A report from Taiwan says Google's Tensor G5 has successfully taped out. The 3 nm chip is slated to use Google's fully in-house design and manufactured by TSMC.
Well, by the time the Pixel 10 comes out, it’ll be 2 generations after the iPhone that used a SoC from TSMC’s 3nm node (the A17, used in iPhone 15 Pro, launched September 2023). I’d imagine it’ll have caught up some, but will still behind while Apple is presumably launching something from TSMC’s 2nm or A14 node at the same time.
It’s definitely a moving target and I can see a world where it stays in the same relative position. Moving to a better node is a necessary but not sufficient condition to be competitive.
I will add that nodes don’t stay still, either. A 2025 run on a node may have a bunch of improvements over a 2023 run on that same node.
And Google’s jump from Samsung to TSMC itself might be a bigger jump than a typical year over year improvement. Although it could also mean growing pains there, too.
Legacy nodes (known in the industry as “mature” nodes) remain in use after they’re no longer cutting edge. Each run teaches lessons learned for improving yield or performance, so there’s still room for improvement after mass production starts happening.
Very exciting. I wonder if this means we’ll get some competitive benchmarks instead of the generation or two behind we have been seeing.
Well, by the time the Pixel 10 comes out, it’ll be 2 generations after the iPhone that used a SoC from TSMC’s 3nm node (the A17, used in iPhone 15 Pro, launched September 2023). I’d imagine it’ll have caught up some, but will still behind while Apple is presumably launching something from TSMC’s 2nm or A14 node at the same time.
It’s definitely a moving target and I can see a world where it stays in the same relative position. Moving to a better node is a necessary but not sufficient condition to be competitive.
I will add that nodes don’t stay still, either. A 2025 run on a node may have a bunch of improvements over a 2023 run on that same node.
And Google’s jump from Samsung to TSMC itself might be a bigger jump than a typical year over year improvement. Although it could also mean growing pains there, too.
Wow, I didn’t know that the quality of a process changes with time.
Legacy nodes (known in the industry as “mature” nodes) remain in use after they’re no longer cutting edge. Each run teaches lessons learned for improving yield or performance, so there’s still room for improvement after mass production starts happening.