Low power office computer?

[cRiPpLe_rOoStEr]

Originally Posted by Isaac Zachary

Not exactly true. Hard drives can actually be more reliable, or at least have a better cost to reliability relationship.

This sounds quite surprising to me.

[Isaac Zachary]

Originally Posted by cRiPpLe_rOoStEr

This sounds quite surprising to me.

I suggest watchin Linus Tech Tips.

Plus from my own personal experience I've worked on dozens of computers and have had at least 5 SSD's die that were no older than 7 years. And out of dozens of hard drives I've only had 2 that have gone bad, and they were well over 10 years old. And apparently at least one, if not both of them are repairable unlike the SSD's that are trash and there's nothing that can be done to fix them.

[Daox]

I've been playing with the new to me computer setup and am quite happy with it. Performance is definitely superior and power consumption is light years ahead of the older computer despite only being 2 years newer technology (2010 vs 2012).

After some tweaking, I was able to get idle power consumption down to about 28W. All the readings were taken via a killawatt from the wall, so it is total system power draw. Here is the process I went through:

stock idle - 33W
unplug DVD drive - 31W
swap out bronze 500W to 350W gold rated power supply - 29W
add wifi card - 29W
add HDD - 29-30W

I then started to undervolt the processor. It did quite a great job and is pretty stable at -.18V though I think I may need to increase that to -.17V. My wife mentioned it crashed once on her today. But, all these settings were tested with a cinebench run.
-.03V - 29W
-.05V - 29W
-.10V - 29W
-.15V - 28W
-.18V - 28W
-.20V - crash

While poking around in the bios for undervolting, I also increased the sleep state from S1 to S3 and went from 26W sleeping down to 2W when sleeping.

Finally, I did also monitor power usage while benchmarking in cinebench. At stock it was using about 82W. After the undervolting it dropped it down quite a bit to 70W with the same performance. I did not underclock it, but its possible I could get power consumption down even further. Of course this would sacrifice performance.

I am very happy with these results. The old computer idled around 100W and the new one now runs about 28W at idle. So, the new computer uses less power at full load than the old one did while idling! Also, if my wife leaves the computer on by accident (which is rare), its not guzzling power all night long.

I also did take a months worth of power usage and it really isn't a ton. In the past 30ish days her old computer used about 9 kWh. I'll probably throw the killawatt on it and check the new one in a month.

Thanks again Redpoint!

[freebeard]

I suggest watchin Linus Tech Tips.

Can confirm. The Winchester drive is more reliable, and uses IMHO a recoverable data structure.

Quoth DDG

How Do SSDs Work? - ExtremeTech
https://www.extremetech.com/extreme/...w-do-ssds-work
Because SSDs write data to pages but erase data in blocks, the amount of data being written to the drive is always larger than the actual update. If you make a change to a 4KB file, for example,...

The Effect Of Data Write Size and Patterns on Solid State ...
https://www.delkin.com/blog/the-effe...e-performance/
Solid State Disk Drives. Spinning mechanical disk drives have one big advantage. They are virtually unaffected by the data transfer size or mode on writes. This is because of the nature of the media. It can be overwritten without any intervening erase. The spots on the media are re-magnetized. Convenient and fast. Not so with Solid State Media.

Daox -- Do those figures include a monitor?

[cRiPpLe_rOoStEr]

Originally Posted by freebeard

The Winchester drive is more reliable

I remember it was still quite usual to refer to HDs as a Winchester around 20 years ago, but lately I haven't seen anyone doing such.

[Daox]

Originally Posted by freebeard

Daox -- Do those figures include a monitor?

Nope, that is just the computer's power usage.

[Ecky]

It's been a little while since I did undervolting with my 3570K. Here's some of my data:






I found that, although load power decreased considerably, that just by enabling voltage adjustment, it increased my idle power by around 10 watts, and there was no way to get it back down. Ultimately, since my computer spends more time idle than loaded, I set everything back to stock.

[Daox]

Interesting results Ecky! Thanks for sharing. I also noticed some odd behavior while undervolting. It wasn't always 100% consistent. Every once in a while it would idle down lower on automatic control, but not always...