Power consumption HZ250 vs. HZ1000

Power consumption HZ250 vs. HZ1000

[Marc Ballarin]

Hi,
I did some measurements in order to compare power drain with HZ250 and
HZ1000.
To measure the actual drain, I used the "smart" battery's internal measurement.
(Available with acpi-sbs in /proc/acpi/sbs/SBS0/SB0/state.)
No clue how accurate this is.

Here some battery details, in case someone knows:
charge reporting error:  25%
SB specification:        v1.1 (with PEC)
manufacturer name:       Panasonic
manufacture date:        2004-11-27
device name:             02ZL
device chemistry:        Lion

Kernel: 2.6.13-rc3-mm1 + acpi-sbs

CPU:
cpu family	: 6
model		: 13
model name	: Intel(R) Pentium(R) M processor 1.60GHz
stepping		: 6

The "ondemand" governor was running, using acpi_cpufreq. (Idle at 600MHz).

Systems was running X11/KDE to get a more or less realistic scenario. No
cron jobs, network traffic or additional applications. WLAN and built-in
display were disabled completely, all fans and LEDs were off, internal hard
disc was running. Additional peripherals: external keyboard, mouse, display
and externally-powered hard disk (USB).

The results are quite simple:
In both configurations the current settles between 727-729 mA
(Voltage ~16.5 V).

Some issues:

- C-states look strange:
active state:            C2
max_cstate:              C8
bus master activity:     00887fff
states:
    C1:                  type[C1] promotion[C2] demotion[--]   latency[000] usage[00000010]
  *C2:                  type[C2] promotion[C3] demotion[C1] latency[001] usage[01367471]
    C3:                  type[C3] promotion[--]   demotion[C2] latency[085] usage[00000000]

- I don't know, how much polling of the battery affects results. Reads always
block for ~10 seconds, and I used this behaviour for rate-limiting.

- Is this approach valid at all?

- I could repeat the test in single user mode with internal hard disc turned off.

Regards

Re: Power consumption HZ250 vs. HZ1000

[Pavel Machek]

Hi!

> I did some measurements in order to compare power drain with HZ250 and
> HZ1000.
> To measure the actual drain, I used the "smart" battery's internal measurement.
> (Available with acpi-sbs in /proc/acpi/sbs/SBS0/SB0/state.)
> No clue how accurate this is.
> 
> Here some battery details, in case someone knows:
> charge reporting error:  25%
> SB specification:        v1.1 (with PEC)
> manufacturer name:       Panasonic
> manufacture date:        2004-11-27
> device name:             02ZL
> device chemistry:        Lion
> 
> Kernel: 2.6.13-rc3-mm1 + acpi-sbs
> 
> CPU:
> cpu family	: 6
> model		: 13
> model name	: Intel(R) Pentium(R) M processor 1.60GHz
> stepping		: 6
> 
> The "ondemand" governor was running, using acpi_cpufreq. (Idle at 600MHz).
> 
> Systems was running X11/KDE to get a more or less realistic scenario. No
> cron jobs, network traffic or additional applications. WLAN and built-in
> display were disabled completely, all fans and LEDs were off, internal hard
> disc was running. Additional peripherals: external keyboard, mouse, display
> and externally-powered hard disk (USB).


USB devices prevent entering C3 and any interesting powersaving,
try without USB...
-- 
64 bytes from 195.113.31.123: icmp_seq=28 ttl=51 time=448769.1 ms         

Re: Power consumption HZ250 vs. HZ1000

[Marc Ballarin]

On Mon, 25 Jul 2005 17:53:22 +0200
Pavel Machek <pavel@ucw.cz> wrote:
 
> USB devices prevent entering C3 and any interesting powersaving,
> try without USB...


Hmm, just did. I even tried the rather minimalistic configuration below.
Still no C3. (And what seems even stranger: no C1.)

Is this a BIOS Issue?

