Confused about lowest-power AMD CPUs

All about them.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

corevalue
Posts: 8
Joined: Thu Nov 02, 2006 3:51 pm

Post by corevalue » Fri Nov 03, 2006 6:27 am

The "Performance Rating" (3200+ and 3500+) doesn't actually refer to the clock speed; it's a relative benchmark based on some reference proc, can't remember which. So the Semp 3200+ runs @ 1.8Ghz and the A64 3500+ runs at 2.2Ghz.
Yep, IIRC, the Sempron is (supposedly) rated in equivalent Celeron MHz. From all the benchmarks I've been able to find, the AM2 Manila core is actually a bit slower than the 754 Palermo of the same "speed".

There is a thread devoted to this board (Abit NF-M2 nView) in this forum; apparently it doesn't allow you to undervolt lower than 1.35V in BIOS, which is bad for power consumption; but CnQ + RMClock/CrystalCPUID should enable you to undervolt by software.
Thanks, found the thread.

smilingcrow
*Lifetime Patron*
Posts: 1809
Joined: Sat Apr 24, 2004 1:45 am
Location: At Home

Post by smilingcrow » Fri Nov 03, 2006 2:00 pm

corevalue wrote:Hi Matt, Im a Newbie here, but I got an amost similar dilemna to you. I need to build a PC for use as home entertainment centre, and to do "real stuff", like CAD occasionally. Snag is, it's going on a boat where power is SEVERELY limited. So it's got to be capable, but must be (a) quiet and (b) frugal.
If you are looking at using a PC in such a frugal environment then the Intel on Desktop platform seems the best bet. If you use a 2.5â€

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Fri Nov 03, 2006 9:18 pm

corevalue wrote:
The "Performance Rating" (3200+ and 3500+) doesn't actually refer to the clock speed; it's a relative benchmark based on some reference proc, can't remember which. So the Semp 3200+ runs @ 1.8Ghz and the A64 3500+ runs at 2.2Ghz.
Yep, IIRC, the Sempron is (supposedly) rated in equivalent Celeron MHz.
It is? If *I* recall correctly it's rated based on the original K6 core. Time to do some actual research...

Mariner
Friend of SPCR
Posts: 260
Joined: Thu Jan 06, 2005 11:25 am

Post by Mariner » Sat Nov 04, 2006 2:52 am

Devonavar wrote:It is? If *I* recall correctly it's rated based on the original K6 core. Time to do some actual research...
I'm pretty sure you are correct in this recollection.

Bear in mind that the Semprons had much better performance than the crippled netburst Celerons.

The new 65nm Yonah/(Conroe?) based Celeron M's look pretty decent however.

corevalue
Posts: 8
Joined: Thu Nov 02, 2006 3:51 pm

Post by corevalue » Sun Nov 05, 2006 3:51 pm

[quote="smilingcrow"]If you use a 2.5â€

autoboy
Posts: 1008
Joined: Fri Dec 10, 2004 8:10 pm
Location: San Jose, California

Post by autoboy » Sun Nov 05, 2006 4:34 pm

Did you ever consider using a laptop for that boat? Might be a better, cheaper, smaller, and easier option. They can be just as fast as a basic desktop.

smilingcrow
*Lifetime Patron*
Posts: 1809
Joined: Sat Apr 24, 2004 1:45 am
Location: At Home

Post by smilingcrow » Mon Nov 06, 2006 2:06 am

corevalue wrote:
The standby power consumption for my desktop systems is typically 3 or 4W, which is reduced by only 1W when I switch them off. In other words, they need to be disconnected from the mains power when they are off to be the brightest shade of green.
Very impressive - does that mean your system uses only 1W in idle, the rest is the PSU? I checked the PC's in my house, one of them was 19 Watts when off! The best was only 5 Watts.
First off, I don’t know how accurate my power meter is when working with such small wattages, so there’s potentially a lot of room for error here.
When a PC is switched off but still connected to the mains power, it still consumes energy, most of which is possibly consumed by the power supply as you suggested.
The extra 1W is the difference between standby and off and nothing to do with idle. The only way to get to 0 watts is to disconnect the PC from the mains power.

I’ve noticed that even my PicoPSU/Core Solo system which consumes 20W at idle, still consumes 3-4W on standby.
My Core Duo laptop registers 0 Watts on standby though and my TV 1W, so maybe the power meter is up to the job.

This 3 or 4W on standby/off is not insignificant, as if my low power desktop is on 4 hours a day and off or on standby the other 20 hours, its daily power consumption is:

4 * 20 = 80 +
20 * 4 = 80
= 160W

Half the daily power consumption is from when the PC is not even on. :o

I have a 20â€

jaganath
Posts: 5085
Joined: Tue Sep 20, 2005 6:55 am
Location: UK

Post by jaganath » Mon Nov 06, 2006 2:26 am

I don’t want to give up on a high resolution screen as it’s such a productivity boost. Also, if people are going to move to lower power PCs, they generally aren’t going to accept a large compromise. That’s one thing I’m very clear on.
Very true. People like to be green, but not if it means they have to make huge changes to their lifestyle or accept a reduction in their standard of living in some way.
I’m not going to be around these forums so much for the immediate future as I need to focus on the Green PC website that I’m developing. See you later guys (& gals).
Best of luck in your new venture. You have done an enormous amount of tremendously valuable testing and experimentation, all fully documented here, and it wouldn't surprise me if the editorial team saw fit to make you an honorary staff member! :)

Poodle
Posts: 254
Joined: Fri Apr 01, 2005 3:17 am
Location: Gothenburg, Sweden.

