All Activity

Hey there! Would it be possible for you to make a dump of the ColorBoard’s BIOS, or whatever that chip with the blue dot is? I know that it is quite the hassle, but considering that these boards are unobtainable and there is quite possibly no one else on earth both having the board and capable of dumping the firmware, here’s me asking for your help. Thank you!

Hmmm🙃 I fix it for you.

Looking good! I prefer liquid metal over kapton for the absolute 100% seal and it is easy to remove anyhow. Asus being Asus for the sake of being Asus..... When I haven't worked with a particular MB maker in a bit and I start digging into the BIOS, I always hit little irritants of localized naming nomenclature and routinely think, "Why can't they all just use the actual terms?" You would think by the time someone is digging into timings they most likely kind of know what they're doing.... Sadly, prices are already on the rise for AMD cards, 5090 stock is drying up even more now along with 5080 cards on the lower pricing end....yay. A solution looking for a problem........................ The logical solution is still to use the included cable if only for warranty purposes or at least use a PSU from the same GPU maker if you want a straight 12v2x6 run for the same reason. AIBs will look for any excuse to deny a warranty claim whether legit or not as we saw with the Cablemod fiasco. Outside of that, some of the twists, turns and pressure I've seen put on these cables is concerning too. When we see reports of burning/melting cables, all we really get most of the time is a, "trust me bro" guarantee there were no other mitigating circumstances contributing to the meltdown. You have AIBs who have literally tested these connectors under strict conditions and they are fine. Then you have end users who run into scenarios where they melt down but yet again in the OC forums you just really don't see too many cable melt downs across many brands and connectors and users on the regular shoving way more watts into their cards. 5080's with 32GB I can definitely see affecting supply.....AI needs to burst sooner than later.....

In the Control Center's Custom Performance tab, check the "Advanced CPU Settings" box (or similar) to unlock CPU settings in the BIOS. Close ThrottleStop and delete the settings file. It's an automatically created .ini file in the same folder as the icon. Disable Core Isolation in Windows. Restart the system and access the BIOS. There will be an "Advanced CPU" tab. Click on it. There, you can disable virtualization features, Overclocking Lock, and Undervolting Protection. Restart the system and you'll be ready to adjust the throttlestop!

I wonder how many will support this... But as you know, there is people out there also for this garbage. Thermaltake’s new AIO uses “AI Forge” software that requires OpenAI subscription to generate images

No thanks Brother @Papusan. I have no space in my life and no desire at all to waste money on mobile computing garbage. I hate turdbook filth more than I ever have. Just the thought of it makes me feel sick.

In principle, what you said has been given, but there is not much change, but the maximum temperature is better, but the consumption seems to increase. 😁

Yes, I leave it on the custom profile with maximum performance and boost fans for benchmarks.

Thanks! Are you using the Performance section of Control Center in addition to TS? I haven't used the TS program in a long time and I see that a lot of changes have happened to it in the past few years. 😄

GPU sum up for 2025. And the trash will probably be more expensive in 2026... AMD and NVIDIA graphics cards will be more expensive in early 2026 because of DRAM crisis https://www.tweaktown.com/news/109484/amd-and-nvidia-graphics-cards-will-be-more-expensive-in-early-2026-because-of-dram-crisis/index.html Tempted bro @Mr. Fox? Coming Jokebooks from Eurocom at +$15000. Eurocom Raptor X18 laptop rocks RTX 5090 GPU, 256GB DDR5 RAM, 32TB SSDs, costs over $15000 https://www.tweaktown.com/news/109462/eurocom-raptor-x18-laptop-rocks-rtx-5090-gpu-256gb-ddr5-ram-32tb-ssds-costs-over-dollars15000/index.html

This is the card I bought on ebay. It is an mxm Quadro T1000 type b for a 17 inch HP zbook gen6. I want to put it in a dell precision m6800 LVDS model. The card has two form factor problems. There are no notches on either side of the connectors that fit into the mxm slot on this card, so the dell mxm slot has tabs that expect and fit into these notches that block the slot. The other issue is the two retain screw holes to retain the mxm card to the motherboard are in the wrong place, which i think can be dealt with by judicious care in use of tape instead of the screws. I will have either notch the card by hand or demo those tabs on mobo to a low enough height to physically be able to insert card. Any advice on which of these mods is more advisable & reasonable tools and methods to achieve either one? I am a linux only guy so I don't worry if windows is unhappy with card. I am ready to mess around with pad placement until cooling is ok. This should not be high wattage card.

There's no OC on the CPU, only undervolting. I tried increasing the ratios using ThrottleStop, but although I can change it in the FIVR window, it doesn't seem to be applied. Both the P-cores and E-cores remain at their standard clock limit of 5.2GHz and 4.6GHz under full load. My undervolt is -60mV for the CPU Core and CPU P Cache and -30mV for the iGPU, with IccMax at maximum on all. I also unchecked the TVB option in FIVR so that the CPU doesn't reduce its clock speed due to the target temperature (I believe that with TVB enabled, the P-cores decrease by 100MHz starting at 70°C and another 100MHz when the temperature reaches 95°C).