Regards


Tested config:

CONFIG_X86=y
CONFIG_MMU=y
CONFIG_UID16=y
CONFIG_GENERIC_ISA_DMA=y
CONFIG_GENERIC_IOMAP=y

CONFIG_EXPERIMENTAL=y
CONFIG_CLEAN_COMPILE=y
CONFIG_BROKEN_ON_SMP=y
CONFIG_INIT_ENV_ARG_LIMIT=32

CONFIG_LOCALVERSION="-hztest"
CONFIG_SWAP=y
CONFIG_SYSVIPC=y
CONFIG_POSIX_MQUEUE=y
CONFIG_BSD_PROCESS_ACCT=y
CONFIG_BSD_PROCESS_ACCT_V3=y

CONFIG_SYSCTL=y
CONFIG_HOTPLUG=y
CONFIG_KOBJECT_UEVENT=y
CONFIG_IKCONFIG=y
CONFIG_IKCONFIG_PROC=y
CONFIG_KALLSYMS=y
CONFIG_PRINTK=y
CONFIG_BUG=y
CONFIG_BASE_FULL=y
CONFIG_FUTEX=y
CONFIG_EPOLL=y
CONFIG_SHMEM=y
CONFIG_CC_ALIGN_FUNCTIONS=0
CONFIG_CC_ALIGN_LABELS=0
CONFIG_CC_ALIGN_LOOPS=0
CONFIG_CC_ALIGN_JUMPS=0
CONFIG_BASE_SMALL=0

CONFIG_X86_PC=y
CONFIG_MPENTIUMM=y
CONFIG_X86_CMPXCHG=y
CONFIG_X86_XADD=y
CONFIG_X86_L1_CACHE_SHIFT=6
CONFIG_RWSEM_XCHGADD_ALGORITHM=y
CONFIG_GENERIC_CALIBRATE_DELAY=y
CONFIG_X86_WP_WORKS_OK=y
CONFIG_X86_INVLPG=y
CONFIG_X86_BSWAP=y
CONFIG_X86_POPAD_OK=y
CONFIG_X86_GOOD_APIC=y
CONFIG_X86_INTEL_USERCOPY=y
CONFIG_X86_USE_PPRO_CHECKSUM=y
CONFIG_PREEMPT_NONE=y
CONFIG_X86_TSC=y

CONFIG_EDD=y
CONFIG_NOHIGHMEM=y
CONFIG_SELECT_MEMORY_MODEL=y
CONFIG_FLATMEM_MANUAL=y
CONFIG_FLATMEM=y
CONFIG_FLAT_NODE_MEM_MAP=y
CONFIG_MTRR=y
CONFIG_REGPARM=y
CONFIG_SECCOMP=y
CONFIG_HZ_250=y
CONFIG_HZ=250
CONFIG_PHYSICAL_START=0x100000

CONFIG_PM=y

CONFIG_ACPI=y
CONFIG_ACPI_BOOT=y
CONFIG_ACPI_INTERPRETER=y
CONFIG_ACPI_SLEEP=y
CONFIG_ACPI_SLEEP_PROC_FS=y
CONFIG_ACPI_BUTTON=y
CONFIG_ACPI_VIDEO=y
CONFIG_ACPI_HOTKEY=y
CONFIG_ACPI_FAN=y
CONFIG_ACPI_PROCESSOR=y
CONFIG_ACPI_THERMAL=y
CONFIG_ACPI_BLACKLIST_YEAR=0
CONFIG_ACPI_BUS=y
CONFIG_ACPI_EC=y
CONFIG_ACPI_POWER=y
CONFIG_ACPI_PCI=y
CONFIG_ACPI_SYSTEM=y
CONFIG_X86_PM_TIMER=y

CONFIG_PCI=y
CONFIG_PCI_GOANY=y
CONFIG_PCI_BIOS=y
CONFIG_PCI_DIRECT=y
CONFIG_PCI_MMCONFIG=y
CONFIG_PCI_NAMES=y
CONFIG_ISA_DMA_API=y
CONFIG_ISA=y

CONFIG_BINFMT_ELF=y
CONFIG_BINFMT_MISC=y

CONFIG_NET=y

CONFIG_PACKET=y
CONFIG_UNIX=y
CONFIG_INET=y
CONFIG_IP_MULTICAST=y
CONFIG_IP_FIB_HASH=y
CONFIG_SYN_COOKIES=y
CONFIG_IP_TCPDIAG=y
CONFIG_TCP_CONG_BIC=y

CONFIG_PREVENT_FIRMWARE_BUILD=y

CONFIG_PNP=y

CONFIG_PNPACPI=y

CONFIG_BLK_DEV_RAM_COUNT=16
CONFIG_INITRAMFS_SOURCE=""

