the VRAM overclock. If I switch to a non-Zotac vBIOS then how would I know if they have the Micron VRAM fix? Which brand/model of GTX 1070 vBIOS do you think could give the best results?

EDIT: I've worked out which vBIOS would have the Micron fix based on release date & a list of manufacturers that implemented the Micron fix. Tested a number of vBIOS now so far, some which I didn't like the fan curve, doesn't hold clocks at voltage, or card was bouncing off power limits, doing more extensive testing with one now from Gigabyte that might be a keeper (Gigabyte.GTX1070.8192.170628, https://www.techpowerup.com/vgabios/196576/gigabyte-gtx1070-8192-170628), this one has good fan curve, good Mhz holding, and gonna find out if I can overclock further & no game glitches, etc.

Wow, you put a lot of work into that taking apart the card, I guess numerous times whilst you work out which ones needed a resistor & which ones had to have the resistor removed, along with also having to experiment with different resistor values. It sounds like you might have had to have taken apart & rebuild your card maybe 4 or 5 times to work that out?!

I notice that the shunt mod gives you about 20-30W extra power draw in some short sections of that 3DMark11 benchmark, so like you say it's not hitting up against the Power Limit Reasons anymore. That's good.

 

Happy Birthday brother @Mr. Fox !

 

First of all, Happy Birthday brother! I hope it is amazing, with many many more to come.

I've been trying to get the Lightning's LN2 vBIOS. The one you've used was the "standard" Lightning vBIOS. The LN2 vBIOS was used by a lot of the OC'ers who were previewing the hardware like Gunslinger...breaking several world records according to the tech media. Those vBIOS along with your mods will unleash your GPU.

 

i9 9900k 5ghz, uncore ratio 45,voltage cpu 1.2v,ram Gskill 4000mhz cl18

 

Cool. I am looking forward to hearing how the Gigabyte vBIOS is going to work out for your GPU.

Yes, I had to take it apart about that many times and might take it apart again for another experiment. I'd already taken it apart probably like 20 times for experimental vBIOS flashing with the SPI programmer. It's not much effort compared to the probably 200+ times I took apart my P870DM3 for firmware flashing with Brother @Prema. This is only 3 thumb screws holding the GPU to my chassis, two 8-pin cables, 1 DP cable, 2 quick-disconnect hose fittings and 16 screws to separate the GPU from the water block. So, easy-peasy. It's tougher on the Philips screw heads than it is on me. They wear out faster than I do, LOL. (I might have to order some stainless steel pan head screws for this GPU like I did for the P870DM3 chassis. Those don't wear out quickly like the soft mild steel screws do.)

The thing I have found (like almost everything in life) is that anything worth doing is kind of a pain in the butt before it's done. Sometimes, even a huge pain in the butt that you start to question whether or not it's worth the effort. Yet, once it is all done and you get to enjoy the results, it doesn't take long to forget about the inconvenience or hassle you had to put up with prior to breaking through to the other side. It's usually worth it.

Thanks, bro! God has blessed me with an awesome wife and family, great friends, and good health. My best friends are here in this forum, and you are definitely one of them. I can't believe I have been alive 56 years and married to the same amazing lady 32 of them. That's just nuts (in a very good way) and doesn't even seem possible, LOL. Wow... blink twice and you're an old man. Remember, "It's better to look good than to feel good."

Thanks, Brother @Rage Set. It's going to be a great day, I can feel it already.

Man, if you can get that and it unlocks the voltage beyond the silly low 1.093V limit, I will be doing my happy dance, big time.

Looking good, bro. Performance and temps are nice. Looks like you have a good CPU sample.

 

Played with a few settings - jumped up 100 points.

 

Another TSE Bench: Almost another 100 points increase. Using my cold garage advantage!

 

Yes, and Happy Birthday to you!