Has anyone done this for a X170, was debating grabbing a 4090 card but the heatsink comments have me out of depth on this.

Interesting... Setting tRDPRE and tRWPRE from "Auto" (4) to 3 in the BIOS seemed to have made a minor improvement in performance and equalized my tPHYRDL. It certainly is stupid that motherboard OEMs are not using common nomenclature for all BIOS menu verbiage as we are on the verge of entering 2026. In their infinite wisdom A$$zeus has decided to refer to these settings as "Read Preamble P0" and "Write Preamble P0" for some idiotic reason. Why not just tRDPRE and tRWPRE? Keep it simple and speak the same language... dumb-dumbs.

Thanks again for your words! 🙂 Yes, you're right, the vapor chamber is very good but reaches its limits when maximum heat output is achieved. Still, we can think about that later. But I would like to point out that the vapor chamber for the P870xx was designed for two GTX1080s. That's just under 400W of electrical power. Anyway, water cooling is quicker to implement for me when I have the spare cash! I've also tested a few desktop cards and the power consumption varied between 250 and 400W! I would still try to reduce the power consumption as much as possible. It's all possible! You also have to remember that the power supply in the hard drive bay gets correspondingly warm at the coils and MOSFETs when running at high power. That also needs to be cooled a little. My power board is very conservatively designed and is geared towards minimum heat conduction! So far, I've only had one failure of my power board with an RTX4090! I think everyone can understand that this is too much for the P870xx! With an RTX5070Ti or RTX5080 with reduced performance, the P870xx is faster than any current laptop with maximum gaming equipment up to UHD! As I said, I want to test my latest version of the MXM board after assembly to see if all resolutions and frequencies work. That's the most important thing for internal operation in the P870xx! At first, I underestimated the DisplayPort lanes and the routing through the MXM board in particular! In the latest version, I paid particular attention to laying the lanes according to all the rules of the art! And it wasn't just the internal display that was important, support for the external display was also important. I would have liked to have implemented support for all DisplayPort and HDMI channels, but unfortunately, even with a 12-layer board, it wasn't possible! After that, I'll pick out an RTX 50xx and do the cooling. Functionally, my MXM board and power board are OK and everything is working as desired! When my financial situation improves, I can do more again.

Nice! Would you mind sharing the settings you used (CPU OC) with me? 😮 🙂 I would be happy to try it out and share the results.

That looks totally awesome, brother. I love it. To protect the GPU SMDs around the core you can use clear nail polish, the red Thermal Grizzly conformal coating (or other conformal coating) and the Thermal Grizzly GPU Guard gasket. I use clear nail polish covered by the GPU Guard, so double-protection. My recommendation would be doing that. You not only protect the SMD components, but everything else beyond them. The gasket also minimizes exposure to air, so the liquid metal should be less likely to oxidize. https://www.titanrig.com/thermal-grizzly-tg-gpu-guard-nvidia-gpu-set.html

problem with shunt mod is not temperature related. it give the gpu more current , regardless of the manufacturer standard. for silicon propriety, more current u give in long term, more it will need in the future. for esample, now gpu will need 0,8volt 100A. after 2 years it will need 0,82volt and 102A. with shunt mod u give gpu 1.1volt and 120A. after 2 year it will need to stay stable at same clock with first shunt mod install 1,2volt and 130A. after a while the bga will not work even with much more overvolting. there are a lots of discussion and video on yt about shunt mod vga failure.

Well after a dozen mock ups, test fits, cuts, more mock ups and test fits this is the current state of the watercooling project. Was able to mount both pumps and res on to the UT45 like I wanted. Essentially used all the fittings save for maybe 2. I assumed the 90 degree rotary fittings were stand alone, didn't occur to me until a bit ago that they need to be coupled to a standard compression fitting. Luckily I purchased enough for the loop (including the 2 filters) Radiator outlet may be better served by a 45 degree fitting long term but tube isn't kinking so it may be OK. Still haven't been able to get the shelves yet, my aunts been using my car for her side hustle quite often while I've been on vacation so I haven't been able to go anywhere. Ideally when I go the shelf approach the rad/res/pump will all be above the pc and tubing will be fed through the shelf directly to the pc below. Probably going to put in my 3090ti to hold me over while I install the ek block. To protect the diodes from liquid metal, is clear nail polish still the preferred method? I did buy some kapton tape and still have some 33/88 electrical tape.

Need to buy a spare ac adapter for my clevo x580 laptop (the one with an rtx 5090 gpu) Just as a precaution incase the one I have breaks

However, you did an amazing job and completed the most difficult part, which is transferring the signal and data from the GPU to the internal screen using an MXM adapter. This is truly the hardest part of the process. You are a legend, and you have all my respect and appreciation, my friend, for this great effort. I hope to be one of the first buyers of this project once it’s completed

I think the vabour chamber will not succeed to cool 250w gpu the air cooling not be effective

Even if the GPU temperature doesn’t exceed 70°C ? And don’t forget the gpu vbios can handel 165w with Nvidia old drivers

most of the shunt gpu mod killed the bga in long term use. i wouldnt do that. u can try oc the gpu core based on the optimal vcore. it's much better to have a stable notebook then a faster one 🙂