Home / Tech News / Featured Announcement / TecLab modded a RTX 2080 Ti with 16 Gbps memory chips

TecLab modded a RTX 2080 Ti with 16 Gbps memory chips

Over the last few days, we’ve been hearing rumours that Nvidia may well release the RTX 2080 Ti SUPER after all. The folks over at the TecLab YouTube channel decided to get a headstart by modding a RTX 2080 Ti with RTX 2080 Super memory – creating what they called the first “RTX 2080 Ti Super”.

By taking the GDDR6 memory chips from two Galax RTX 2080 Super cards, the enthusiasts at TecLab modded them into a ‘Galax RTX 2080 Ti HOF OC LAB Edition’. With this mod, they created a RTX 2080 Ti capable of reaching 17.2Gbps memory clock speed, when overclocked. This wouldn’t be possible with the original 14Gbps GDDR6 memory chips that come with the stock 2080 Ti.

In this 16-minute video, we can see the process of replacing the sensitive memory components of the GPU. The 2080 Ti’s 11 memory modules had to be removed, and then 11 memory chips from the two RTX 2080 Super cards were removed and re-soldered onto the 2080 Ti. This meant TecLab performed a total of 33 manual solder operations to complete this mod.

Image credit: tomshardware.com

They also described and showed how they took the memory chips by heating them, reballed and removed the old solder from them, and put everything together by soldering the chips onto the 2080 TI PCB.

For this mod to work, no BIOS update was needed. The RTX 2080 Ti BIOS detected the newly installed chips and initially ran them at 14Gbps. After that, they overclocked the memory on the GPU to 17.2Gbps and ran SUPERPOSITION, scoring 11460 in the “1080p extreme” benchmark, or around 2000 points more compared to a stock 2080 Ti.

Discuss on our Facebook page HERE.

KitGuru says: Do you think that you would be able to pull off this mod? Would you buy an RTX 2080 Ti Super if Nvidia officially announced it?

Become a Patron!

Check Also

Apple expected to release first ARM powered Mac in 2021

It has long been rumoured that Apple is looking to move away from Intel and x86 onto …