Re: cool high-perf drives?

Mike wrote:
"timeOday" <timeOday-UNSPAM@xxxxxxxxxxxx> wrote in message news:Y7CdndeLULvgjszZnZ2dnUVZ_v6dnZ2d@xxxxxxxxxxxxxx

Mike wrote:

Is there a drive that has a good balance of high perf and low ambient power output?


I searched google for HTS7201 and got 0 results.


Yeah, I did the same thing. It's the model off of the drive itself.
It came when I ordered my Dell with a 7200 -- I'm not sure if it's an OEM designation for the 7K100 or some other drive.


I just got a Hitachi Travelstar 7K100 and it seems, if anything, just slightly warmer than the old 4200 RPM drive. According to ACPI, the battery discharge rate is now 1% higher than with the 4200 drive, and with a new battery my T40 gets almost 7 hours battery life, so it certainly is not generating heat like you're describing.


I'm getting about 4.5 hrs if I use the big (85wh) battery.
(Not sure about the T40- my version of the 5150 is a dual-core T2400@xxxxxxx, 2GB RAM, 1650x1280 ati mobility, and the 7200 drive - so it's pretty power-hungry.)

I'm amazed that it could generate so much heat from the hard drive alone and still last 4.5 hours. I guess you're sure it's the drive and not the chipset or graphics adapter or something?

I'm underclocking by 50% on battery and doing all the power-saving I can on the HD.

I'm not sure if doing spin-down is worth it or not? I know it takes more power and drive stress to spin up, so I'm not sure where the tradeoff time for spindown (re:heat and power) is.


I use Linux and without a lot of modification, it has daemons such as the system log that write to the drive every minute or so, and for that reason I don't bother.

I did find a kernel modification once that would cache all writes, and putting the drive to sleep only boosted the runtime projection by about 10 minutes IIRC which wasn't worth it to me.

Are you using Linux? If so, and have ACPI working, you can get the battery drain rate:

grep 'present rate:' /proc/acpi/battery/BAT0/state
present rate: 11232 mW

Compare with the drive spun up vs asleep.


Why do you say rotational latency doesn't go hand-in-hand with access time? Rotational latency sets a lower bound for access time.


I was talking specifically about track-to-track (seek) time - i.e., the actuator - then of course you're right you'll have to wait for 1/2 a spin on average (4.2ms @ 7200rpm)
I think seek time is at least as a factor as is rotational latecy in general usage. (Apparently even among manufactures, "acess time" doesn't always mean command+seek+settle+latency.)

I agree those are all factors, but they're not additive because, surely, the seek and the rotational latency occur in parallel. I don't know that for a fact but it would seem crazy to do otherwise.

(I may not have been careful and may have been comparing apples of one drive to oranges of another.)


Finally, if you are thrashing I suggest getting more RAM instead of a faster drive. Swapping is pointless with RAM so cheap.


Well, there's real thashing due to VM (I'm okay here for the moment as I have 2GB) and there's thrashing because of Windows and teams of random apps hitting the disk loading DLLs and writing state, Outlook skattering stuff all over a big PST file, anti-virus checking the files that the spyware checker is checking (and probably visa-versa) -- and who knows what else - that kind of thrashing I'm still getting :)

I run Outlook and Office under VMWare too. Subjectively, the faster drive seems to speed up the Windows guest quite a bit more than the Linux host. Perhaps storing a Windows filesystem over the Linux filesystem introduces two layers of fragmentation, and the "grow on demand" VMWare virtual disk files probably cause more fragmentation on top of that.
.

Relevant Pages

  • Re: UPS Tab & wiring?
    ... to my car battery and thus can be powered on even while the engine is not ... I'd like Windows to power down after 30 minutes of running on ... signaling does windows use for it's generic ups support. ...
    (microsoft.public.windowsxp.general)
  • Re: MSI Windriver running OS X
    ... being you have to scroll the screen about etc? ... or is this some windows thing? ... better than the letter box we have with Linux. ... When I can get the better battery, ...
    (uk.comp.sys.mac)
  • Re: Ext4 and the "30 second window of death"
    ... you're going to be burning a lot of power. ... Linux is still trying to catch up with Windows when it ... comes to battery life. ... But honestly, if we all start blaming the filesystem for it, I don't ...
    (Linux-Kernel)
  • Re: PC Reboots randomly
    ... Windows started to lock up too. ... >>time (in Linux) to check my mail before it resets itself. ... Also check the specs for the surge protector maybe you have ... Power is flaky where I live, no question about that, but the system ...
    (comp.os.linux.hardware)
  • Re: Toshiba Satelitte 3000-400
    ... Battery life with optional 2nd battery: ... wobei der Prozessor schon gefordert ist. ... Microsoft Windows Millennium, Toshiba Hibernation Utility, online user ... Auf Wunsch auch mit Linux ...
    (de.comp.sys.notebooks)