4GB vs. 8GB VRAM: The 1080p Benchmark

[KnightSayNI]

Sigh...

If this keeps going, it won't be much longer and we will be playing text based games on a Mainframe and a green screen monitor.

But can it run Zork II on high?

 

[Tech420]

As I commented before, I was playing CP2077 kind of fine until last year in a Sapphire Radeon RX470 4GB , yes the 4GB version (will not tell what CPU I use, but 24 GB of RAM is what I still have). a card I got in 2016. the game I played that taxed it the most was Metro Exodus, even in low. 1080p, FSR and tweaked settings (not minimum at all) and CP2077 was above 30 fps all the time, so for me, it was fine, anything above 30 is good for me (I even lock some games to 30 to hold temperature and power). then moved to a RTX3060ti, 8GB, and Im happy, training some AI models several times faster, using some local LLMs, and playing with better settings (CP2077 almost maxed). maybe in 8 or 10 years I may need another card.

Agreed. A lot of people get hung up on a super high FPS, but most people won't be able to notice a difference on anything higher than 35FPS - the human eye cannot physically detect any change above around 40FPS

 

[manowar51]

This article feels petty, a poor attempt to vindicate Techspots vastly misplaced criticism of 4GB budget OEM GPUs.

Honestly, I dont understand the hate from TS. These cards have always existed and nobody I know thinks they are suitable for gaming.

 

[b3rdm4n]

The name is 100% deceptive. Calling it a 9040 in a prebuilt would be perfectly fine.

Yep that's it, that's the whole thread. The card itself can absolutely exist, calling it a 9050 is wrong when the memory size isn't the only difference, and that difference will certainly hamper performance significantly.

It simply won't offer 9050 tier performance.

Would 100% be a nothing burger with an apt name.