Post by Poodle » Tue Nov 07, 2006 1:12 am

W x h = Wh ;)


@ SmilingCrow: Seriously, you must have like the best Performance/W system in the World. :)

smilingcrow
*Lifetime Patron*
Posts: 1809
Joined: Sat Apr 24, 2004 1:45 am
Location: At Home

Post by smilingcrow » Tue Nov 07, 2006 4:43 am

jaganath wrote:
I don’t want to give up on a high resolution screen as it’s such a productivity boost. Also, if people are going to move to lower power PCs, they generally aren’t going to accept a large compromise. That’s one thing I’m very clear on.
Very true. People like to be green, but not if it means they have to make huge changes to their lifestyle or accept a reduction in their standard of living in some way.
The sooner we start voluntarily making many smaller and less disruptive compromises, the less chance that we have to make so many larger ones down the road, some of which will be outside our control.
jaganath wrote:Best of luck in your new venture. You have done an enormous amount of tremendously valuable testing and experimentation, all fully documented here, and it wouldn't surprise me if the editorial team saw fit to make you an honorary staff member! :)
Thank you for your kind words.
Poodle wrote:@ SmilingCrow: Seriously, you must have like the best Performance/W system in the World. :)
That’s a nice challenge to throw out to people; see who can build the most power efficient PC in the way that people currently see who can over-clock the highest. You could have different categories: one could be how powerful a PC can you build using a PicoPSU 120W. That could cover the green over-clocking section. :D

matt_garman
*Lifetime Patron*
Posts: 541
Joined: Sun Jan 04, 2004 11:35 am
Location: Chicago, Ill., USA
Contact:

Post by matt_garman » Tue Nov 07, 2006 9:13 pm

FYI, to all. I finally snagged a S754 Sempron 2800 and a Biostar Tforce 6100 (newegg had a special on the combo not too long ago, $90US for both).

So I went to work on my fileserver. My notes are posted below. In short: I went from about 220 W running to about 130 W. I made many other mods too (not just mobo+cpu).

Feel free to ask questions. The measurements were all taken with a Kill-a-Watt.

Code: Select all

septictank
==========

Note:
    Highest temp seen for WD4000YR hard drives:
        - no fans (except CPU)
        - updatedb and emerge --sync
        - ~45 deg C

Config 1:
    512 MB RAM
    Athlon XP 1800
    Thermaltake HSF
    Stock BIOS config
    4x400 GB WD4000YR SATA HDs
    2x120 GB WD1200JB PATA HDs
    1x160 GB Samsung SP1614N PATA HD
    1x18 GB Quantum Atlas 10K2 SCSI drive
    2x120 mm high speed HD fans (Delta)
    2x92 mm high speed HD fans (Delta)
    1x120x35 mm high speed case fan (Delta)

    Max wattage seen at power-on: 335 W
    Steady-state w/SETI: 211 W
    Steady-state w/o SETI: 177 W

    All fans HD temps: 25--27 deg C

    Baseline: no SETI@home, ~180 W
        - Stopped front two 92mm HD fans: 160 W
        - All above plus stopped front two 120mm HD fans: 145 W
        - All above plus stopped rear 120mm case fan: 140 W

    Baseline: no SETI@home, ~140 W (no fans except CPU)
        - Ran bonnie++ -d . -u 0 on /media/video/stress
        - ~160 W

Config 2:
    512 MB RAM
    Sempron64 2800 (S754)
    Stock CPU cooler
    Stock BIOS config
    4x400 GB WD4000YR SATA HDs
    1x18 GB Quantum Atlas 10K2 SCSI drive
    1x36 GB Fujitsu MAP3367NP 10K SCSI drive
    2x120 mm low speed HD fans (Yate Loon)
    1x120 mm Antec Tri-Cool case fan (at Medium)

    (i.e. removed three PATA drives, added SCSI drive, replaced 120mm
    fans, removed front 92mm fans)

    HD temps: 30--35 deg C

    Max wattage seen at power-on: ???
    Steady-state w/burnK7: 160 W
    Steady-state w/o SETI: 132 W

Config 3:
    Same as #2, except:
    Reduced CPU voltage to 1.1

    Max wattage seen at power-on: ???
    Steady-state w/burnK7: 142 W
    Steady-state w/o SETI: 132 W


Config 4:
    Same as #2, except:
    Reduced CPU voltage to 0.925

    Max wattage seen at power-on: 214
    Steady-state w/burnK7: 137 W
    Steady-state w/o SETI: 126 W

    Running "memtester 400" and updatedb and burnK7: 145 W

nightmorph
Posts: 316
Joined: Thu Aug 10, 2006 11:07 am

Post by nightmorph » Tue Nov 07, 2006 10:32 pm

matt_garman wrote:

Code: Select all

        - updatedb and emerge --sync
Fantastic, another Gentoo Linux user. :D

Post Reply