There's no denying that the ongoing semiconductor shortage has made it a challenging time to get your hands on a new GPU. Even if you can, you're often forced to pay well over the recommended retail price for it, with demand heavily outweighing available supply and thus creating a lucrative market for some unwelcome price gouging. It's what makes new releases of GPUs, such as the RTX 3050, so tricky. On paper, the latest graphics card from Nvidia delivers--it's a good GPU for 1080p gaming at high refresh rates, with the bonus of ray tracing and DLSS support. But if you can't find it for the price it's meant to be sold at, does that even matter anymore?
The RTX 3050 sits at the bottom of Nvidia's current line of new GPUs, positioned as this generation's budget card alongside significantly more expensive options such as the RTX 3060 and RTX 3070. It's evident that the price cut comes with its own share of concessions made on the hardware side. The GA106 GPU that the RTX 3050 uses is a slightly smaller version of the one in the RTX 3060, suggesting that Nvidia is making economical use of silicon it previously might have kept off shelves. That said, you're not going to get close to the same performance, even with the 8GB of GDDR6 VRAM that is shared between the two tiers of cards.
For starters, the RTX 3050 limits its memory to a 128-bit bus, drastically reducing its effective memory bandwidth to just 224GBps. This comes into play when trying to hit higher frame rates at 1440p, where the available memory on offer hits the mark but just doesn't have the raw speed to keep up. Elsewhere, the 2560 CUDA cores and 12 billion transistors initially make the RTX 3050 sound more like a replacement for the RTX 2060 from last
Read more on gamespot.com