Soldato
- Joined
- 22 Nov 2018
- Posts
- 2,841
RTX and Tensor cores are about 8% of the die area, so they have very little no added cost.
You're only taking the manufacturing cost into effect. By your argument, you could say that if a blank dvd cost 37p, how can they charge £10 for a film. The product cost millions upon millions to produce and that's what you're paying for, not the material object.
RTX took 10 years to design, the cost of that would be an enormous sum of money. Nvidia need to recoup that money, hense the cost of the graphics cards. The mistake Nvidia made was making RTX in the first place in my opinion. Either that or they should have started development 5 years later when technology had advanced more.
Transistor have zero cost.
Extra transistors have zero cost to manufacture, that's correct because the foundries charge per wafer, not per transistor (providing the die size doesn't increase). However, the R&D cost for adding these transistors is huge if it takes a team of engineers years to design
Last edited: