Subtlesnake

In the graphs I have seen here, the cost per transistor is basically flat, not increasing sharply. The reason to go for a newer node in...

You have a source for the claim that 5LPP is >3x more per wafer?

Above a certain clock speed, then the voltage will need to rise beyond the minimum level. But we are talking about the situation where...

Thraktor is making assumptions based on publicly available information, such as Nvidia's power estimator tool for Ampere GPUs. Of course...

For a given power budget, you can have a wider chip at lower clocks, or a narrower chip at higher clocks. The wider chip will generally...

I am talking about on the same node. And surely thermal limits would come into play for docked, not mobile clock speeds. I guess it...

Yes, and if you could get the same performance with a smaller chip at a higher clockspeed (and no loss in efficiency), why would you not...

Again, the point is that once you are at the lowest voltage, decreasing the clock further just loses you performance with no gain to...

Valve have made it pretty clear they are working on a new Half-Life game. Whether anything sees the light of day is another question.

What I mean is that if the analysis is correct, Nintendo could have built a smaller chip with the same performance.

The issue is with Thraktor's analysis showing that 12 SMs would need to be clocked below the voltage floor on 8nm, so essentially...

I'm not sure what I'm supposed to be seeing. 144 down to 85. 92 down to 70. These are still big drops. It's important to bring back the...

Well 0.6 of 200 gives 120, so that leaves 80 mm^2 for the CPU and residual IO, or 3.48 billion transistors. But the die size won't scale...

I guess only for cost reasons? But I am not the one pointing to SN in the chip signature. I am just making the point that 8nm doesn't...

The "than is needed" is the important bit. Based on the Famiboards analysis, T239 is too big to be reasonably clocked down in portable...