Comments by "IIIRattleHeadIII" (@badass6300) on "NVIDIA... Stop trying to trick consumers! (RANT)" video.

The RTX 4080 desktop is literally a 70 class card. The RTX 4080 has a die size of 389mm^2 while the RTX 3070 has a die size of 392mm^2... The RTX 4070Ti is a 60 class card, the 4070 is a 50Ti class card, the 4060Ti is a 50 class card and the 4060 is a 30/40 class card... What is worse is when they sold a sub-mid-range GPU like the gtx 1080, which was in fact a 60 class GPU for 600-700$, when it should have been 350$ at launch. And now they are doing the same with the rtx 4000 series...
3
What is worse is when they sold a sub-mid-range GPU like the gtx 1080, which was in fact a 60 class GPU for 600-700$, when it should have been 350$ at launch. And now they are doing the same with the rtx 4000 series...
2
@Cube930 Die size is the cost factor, along with the R&D. Nvidia doesn't pay for performance, they pay for die size, because they buy wafers, not individual chips and the smaller the die size the more chips per wafer, the cheaper the chip is, on top of being more power efficient and requiring a cheaper PCB and cooling. Performance is a byproduct, it's not a cost factor. And in gneeral the definition for a mid-range chip, regardless if it's a CPU, GPU, ASIC, FGPA, Trigger, etc is a chip that is roughly half the reticle limit. So far the reticle limit has been 858mm^2 for years, until they invent something isn't based on UV. 389mm^2 and 392mm^2 is roughly half the reticle limit of 858mm^2, considering you never want to hit said reticle limit, never go above 95% of it. I'd say both the rtx 3070 and 4080 are mid-range GPUs. Just like the gtx 1080 was a sub-mid-range GPU that should have costed 350$ but instead costed 600-700$... And if you go and see Nvidia's financial reports from 2016-2017 you'll see that their profit margins were between 90 and 138%, which is why their stock prices exploded back then and this was before AI... The RTX 4000 is just as bad, if not worse than the gtx 1000 series.
2
@thewinner4x873 It's on an much, much better process node and on a new architecture... Die size is what matters, performance is a byproduct.
2
@τετέλεστα Not always, when desktop GPUs didn't consume 450W the laptop GPUs were just downclocked desktop GPUs, example being the gtx 900 and gtx 1000 series.
2
@τετέλεστα Yup. RIP the good old days, killed by people who bought the gtx 1000 series at or above MSRP.
2
The RTX 4080 desktop is literally a 70 class card. The RTX 4080 has a die size of 389mm^2 while the RTX 3070 has a die size of 392mm^2... The RTX 4070Ti is a 60 class card, the 4070 is a 50Ti class card, the 4060Ti is a 50 class card and the 4060 is a 30/40 class card... What is worse is when they sold a sub-mid-range GPU like the gtx 1080, which was in fact a 60 class GPU for 600-700$, when it should have been 350$ at launch. And now they are doing the same with the rtx 4000 series...
2
And the funny part is that the rtx 4080 is a 70 class card. The 4080 has a die size of 389mm^2 the rtx 3070 has a die size of 392mm^2...
1