cool high-perf drives?

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

I ordered a 7200rpm drive (turns out to be a hitachi HTS7201 SATA) and
although it seems fast, it runs really hot.
(Gets to about 53 degrees C.) This is hot enough that it can't be used on
the lap directly, even with all power saving modes on.
I imagine the seagate 7200.1 would be similar, given it's specs.

It seems like the 5400rpm Western Digital Scorpio (1200BEVS) drive has
gotten good ratings for being reasonably performant and quiet/cool -
anyone have any direct experience?

I don't care as much about rotational lateny and bust throughput as I do
acess-time, which is probably the most important when things start
thrashing, but it looks like these specs go hand-in-hand even though they
need not, as the 7200 has ~9ms, where the 5400's are about 12ms.

thanks,
m

Mike wrote:
> Is there a drive that has a good balance of high perf and low ambient power
> output?
>
> I ordered a 7200rpm drive (turns out to be a hitachi HTS7201 SATA) and
> although it seems fast, it runs really hot.
> (Gets to about 53 degrees C.) This is hot enough that it can't be used on
> the lap directly, even with all power saving modes on.
> I imagine the seagate 7200.1 would be similar, given it's specs.
>
> It seems like the 5400rpm Western Digital Scorpio (1200BEVS) drive has
> gotten good ratings for being reasonably performant and quiet/cool -
> anyone have any direct experience?
>
> I don't care as much about rotational lateny and bust throughput as I do
> acess-time, which is probably the most important when things start
> thrashing, but it looks like these specs go hand-in-hand even though they
> need not, as the 7200 has ~9ms, where the 5400's are about 12ms.
>
> thanks,
> m
>
>

I searched google for HTS7201 and got 0 results.

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.

As to the performance increase over the 4200 rpm drive, it's huge when
copying a directory tree (from 2m17s to 1m17s), but fairly small during
bootup (54s to 47s) and very small when compiling a program (58s vs
56s). It is subjectively more responsive when loading a small program
while a large program is already loading.

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

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

"Mike" <vimakefile@yahoo.com> wrote:
>Is there a drive that has a good balance of high perf and low ambient power
>output?

>(Gets to about 53 degrees C.)

Kinda depends on the laptop, and the cooling it provides. My Dell
Latitude D610 has a Hitachi 7K100 (PATA) drive that runs around 45C,
FWIW...

You can tell something about the heat buildup by looking at the power
consumption numbers in the manufacturers detailed specs on their
WWWebsites.

"timeOday" <timeOday-UNSPAM@theknack.net> wrote in message
news:Y7CdndeLULvgjszZnZ2dnUVZ_v6dnZ2d@comcast.com. ..
> 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@1.83mhz, 2GB RAM, 1650x1280 ati mobility, and the 7200 drive - so it's
pretty power-hungry.)
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.

> 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 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

m

Mike wrote:
> "timeOday" <timeOday-UNSPAM@theknack.net> wrote in message
> news:Y7CdndeLULvgjszZnZ2dnUVZ_v6dnZ2d@comcast.com. ..
>
>>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@1.83mhz, 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.