Skip to content

Your GeForce runs AI at half speed. This fixes it for 20/30/40 series cards. Up to 2× FP16 / BF16 / TF32 tensor throughput

C

Add GA106 support: GeForce RTX 3060 12GB (10de:2504)

jesvs9eeeb4c09810
diff --git a/kernel/gpu_unlock/main.c b/kernel/gpu_unlock/main.c
index d6c2a23..9ad0ea2 100644
--- a/kernel/gpu_unlock/main.c
+++ b/kernel/gpu_unlock/main.c
@@ -23,6 +23,8 @@ static const unsigned short supported_ids[] = {
0x248a, 0x248c, 0x248d, 0x248e, 0x249c, 0x249d, 0x249f, 0x24a0,
0x24a4, 0x24ac, 0x24ad, 0x24af, 0x24bf, 0x24c0, 0x24c7, 0x24c8,
0x24c9, 0x24dc, 0x24dd, 0x24df, 0x24e0,
+ /* GA106 (experimental: same family; validated on RTX 3060 12GB 0x2504) */
+ 0x2504,
};
static ASYNC_DOMAIN_EXCLUSIVE(unlock_domain);