I tested out that Gigabyte vBIOS quite thoroughly now. It's given me +600Mhz on the VRAM in comparison to the +566Mhz I had before, so it looks like it's implementation of dealing with the fussy Micron VRAM issue is a better one. Max Core overclock was unchanged with this vBIOS, although it took longer to fail at known unstable overclocks. I noticed that it's under reporting power usage, my previous Firestrike Extreme GT1 GPU consumption was 200W, but it's showing only 150W now, I'm not sure if that's an issue apart from the GPU could be overstressed if I ran Furmark (which I won't), praps I'll reduce power limit down to 90% seeing as it's under reporting by 50W, and doing so it'll still be good for 240W (accounting for the under reporting). The auto fan speed curve built into the vBIOS works well for my card, it's a little bit more aggressive than my previous, about +200rpm in known test runs. It also shows a lower % fan speed whilst providing this higher RPM, so I'm assuming the vBIOS has given the card a higher max allowable fan RPM - is that possible, I've not tested it, didn't want to burn out the fans?

I'll keep this vBIOS for now given it's allowed for a little bump in the VRAM overclock, and the slightly more aggressive fan profile is probably healthy for the card whilst still not being noticeably louder than before.

 

It looks like it is available here... https://www.techpowerup.com/vgabios/207629/msi-rtx2080ti-11264-181203

It's a different version than the one I tested before, so I will give this one a go. It seems the specs are listed incorrectly. It shows the LN2 vBIOS has an 80W lower power limit than the default 380W.

 

I may try it as well, but that is suspiciously low...unless they (MSI) did that on purpose so not to anger Nvidia.

 

Thank you for the birthday wishes. I'm glad to see the vBIOS testing was useful in identifying firmware that was a bit better than what you started with.

Yeah, don't bother. Totally a waste of time. This vBIOS absolutely SUCKS at holding the voltage to the max and doesn't give any more voltage, so it's totally worthless and under-performs compared to what I already had. Going back to the Aorus vBIOS for now. Either MSI gave their "special big boys" secret non-production firmware for marketing hype, and so nobody else could ever legitimately compete with their benchmark results, or the 2080 Ti will run 2400MHz and higher with the crappy stock 1.093V limit on LN2... not sure which one.

I don't know if I should hold out to see if ASUS actually releases one with the voltage unlocked like the Strix XOC 1.200V 1080 Ti vBIOS or just say screw it and buy the trim potentiometers and hard mod it so I can do whatever I feel like doing with the voltage. I'm getting just about fed up with NVIDIA's stupidass firmware bullcrap, but AMD hasn't (and I am 99.9% sure never will) released a GPU worth buying.

 

I suspected it wasn't that easy to acquire those LN2 vBIOS. During the livestream of the Lightning Z, the presenter (who admitted he wasn't much of an OC'er) flashed those LN2 vBIOS on the card. He mentioned on the card, there were higher stock vBIOS but not the ones he flashed. I asked if those LN2 vBIOS will be available for download from MSI and he said no. He did say, however, there are a couple places where XOC'ers go to get those secret vBIOS. I assume he is talking about Overclock.net. Since this card has not officially launched yet, it may be another week or so until the real LN2 vBIOS for the Lightning make an appearance. I'm going to keep watch.

 

Yes, we should keep watch and keep testing until we find something that gives us the freedom to make our own decisions on the voltage.

But, it's pretty damned sad that the ONLY firmware I have tested (so far) that actually functions correctly are the XC2, FTW3 and Gigabyte Aorus Extreme vBIOSes. You select the voltage dot you want--as long as it is not more than the sick little baby girl 1.093V limit--slide it up to the clock you want, click apply and *BOOM* it goes there, and it stays there, and it never moves, until you reset/release it. I don't know what is wrong with the idiots that authored all of the other firmware versions out there. (The 1080 Ti had exactly the same issue with most of the firmware available for it.) The only thing wrong with the XC2, FTW3 and Aorus Extreme firmware is the fact that they are limited to a ridiculously low 1.093V. Unlimited power is worthless with castrated voltage because it will never get utilized.

Here is a 40 second video showing what properly functioning firmware looks like, and this is not possible with most of the worthless crap vBIOSes.



It is also really stupid and dishonest that they make it so difficult for those of us who aren't famous yet in the pro overclocking cliques to get our hands on good firmware. They must be mortified by the thought that someone of lower status could actually compete and win if they were given exactly the same courtesy. That makes the whole thing a rigged show, just like "professional" wrestling, LOL.

