Clevo Overclocker's Lounge

I have a RT809F flasher but I have no idea if it came with the alligator chip. I don't think it did at all.
Um I found the device
Will this be able to flash it without removing the chip?

 

Even if it did, if you put more than 1.8V to the vBIOS chip you'll potentially fry it. That's why I mentioned the 1.8V adapter.

 

Before that you will need to check which BIOS chip your GPU has. I have , mixed results flashing both MXIC and Winbond chips inline using the TL866 (2 different units), however for @Mr. Fox and a few other with older units can flash inline without issues, while i have to do a lot of shenanigans.
What actually does work for me is the Skypro programmer with the 1.8V adapter. That one flashes both without any issues at all.

End of the day, before you begin, find which CHIP your GPU has, then make sure you have the right part - Programmer, good clips, 1.8V adapter etc. And then proceed by first making a full backup and comparing it with a back up for the same vBIOS made using GPUz.

 

There is a good chance the RT809F will work if it has the vBIOS chip in its database.
http://topmarketingsite.com/wp-cont...V-adapter-VGA-LCD-programmer-ICSP-board-1.jpg

Here's a photo showing the 1.8V adapter.

 

I have 2 of them and both have the same issue, however if i remove overvolt protection by hard modding the programmer inline flashing works fine. And this is using 2 different TL866's. Which leads me to believe something has been changed in the newer units (last year or so). The Skypro however flashes faster and inline without any issues using the same 1.8v adapters. -
http://imgur.com/a/MvkuU

 

Thank you guys. I'm not going to try anything, since I don't want the Magic Smoke to appear. I'll just have to wait for NVflash developments.

 

That is very possible. Maybe some of our enemies got to them and effed things up on purpose with TL866A. It could be brand-specific as well, since there are several companies that sell TL866A programmers.

Good luck with that one. The NVIDIOTS screwed us all on purpose. There won't be a change of heart. Their evil deeds are premeditated and calculated, and the damage is permanent. As long as your GPU has Falcon chit on it, you'll have to use a sledgehammer to program it.

As long as the software supports the chip model that is soldered to the GPU you probably can. Unless you run into some snags like @bloodhawk did. You'll need the 1.8V adapter, but it looks like you have everything else.

 

I don't even have the 1.8v adapter. I have all the other IC's shown in the RT809f picture but not the 1.8v adapter...I'm doomed...

 

Those are very cheap on Amazon and even cheaper on eBay.

 

https://www.amazon.com/XINY-adapter-Iphone-motherboard-Memory/dp/B01GFMYICM

@bloodhawk - did you ever have a go at flashing 1080 vBIOS with this type of programmer?

 

Right now, a programmer is still needed to flash the vBIOS in. Like @bloodhawk mentioned, some programmers required a BIOS chip soldered off, some can do inline flashing. I used a Sofi SP16-B to flash it, it natively supports 1.8V and inline flashing for the MXIC 25U8033E chip. Please see below image.



If the 8bit checksum is 00, the app will show green, otherwise it will show red.
If the official vBIOS that you are using showing red then it's okay. Showing red can be interpreted as the vBIOS maybe modded before, if you don't know where the vBIOS is from, then you might need to be careful. Showing green normally it's safe.

 

It is as you said, a guess.
Both mobile and non-mobile bioses indeed contain string 'BIOS Certificate Check Failed'.
In addition most (but not all) non-mobile bioses contains additional strings related to certs 'GPU Pers Req match found in cert', 'Pers Req match found in cert', 'vendor cert parse complete', 'BIOS Cert Verify Successful', 'Ready to change personality', 'Personality in Req', 'Ready to change personality' etc.
So I assumed some additional checks happen at boot in non-mobile bioses.

 

Yeap. Though im yet to see any Pascal vBIOS without the certificate check strings.

 

I wish my MSI 1080 had the vBIOS chip on the same side as the die like my Clevo 1080s are. It makes flashing so much easier. I'm able to flash in-line without removing the GPUs from their MXM slots and I have to remove both CPU and GPU heat sinks and remove the 1080 from the MXM slot on the MSI. Not very convenient in comparison.

 

Well, all of them (or at least entire techpowerup collection) contain string 'BIOS Certificate Check Failed', but most 1050, all mobiles and maybe 20-30% other 10xx do not contain other mentioned strings.

 

Can you link me to either those vBIOS' without the strings?