CONFIG_IOSCHED_NOOP=y
CONFIG_IOSCHED_AS=y
CONFIG_IOSCHED_DEADLINE=y
CONFIG_IOSCHED_CFQ=y

CONFIG_IDE=y
CONFIG_BLK_DEV_IDE=y

CONFIG_BLK_DEV_IDEDISK=y
CONFIG_IDEDISK_MULTI_MODE=y
CONFIG_BLK_DEV_IDECD=y

CONFIG_IDE_GENERIC=y
CONFIG_BLK_DEV_IDEPCI=y
CONFIG_IDEPCI_SHARE_IRQ=y
CONFIG_BLK_DEV_GENERIC=y
CONFIG_BLK_DEV_IDEDMA_PCI=y
CONFIG_IDEDMA_PCI_AUTO=y
CONFIG_BLK_DEV_PIIX=y
CONFIG_BLK_DEV_IDEDMA=y
CONFIG_IDEDMA_AUTO=y

CONFIG_INPUT=y

CONFIG_INPUT_MOUSEDEV=y
CONFIG_INPUT_MOUSEDEV_SCREEN_X=1280
CONFIG_INPUT_MOUSEDEV_SCREEN_Y=800

CONFIG_INPUT_KEYBOARD=y
CONFIG_KEYBOARD_ATKBD=y

CONFIG_SERIO=y
CONFIG_SERIO_I8042=y
CONFIG_SERIO_LIBPS2=y

CONFIG_VT=y
CONFIG_VT_CONSOLE=y
CONFIG_HW_CONSOLE=y

CONFIG_UNIX98_PTYS=y

CONFIG_VIDEO_SELECT=y

CONFIG_VGA_CONSOLE=y
CONFIG_DUMMY_CONSOLE=y

CONFIG_SPEAKUP_DEFAULT="none"

CONFIG_USB_ARCH_HAS_HCD=y
CONFIG_USB_ARCH_HAS_OHCI=y

CONFIG_EXT3_FS=y
CONFIG_EXT3_FS_XATTR=y
CONFIG_EXT3_FS_POSIX_ACL=y
CONFIG_EXT3_FS_SECURITY=y
CONFIG_JBD=y
CONFIG_FS_MBCACHE=y
CONFIG_FS_POSIX_ACL=y

CONFIG_DNOTIFY=y

CONFIG_PROC_FS=y
CONFIG_PROC_KCORE=y
CONFIG_SYSFS=y
CONFIG_DEVPTS_FS_XATTR=y
CONFIG_DEVPTS_FS_SECURITY=y
CONFIG_TMPFS=y
CONFIG_RAMFS=y

CONFIG_MSDOS_PARTITION=y

CONFIG_NLS=y
CONFIG_NLS_DEFAULT="iso8859-1"
CONFIG_NLS_CODEPAGE_437=y
CONFIG_NLS_CODEPAGE_850=y
CONFIG_NLS_CODEPAGE_1250=y
CONFIG_NLS_ASCII=y
CONFIG_NLS_ISO8859_1=y
CONFIG_NLS_ISO8859_15=y
CONFIG_NLS_UTF8=y

CONFIG_LOG_BUF_SHIFT=14
CONFIG_DEBUG_BUGVERBOSE=y
CONFIG_EARLY_PRINTK=y

CONFIG_GENERIC_HARDIRQS=y
CONFIG_GENERIC_IRQ_PROBE=y
CONFIG_X86_BIOS_REBOOT=y
CONFIG_PC=y

Re: Power consumption HZ250 vs. HZ1000

[dean gaudet]

On Mon, 25 Jul 2005, Marc Ballarin wrote:

> Hmm, just did. I even tried the rather minimalistic configuration below.
> Still no C3. (And what seems even stranger: no C1.)

there's no point to going into C1 if the C2 entry/exit latencies are 
acceptable.  (winxp generally never uses C1 if C2 is available and within 
the specs msft dictates for C2 latencies...)

-dean

Re: Power consumption HZ250 vs. HZ1000

[Tony Lindgren]

* Pavel Machek <pavel@ucw.cz> [050725 08:57]:
> 
> USB devices prevent entering C3 and any interesting powersaving,
> try without USB...

Why do USB devices prevent C3? If it was because of the timer polling
in the root hub, I believe that should be fixed now.

Or is there some other reason?

Tony

Re: Power consumption HZ250 vs. HZ1000

[Pavel Machek]

Hi!