 

Nice of Steve to make a video for my birthday.

 

You hit the nail on the head there. It is very unfortunate for anyone that isn't in the know to compete with the "pro" overclockers. That is why I asked the MSI reps during the livestream about those LN2 vBIOS. I knew it would be very difficult to acquire them but it seems anyone in MSI's circle not only has those vBIOS but sample Lightning cards. There are already several MSI Lightning unboxing videos from people in Europe and I have a feeling that card (I do want one) is going to be extremely difficult to buy over here in the states.

While I understand the reason why MSI, EVGA and other AIB partners have to make it hard to get these vBIOS (they don't want beginners killing their cards), there should be an official way for us to sign up to get them. I wouldn't have a problem agreeing that if my card blows up, it is my fault.

 

Yes, Johnksss was in the chat with me.

 

Definitely... I wouldn't have a problem agreeing to that either. And, I think the risk is not really any greater than a noob with an enthusiast motherboard. Some of them could kill anything, even running it stock, just by looking at it with crossed eyes. I think it is a combination of being overly (to a fault) protective of the hardware and being protective of their most-favored "golden boys" status.

 

I wonder if the move to 12nm from the 14nm Pascal means that they have to be more careful with maximum voltages? Perhaps they've found out that over 1.093 will start to kill cards if they're close to the max temperature limits. Although 1.2V is only 100mv above the present 1.093V, but I think it's true that increased voltage can suddenly make you fall off the cliff of video card destruction once you go beyond a certain limit.

 

Yes, using too much voltage can cause damage. I don't believe 1.200V is enough to cause any. It wasn't enough to cause any for 1080 Ti, but yet they still locked it to an asinine low value the same as 2080 Ti.

 

Yes, but you were running your 1.2V 1080ti at something like 20 degC on chilled water, that's a world away in terms of potential silicon degeneration if it were running at the normal air cooled 60-83 degC. Low temperatures can protect the silicon from excess voltage (to an extent), and high temperatures exacerbate degredation caused by voltage. They don't want folks running 1.2V while bouncing around at 70's & 80 degC's I imagine. And you'd probably be in the 80's & 90's on standard air cooling at 1.2V I would think.

 

Perhaps. They run too hot running stock without water cooling. And, the way they throttle at room temperature running stock without water cooling, I suspect they might just throttle even more instead of damaging the GPU.

Here is the power draw with the Aorus Extreme vBIOS in 3DMark 11 Test #1.

 

gskill stock 4000cl18 oc 4300cl18

 

Happy Birthday Brother Fox!!!

 

Good testing regime, I do it the exact same way for testing RAM OC's for stability. I've found that it's also the quickest way to find errors.

 

https://hwbot.org/submission/4055563_ | https://www.3dmark.com/spy/5990330

 

How much more percentage performance did that shunt mod give you? I know you're limited to 1.093V for the moment.

 

About 2%. https://www.3dmark.com/compare/spy/5990330/spy/5547809

 

Oh well, shunt mod in preparation for voltage mod I guess!

 

Looks pretty even across the board...

https://www.3dmark.com/compare/3dm11/13135126/3dm11/13125441

 

Tight!

 

Looks like the benchmark charts are going to be dominated by the select few that A. can afford an X599 rig from an SI, B. actually OC the CPU in benchmarks. Looks like Steve from GamersNexus may try to delid the W-3175X.

 

LOL - So I let my rig stay in my garage for too long without it being powered on. The water inside my loop was frozen. I managed this before the freeze...

 

Nice score!

I hope the freezing didn't break anything.

 

No time for me to watch these, 2080Ti Mods Buildzoid
Jan 25 90min

Today's Follow Up 23min

Spoiler: Der8aur Tackles Intel 28Core @1225watts

 

Noice!
Never used two different vBIOS bases in SLI (only different Mods adjusted for each card but using the same base)....

 

That "Progress" video is pretty discouraging. The NVIDIOTS strike again, LOL.

 

If this turns out to be true, then I may have to build one of these

View attachment 169286