 

For example, mobile bioses https://www.techpowerup.com/vgabios/186329, https://www.techpowerup.com/vgabios/186530 do not contain additonal strings, 1080 https://www.techpowerup.com/vgabios/187224 does not contain additional strings, but 1080 https://www.techpowerup.com/vgabios/185388 does contain additional strings.
On the other hand maybe it is just different bios versions.

EDIT: Ok, it do looks like different bios versions.
Versions for bioses not containing 'Ready to change personality' and 'BIOS Cert Verify Successful' are: 86.02.15.00.xx, 86.04.2A.00.xx, 86.04.3B.00.xx, 86.04.3B.40.xx, 86.04.50.00.xx, 86.06.11.00.xx, 86.06.27.00.xx, 86.06.39.00.xx, 86.06.3C.00.xx, 86.07.22.00.xx, 86.07.2F.00.xx.
And versions for bioses containing 'Ready to change personality' and 'BIOS Cert Verify Successful' are: 86.02.39.00.xx, 86.02.3D.00.xx, 86.04.11.00.xx, 86.04.17.00.xx, 86.04.17.40.xx, 86.04.1E.00.xx, 86.04.26.00.xx, 86.04.66.00.xx, 86.06.0A.00.xx, 86.06.0E.00.xx, 86.07.39.00.xx.

 

Umm.. all of them seem to have that "BIOS Certificate Check Failed" string.

 

Yeah tried, once. I dont think that supports our chips. At least i couldnt find it in the list/database.

 

I said earlier that _all_ of them contain "BIOS Certificate Check Failed" but some of them contain additional strings.

 

Ah, i think i got confused then. Those extra could definitely be there because of the BIOS versions. End of the day they dont matter in the scheme of things as they are right now.

 

Updated post http://forum.notebookreview.com/threads/clevo-overclockers-lounge.788975/page-1262#post-10539707 with version lists for both cases.
Maybe these messages are just 'debug leftovers'.

EDIT: Most of 980ti contain these messages too but most of other 9xx do not.

 

You know, we all say we're laptop enthusiasts (at least most of us who frequent here) but the last 5 pages show a level of enthusiast far beyond what should ever be required to get a machine working properly, $5000 enthusiast laptop or not.

Period.

Big difference between getting more and the most out of something far beyond the initial design specs and getting something to simply behave in a nice manner.

 

Wow, that's a major realization, good job for sticking it out to get there

We shouldn't need to pull apart our brand new laptops to have them working correctly, it's the job of the maker / vendor and we should reject poorly made and tuned laptops, and send them back instead of spending so much of our own lives getting them right.

We should be spending that time, our time, doing what we bought the laptop in the first place

It's too bad we've all trained the makers / vendors to know that they can depend on their customers to do their finishing and tuning work to take the laptop that last 10% needed to get them working.

Less time spent by the maker / vendor for the same profits, why wouldn't they?

 

I agree with both of you and this level of utter bull crap will potentially be the death of the PC enthusiast niche on a global sense, not only mobile. To be clear, this includes desktops. One of the new members of our forum that began posting in this thread, on this subject, is a desktop PC guy and the NVIDIOTS have pulled the same nonsense with desktop video cards. (Welcome @niko2004x to the forum, guys.)

It should NOT be necessary to do this to brand new laptops. Understand, this applies to all of them and if you are going to reject one on this basis you must universally reject everything available because everything available today is a malfunctioning piece of trash. BGA, LGA, onboard GPU or MXM... every single laptop sold on the face of the earth is a castrated, malfunctioning, poorly engineered pile of feces that is unworthy of consideration for purchase. Every cotton-picking option available for purchase is messed up, and anyone who doesn't know this is either not paying attention or they're a Kool-Aid drinking fanboy that deserves to get screwed for being so stupid.

The ONLY reason the MSI 16L13 and the Clevo notebook models with desktop CPUs and MXM are even remotely legitimate options is thanks to @Prema. Remove him from the picture and the only intelligent move is flat-out rejection and ridicule of every single make and model of notebook computer being sold today. That's really sad, but it's the reality of the suckiness of the situation.

Know this: if you are not willing to purchase a programmer and jump through these hoops, you will NEVER have a Pascal notebook GPU that functions well, unless you buy a Clevo from a @Prema Partner shop. Everyone else is on their own and can count on a product that underperforms and misbehaves, and gives the Green Goblin lots of headroom to rebrand it as something better later using their firmware and driver voodoo and charge extra for the regurgitated filth.

