igneum-network/igneum
1
0
Fork 0
Code Issues Pull requests Releases Activity

0.3.25 AMD knob: the RX 9070 XT's measured efficient point, 149 W, in the tuner's ceiling table (the first grid on the knob, PC 1, 8 October 2026, 12:10 BST, run-ca3-pc1-amd-grid-9070-20261008-b, 24 of 24 rows ok, the app's own hash_now, the card reset at the end). The rate is flat at 18.93 to 18.98 MH/s over the whole ladder: the knob moves watts only, so the knee never comes and the stop rule reaches the floor; the efficient point is both floors (clock offset -500, about 2,920 MHz where ADLX clamps; power limit -30): 149.3 W at 18.96 MH/s, 0.127 MH/W, 24 percent under stock's 195.8 W at the same rate. The clock offset alone takes 196 to 159 W; the power limit alone does nothing until -30 (184 W). Test: efficient_watts reads 149

Browse source 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
This commit is contained in:
igneum-labs 2026-10-08 11:23:54 +00:00
parent 3a15e3b0d0
commit aa2b865295
1 changed files with 7 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
app/igneum-app/src/ember.rs
View file

@ -37,7 +37,12 @@ pub const CLOCK_FINE_STEP_MHZ: u32 = 100;
/// 5090 for 575 W with proving on the same card): the tuner never requests more than the card's measured efficient point
/// unless Power control is on and the user raised the cap above the default; a card not in this table is held at the
/// cap it runs at. A point below the vendor's minimum limit clamps up to that minimum (the 5090 floors at 400 W).
pub const EFFICIENT_W: &[(&str, f64)] = &[("RTX 5090", 308.0), ("RX 9070 XT", 199.0)];
/// RX 9070 XT: 149 W from the first measured grid on the AMD knob (PC 1, 8 October 2026, 12:10 BST, run
/// run-ca3-pc1-amd-grid-9070-20261008-b, 24 of 24 rows ok): the rate is flat at 18.93 to 18.98 MH/s over the whole ladder,
/// the knob moves watts only, and the efficient point is both floors (clock offset -500, about 2,920 MHz, the ADLX clamp;
/// power limit -30): 149.3 W at 18.96 MH/s, 0.127 MH/W, 24 percent under stock's 195.8 W. The 199 W before it was the
/// 7 October read at stock.
pub const EFFICIENT_W: &[(&str, f64)] = &[("RTX 5090", 308.0), ("RX 9070 XT", 149.0)];
/// The app's cap when the user chose none (engine.rs requested_watts); a cap above it is one the user raised.
pub const DEFAULT_CAP_PCT: u32 = 80;
@ -1416,7 +1421,7 @@ mod tests {
assert_eq!(l.ceiling(), 403.0, "70% of 575 W, the card's own cap");
let zero = with_ceiling("NVIDIA GeForce RTX 4070", Point::default(), false);
assert_eq!(zero.ceiling(), 460.0, "no chosen cap: the app's default 80%");
assert_eq!(efficient_watts("AMD Radeon RX 9070 XT"), Some(199.0));
assert_eq!(efficient_watts("AMD Radeon RX 9070 XT"), Some(149.0), "the 8 October grid: both floors, 149.3 W at 18.96 MH/s");
assert_eq!(efficient_watts("gfx1036"), None);
}