53 Removed incorrect observation.
source | link

I've finally accessed the Enhanced Core 2 C-states, and now that my cluster is rebooted the fan noise has dropped to almost nothing in this room. And it looks like there is a measurable drop in power consumption - my meter on 8 nodes appears to be averaging at least 5% lower (with one node still running the old kernel), but I'll try swapping the kernels out again as a test.

I've finally accessed the Enhanced Core 2 C-states, and now that my cluster is rebooted the fan noise has dropped to almost nothing in this room. And it looks like there is a measurable drop in power consumption - my meter on 8 nodes appears to be averaging at least 5% lower (with one node still running the old kernel), but I'll try swapping the kernels out again as a test.

I've finally accessed the Enhanced Core 2 C-states, and it looks like there is a measurable drop in power consumption - my meter on 8 nodes appears to be averaging at least 5% lower (with one node still running the old kernel), but I'll try swapping the kernels out again as a test.

52 Added slide from Intel presentation.
source | link

I managed to find a slide from a 2009 Intel presentation on the transitions between C-states (i.e., Deep Power Down):

Deep Power Down Technology Entry/Exit

I managed to find a slide from a 2009 Intel presentation on the transitions between C-states (i.e., Deep Power Down):

Deep Power Down Technology Entry/Exit

51 Tested QX9650 Extreme processor.
source | link

I tested this on a Core 2 Duo E2220 (Allendale), a Dual Core Pentium E5300 (Wolfdale), Core 2 Duo E7400, Core 2 Duo E8400 (Wolfdale), Core 2 Quad Q9550S (Yorkfield) and Core 2 Quad Q9550Core 2 Extreme QX9650, and I have found no issues beyond the afore-mentioned preference for state C2/C2E and C4/C4E.

  • Core 2 Solo SU3300/SU3500 (Penryn-L) are family 6, model 23 and will be detected by this driver. However, they are not Socket LGA775 so they may not support the C1E Enhanced Halt C-state. Likewise for the Core 2 Solo ULV U2100/U2200 (Merom-L). However, the intel_idle driver appears to choose the appropriate C1/C1E based on hardware support of the sub-states.
  • Core 2 Extreme QX9650 (Yorkfield) reportedly does not support C-state C2 or C4. I'm skeptical of this because this is contrary to all other documentation I've found. I ordered a used chip for testing, so I'll provide an update when it arrives.
  • After testing the Q9550Core 2 Extreme QX9650 (95WYorkfield) processor onreportedly does not support C-state C2 or C4. I have confirmed this by purchasing a Dellused Optiplex 780, I have found that although the and QX9650 Extreme processor on eBay. The processor supports C-state C4Estates C1 and C1E. With this driver modification, it spends more timethe CPU idles in state C1E instead of C1, so there is presumably some power savings. I expected to see C-state 3C3, but it is not present when using this driver so I may need to look into this further.

I tested this on a Core 2 Duo E2220 (Allendale), a Dual Core Pentium E5300 (Wolfdale), Core 2 Duo E7400, Core 2 Duo E8400 (Wolfdale), Core 2 Quad Q9550S (Yorkfield) and Core 2 Quad Q9550, and I have found no issues beyond the afore-mentioned preference for state C2/C2E and C4/C4E.

  • Core 2 Solo SU3300/SU3500 (Penryn-L) are family 6, model 23 and will be detected by this driver. However, they are not Socket LGA775 so they may not support the C1E Enhanced Halt C-state. Likewise for the Core 2 Solo ULV U2100/U2200 (Merom-L). However, the intel_idle driver appears to choose the appropriate C1/C1E based on hardware support of the sub-states.
  • Core 2 Extreme QX9650 (Yorkfield) reportedly does not support C-state C2 or C4. I'm skeptical of this because this is contrary to all other documentation I've found. I ordered a used chip for testing, so I'll provide an update when it arrives.
  • After testing the Q9550 (95W) processor on a Dell Optiplex 780, I have found that although the processor supports C-state C4E, it spends more time in C-state 3.

I tested this on a Core 2 Duo E2220 (Allendale), a Dual Core Pentium E5300 (Wolfdale), Core 2 Duo E7400, Core 2 Duo E8400 (Wolfdale), Core 2 Quad Q9550S (Yorkfield) and Core 2 Extreme QX9650, and I have found no issues beyond the afore-mentioned preference for state C2/C2E and C4/C4E.

  • Core 2 Solo SU3300/SU3500 (Penryn-L) are family 6, model 23 and will be detected by this driver. However, they are not Socket LGA775 so they may not support the C1E Enhanced Halt C-state. Likewise for the Core 2 Solo ULV U2100/U2200 (Merom-L). However, the intel_idle driver appears to choose the appropriate C1/C1E based on hardware support of the sub-states.
  • Core 2 Extreme QX9650 (Yorkfield) reportedly does not support C-state C2 or C4. I have confirmed this by purchasing a used Optiplex 780 and QX9650 Extreme processor on eBay. The processor supports C-states C1 and C1E. With this driver modification, the CPU idles in state C1E instead of C1, so there is presumably some power savings. I expected to see C-state C3, but it is not present when using this driver so I may need to look into this further.
50 Tested Q9550 processor.
source | link
49 Finished testing E8400.
source | link
48 Added update.
source | link
47 Added MWAIT tuning resources.
source | link
46 Added MWAIT tuning resources.
source | link
45 Re-wrote introduction.
source | link
44 Added link.
source | link
43 Added link to current driver.
source | link
42 Added more information on C2/C2E and C4/C4E.
source | link
41 Changed source of C-state data.
source | link
40 Changed source of C-state data.
source | link
39 Added MWAIT link.
source | link
38 Added CxE information.
source | link
37 Updated status.
source | link
36 Added C4E and C6 information.
source | link
35 Added more information regarding C2/C2E.
source | link
34 Expanded explanation of C1/C1E. Added support for Conroe processors.
source | link
33 Expanded explanation of C1/C1E.
source | link
32 Added information from linux-family.h.
source | link
31 Added information from linux-family.h.
source | link
30 Moving a table from the answer into the question.
source | link
29 Updated CPU list.
source | link