The fact that they have done this same idiotic bull crap to desktop Pascal GPUs merely emphasizes how utterly despicable and worthy of contempt they are as a company. They are not our friends and we are not their customers. OEMs that sell proprietary versions of their trash are their customers. There is a small group of customers that are able to buy desktop GPUs directly from NVIDIA, but they treat them to the same broken firmware fiesta. The bottom line is they don't give a damn about any of us, and most PC gamer-boys are too misinformed or inexperienced to even recognize that. Most are pre-occupied with the folly of having some thin and light piece of trash to play DOTA, WoW and Minecraft on battery power on the playground during recess.

Simon says, " Meet me between the monkey-bars and the seesaw boys and girls. Last one there is a rotten egg."

 

Why hasn't anyone tried flashing through Ubuntu? Flashrom was able to flash Benq monitors this way (the only other way was to purchase a programmer like the Mstar ISP device or RT809F and flash through the VGA or DVI port).

There's always a way. Just need someone with enough power and knowledge to be able to do it. Someone out there is capable of writing a software flasher, even if it has to be done in Linux....

 

Nobody here is flashing through VGA or DVI ports. Flashing the vBIOS chip directly. With Pascal came Falcon and I believe (I could be wrong) this is more of a hardware-based Nazi tactic that isn't as easy as software hacking.

All of this discussion about workarounds is really a distraction. There should be no basis for us to be having this discussion. The only reason we are is because NVIDIA is a crooked company that enjoys screwing people that buy their products. The sad part is, AMD brings nothing to the table that is as good, so we are left with this discussion about how to make NVIDIA's trash less sucky.

Edit: And, here's the other thing, which is kind of a silver lining with manually programming. Let's say you could flash it with Linux. If you flash with Linux and the firmware mod isn't bootable, what next? You're left with a fancy paperweight unless you have the programmer to resuscitate your bricked GPU.

 

Right! Every time I need to extract the card and flip it over. It became especially inconvenient when the die was applied liquid metal paste or some thermal pads dangled around. I went back to ICD paste because of that. I remember the Clevo 970/980M that I had, the vBIOS chip is also facing up on the same side of the die. Maybe MSI doesn't want users to touch it, haha.

 

When's the last time MSI made a product that was designed to be pro-consumer?

It's not nearly as bad as Razer and ASUS which regularly go out of their way to make anti-consumer products, but they're still not pro-consumer.

 

I don't think any of these gaming laptop companies are pro-consumer anymore.

As time goes by and newer products are released, we see more and more of that.

The peak of high performance laptops to me were the Alienware M18xR2 and the Clevo P570WM.

 

I can tell you based on my experience with the Tornado F5 that there is not an enthusiast focus coming from MSI for sure. It's a great laptop because of the desktop CPU and the potential it has with unlocked firmware, but I am not impressed with the lack of forethought that went into the design of it, and that is consistent with what my opinion of MSI notebooks has always been as an observer. They do one or two things really well to create an illusion of awesomeness, then drop the ball on other things that are more important than what they did right. Good examples are the SteelSeries keyboard (which many people like a lot... not my favorite) or the mechanical keyboard on the Titan SLI machine. They do that, it makes people happy, and then they turn around and do tons of stupid stuff like make the CMOS battery inaccessible and block off the fans with a cover that almost touches the fan housing and use chintzy weak fans to exacerbate the issue, and leave a bare PCH with no cooling, and no cooling of CPU MOSFETs and VRMs. And, then they bake in some kind of power-cutting firmware cancer that makes indefinitely holding CPU clocks above 4.8GHz impossible and block RAM overclocking in the BIOS. So, to support what you said, it is crystal clear to me that MSI's focus is gamer kiddos and wannabe enthusiasts that want a fancy looking cookie-cutter laptop to just open the box, push the button and play games. If I wanted that, I would just spend $500 on an XBOX and be done with it.

Do I still love the Tornado F5? Of course I do. I was able to use it as a foundation for something awesome. It wasn't awesome the way MSI released it. To be fair, however, neither are any of Clevo's products. All laptops suck as shipped from the factory. It's just a fact of life. Only a few models are worthy of being purchased as a foundation to work with to make something excellent.

 

I have other issues with it too. The single SATA port is very limiting, and the poor cooling of the M.2 slots for NVMe drives is another issue. From that perspective, for me, it gets a failing grade. It's fine when tuned, a model from HID or Woodz would come in tip top shape to a user, but to me it would never suffice as a primary PC.

Yeah, this is the "other" stuff. If MSI tried, they would easily take and keep the crown. But they just don't give a crap, apparently.

