News for hackers. Stuff that matters.

A comment on A Pi-Powered Plan 9 Cluster. Parent comment

Re:A Pi-Powered Plan 9 Cluster by bullen on Tuesday October 13, 2020 @11:42AM (#24766649)
Nope 28nm for the Raspberry 4. And 5nm is not going to save you that much energy compared to 28nm, it will however make the circuit much more fragile.

The Raspberry 4 is peak power efficient: http://web.eece.maine.edu/~vweaver/group/green_machines.html

Mobile phones don't have a keyboard, you can only consume with them. I produce things.

Those sticks don't have a company with it's own GPU and linux (Raspbian now RaspberryOS) behind them and they cost more per watt for performance I don't need.

However I have 2x NUC with Streacoms passive case: https://streacom.com/products/nc2-fanless-chassis/

[ Parent | on HN ]
Re:A Pi-Powered Plan 9 Cluster by kingosticks on Tuesday October 13, 2020 @11:53AM (#24766759)
28nm vs 5nm, is what, 3 generations with a 20-30% power saving at each step. ~50% power reduction is something.
[ Parent | on HN ]
Re:A Pi-Powered Plan 9 Cluster by bullen on Tuesday October 13, 2020 @12:09PM (#24766893)
Nope, if you compare theoretical Gflops/watt it's probably 30%, but in practice you'll only get maybe 15%. Memory speed is the bottleneck and you cannot get around that. DDR3 had lower latency than DDR6 and only bandwidth is increasing, and you have to ask bandwidth for what?
[ Parent | on HN ]
Re:A Pi-Powered Plan 9 Cluster by kingosticks on Tuesday October 13, 2020 @12:59PM (#24767433)
Sorry, I don't follow. Why would anything be getting faster if we are talking about hypothetically utilising (just) the power-saving gains from newer nodes?
[ Parent | on HN ]
Re:A Pi-Powered Plan 9 Cluster by bullen on Tuesday October 13, 2020 @02:37PM (#24768617)
Basically Gflops/watt is not going north of 1 for 32-bit CPUs, so humanity is completely out of growth. This is the end. Game Over!
[ Parent | on HN ]