[silicon] BC-250 stable clocks report #54

Activity

1. MTSistemi

   added

   silicon-lotteryStable clock reports from other boards (silicon lottery)Stable clock reports from other boards (silicon lottery)

   bc250Specific to the AMD BC-250 boardSpecific to the AMD BC-250 board

   on Aug 23, 2026
2. MTSistemi commented on Aug 23, 2026

   

   MTSistemi

   on Aug 23, 2026

   Owner

   More actions

   Thanks — this is the first report we have from a board that is not ours, and it arrives on a useful day.

   For comparison, the boards we have numbers for:

   	GPU	CPU	CUs
   yours	2200 @ 1000 mV	3500, no undervolt	40 / 40
   ours	2230 @ 1000 mV	3700 @ −150 mV	40 / 40
   a third, reported by email	artefacts at 40 even at 1100 mV	3500, no undervolt	38 / 40

   So two boards out of three hold all forty compute units, and the third needs one WGP pair switched off. That is the silicon lottery in one table, and it is exactly why the unlock is a switch rather than a default.

   Two things, if you have a minute. Your CPU sits at stock 3500 with no undervolt — ours holds 3700 at −150 mV, so there may be room on yours; the Tuner's Suggest UV proposes a value and Test (benchmark) validates it, stepping back if it fails. And the version field came through empty: uname -r plus the package version would let us tell reports apart once there are a few of them.

   Two stock blowers is worth noting on its own — most people run one.
3. MTSistemi

   added a commit that references this issue on Aug 23, 2026

   Silicon reports move to Discussions, and the Tuner fills in what it k…

   ...

   2352e7b
4. Repository owner locked and limited conversation to collaborators on Aug 23, 2026
5. MTSistemi

   converted this issue into a discussion #55 on Aug 23, 2026