Eh, kind of. I think something from @John@OBSIDIAN-PC or @Donald@HIDevolution would come in a rather A+ state. If not for them I would never, under any circumstances, recommend any of the LGA models to any user. Period. As I said above, it's just very very poor to have to do so much work on a machine JUST to get it in a really nice usable state.

 

Back to Bios flashing, I think someone should continue or make their own work like this guy did:

http://www.overclock.net/t/1521334/...ypassed-for-gtx-950-960-970-980-980ti-titan-x
Of course maybe I'm talking out of my butthole, but as I mentioned, there are VERY intelligent people around here. Someone is capable of it. Even if someone threw a little money their way...

Hell I'd pay $20 for a binary that can flash the modded firmware made with the pascal editor....

 

We are already working with the most intelligent firmware modder than I know of (Brother @Prema) so some of us are going to be OK. As far as I know he does not do any desktop firmware mods. But, I agree with you 100% about paying someone with the talent to do it. I wish I had the knowledge and talent, and I always donate to the cause. The more people contribute, the more incentive there will be.

No, they don't give a crap. It's more than obvious they do not. If they did one of the first things they would do is ditch the BGA feces and make all of their so-called high performance notebooks use desktop CPUs like Clevo. They're just like all the other loser OEMs... kissing Intel's heiny and going with the flow on BGA filth like good little muppets.

 

Once you put on glory fancy colors on the jokebook's. All of the kids is happy camper and ready to game

Spoiler: FANY COLORS

 

Hey all...not really overclocking, but it just feels right to put this here..

Have yall seen this? Super detailed...IOPS!

http://anvils-storage-utilities.en.lo4d.com/

Thanks @richiec77

 

Yeah, have it in my toolbox. But not used much. The review sites use it.

 

I like that it measures IOPS... for what it's worth...

 

Or people can run it and watch their NVMe ssd's throttling as hell

 

When like furmark it's fairly unrealistic to ever see that under normal usage.

 

Just a little side note. Not all checksums will end in 00.

Second thing. Looking at your programmer...How are you able to flash from the back side of the programmer? Or is that a split programmer like the skypro?

 

That's right. A lot of desktop vBIOS their checksums don't end in 00. The reason I made the color green because a lot of mobile Pascal vBIOS (official MSI, ASUS, Clevo) they end in 00. I might remove the color change later to make it less misleading.

Back side of the programmer? The back is connected with a USB cable with a laptop. Maybe the camera didn't capture the bending cable, please see below another picture.



More google images of my programmer: https://www.google.com/search?q=sky...235&bih=698#tbm=isch&q=sofi+sp16-b+programmer

 

Talking about the part circled in black.

 

I haven't been able to find a chip list for that. Is it in its help file?

 

I see what you meant now. It's design like that.
Below is a screenshot of the instruction ( http://www.sofi-tech.com/English/html/574981146.html). Couldn't find the English version, basically it said use this configuration for common chip-sets. Also the programmer software/driver will confirm/advise the position when it's connected.

 

Ah, ok. I see. That particular programmer will program from 3 locations.

 

Here: http://www.sofi-tech.com/upload/product/SP16/SP16F_List.htm
The list was just updated 2 weeks ago.

 

I probably sound like a complete baboon, but what are you guys programming with this? Are you trying to unlock the Vbios or voltage to the GPU? I imagine if you don't know what you're doing you could destroy the card.

 

If you use a 1.8V adapter or a programmer that directly supports 1.8V programming without the adapter, you should not have anything to worry about other than ESD damaging the GPU (which is always a thing to be careful about working inside of electronics). The important thing is that you take a few minutes to pull your stock firmware off the chip and save it. Make sure it is a clean copy and reliable and save it in a safe place where it will not get deleted by accident.

If you flash a firmware mod that is not bootable your GPU will be soft-bricked, but you can go back to the stock firmware without any problem using the programmer. My Clevo 1080 cards have already been soft-bricked by @Prema with test vBIOS mods a couple of times and it was not a big deal using the programmer. It would have been tragic without it, making the GPUs essentially broken and unusable. This is one of the benefits of using a programmer for firmware mods. Same applies to the system BIOS. As long as you have known good firmware to fall back on in the event what you flash is unusable, this is safer that software flashing.

 

Anything that can be programmed with it. Some examples are The Vbios/Bios chips.

As to trying to unlock them...@Prema already unlocked the Bios. Some of us are just experimenting with Vbios files using the programmer.

As to not knowing what one is doing. There is always that possibility with someone who who knows exactly what they are doing. Experiments are just that. Experiments.... In the end...it's how well you can recover from them in the event of a fail