Testing Intel 18A / Xe3 (Arc B390) Throughput on RC5-72

\n studio-striking\n

Message boards : Number crunching : Testing Intel 18A / Xe3 (Arc B390) Throughput on RC5-72
Message board moderation

To post messages, you must log in.

AuthorMessage
Ncard00

Send message
Joined: 7 Sep 24
Posts: 15
Credit: 118,560
RAC: 0
Message 8738 - Posted: 16 Jan 2026, 17:37:24 UTC

Hi everyone,

I’m currently planning a "Green Crunching" upgrade to an 80W Intel Panther Lake Mini PC, which I expect to have in hand by early February.

I’m very curious to see how the Xe3 architecture (Arc B390) performs on Moo! Wrapper. Since the RC5-72 challenge is heavily dependent on integer and bitwise operations, the 18A node's efficiency and the redesigned Xe3 Vector Engines should provide a significant leap in "Keys per Second" compared to the older Xe-LP and Xe2 iGPUs.

I’ll be running 64GB of DDR5-9600 to ensure there are no memory bottlenecks during high-throughput crunching. Once the hardware arrives in Denmark, I'll post my dnetc benchmark results here. Does anyone have baseline data for the Battlemage lineage on this project yet?
ID: 8738 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Link
Avatar

Send message
Joined: 11 Feb 14
Posts: 144
Credit: 7,649,259
RAC: 0
Message 8739 - Posted: 16 Jan 2026, 20:02:45 UTC - in response to Message 8738.  

This project does not support Intel GPUs as you can see on the applications page.
ID: 8739 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : Number crunching : Testing Intel 18A / Xe3 (Arc B390) Throughput on RC5-72


 
Copyright © 2011-2026 Moo! Wrapper Project