> > USB devices prevent entering C3 and any interesting powersaving,
> > try without USB...
> 
> Why do USB devices prevent C3? If it was because of the timer polling
> in the root hub, I believe that should be fixed now.
> 
> Or is there some other reason?

Yes. UHCI zas keeps doing DMA all the time.... It can be worked
around, but it means proper usb powermanagment support.
								Pavel
-- 
teflon -- maybe it is a trademark, but it should not be.

Re: Power consumption HZ250 vs. HZ1000

[Tony Lindgren]

* Pavel Machek <pavel@ucw.cz> [050727 01:00]:
> Hi!
> 
> > > USB devices prevent entering C3 and any interesting powersaving,
> > > try without USB...
> > 
> > Why do USB devices prevent C3? If it was because of the timer polling
> > in the root hub, I believe that should be fixed now.
> > 
> > Or is there some other reason?
> 
> Yes. UHCI zas keeps doing DMA all the time.... It can be worked
> around, but it means proper usb powermanagment support.

OK, thanks for clarifying that.

Tony

Re: Power consumption HZ250 vs. HZ1000

[Bill Davidsen]

Marc Ballarin wrote:
> Hi,
> I did some measurements in order to compare power drain with HZ250 and
> HZ1000.
> To measure the actual drain, I used the "smart" battery's internal measurement.
> (Available with acpi-sbs in /proc/acpi/sbs/SBS0/SB0/state.)
> No clue how accurate this is.
> 
> Here some battery details, in case someone knows:
> charge reporting error:  25%
> SB specification:        v1.1 (with PEC)
> manufacturer name:       Panasonic
> manufacture date:        2004-11-27
> device name:             02ZL
> device chemistry:        Lion
> 
> Kernel: 2.6.13-rc3-mm1 + acpi-sbs
> 
> CPU:
> cpu family	: 6
> model		: 13
> model name	: Intel(R) Pentium(R) M processor 1.60GHz
> stepping		: 6
> 
> The "ondemand" governor was running, using acpi_cpufreq. (Idle at 600MHz).
> 
> Systems was running X11/KDE to get a more or less realistic scenario. No
> cron jobs, network traffic or additional applications. WLAN and built-in
> display were disabled completely, all fans and LEDs were off, internal hard
> disc was running. Additional peripherals: external keyboard, mouse, display
> and externally-powered hard disk (USB).
> 
> The results are quite simple:
> In both configurations the current settles between 727-729 mA
> (Voltage ~16.5 V).
> 
> Some issues:
> 
> - C-states look strange:
> active state:            C2
> max_cstate:              C8
> bus master activity:     00887fff
> states:
>     C1:                  type[C1] promotion[C2] demotion[--]   latency[000] usage[00000010]
>   *C2:                  type[C2] promotion[C3] demotion[C1] latency[001] usage[01367471]
>     C3:                  type[C3] promotion[--]   demotion[C2] latency[085] usage[00000000]
> 
> - I don't know, how much polling of the battery affects results. Reads always
> block for ~10 seconds, and I used this behaviour for rate-limiting.
> 
> - Is this approach valid at all?
> 
> - I could repeat the test in single user mode with internal hard disc turned off.
> 
I think what you did is the most valid test. Testing with any 
configuration other than the one you actually use may give some specious 
appearance of saving, but unless you actually use the system in the 
reduced configuration I don't see how you would save anything by going 
to the lower clock.

A configuration without any external kb/mouse connections would be 
something which many people do use, so that would be meaningful to some 
people if not you. I'm guessing it won't change much, but to satisfy the 
people who think USB precludes power saving you could make the test.

-- 
    -bill davidsen (davidsen@tmr.com)
"The secret to procrastination is to put things off until the
  last possible moment - but no longer"  -me

RE: Power consumption HZ250 vs. HZ1000

[Brown, Len]

yes, this approach is valid.
I've done the exact same measurements for 100HZ vs 1000HZ.
For currently shipping laptops, I didn't see a significant
difference.,

Note that the quality of the instrumentation on
the battery can vary widely, and so if you really
want the best numbers you need to start from a fully
charged battery and run it until the battery dies.

Also, for the most controlled experiment, you can
run in single user mode with no network, no USB plugged
in, and either "performance" or "powersave" governors.
If you don't get into C3 on this box nearly all the time
then something is wrong.

cheers,
-Len