Quiet and Cool High Capacity HDs
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Quiet and Cool High Capacity HDs
what are peoples recommendations these days...
i need 8 drives (in raid 6) currently i'm running with a mix of 400GB drives a couple of WD (500GBs) and a lot of Seagate 7200.8 / 7200.9 drives...
i'm looking to shift capacity up a notch..
maybe to 750Gb drives of even 1TB if that hitachi drive is any good
as the 8 drives are the loudest items in my system every db counts...
so anything above 400GB, ideally with less power / heat output than my WD 500 KS (they seem to run quite hot in proximity to each other)
and with less noise than my 400GB Seagates....
i don't ask for much, i know!
i need 8 drives (in raid 6) currently i'm running with a mix of 400GB drives a couple of WD (500GBs) and a lot of Seagate 7200.8 / 7200.9 drives...
i'm looking to shift capacity up a notch..
maybe to 750Gb drives of even 1TB if that hitachi drive is any good
as the 8 drives are the loudest items in my system every db counts...
so anything above 400GB, ideally with less power / heat output than my WD 500 KS (they seem to run quite hot in proximity to each other)
and with less noise than my 400GB Seagates....
i don't ask for much, i know!
-
- Posts: 35
- Joined: Wed Aug 16, 2006 6:07 pm
- Location: ohio usa
have you checked the recommended hard drives section recommended hard drives section
not many 3.5 inchers but its a start, the 1/2 tb caviar se16 stands out as it is the "only high capicity drive spcr feels comfortable reccomending"
a raid array like that is going to be some what loud no matter what. what case do you have these in btw?
not many 3.5 inchers but its a start, the 1/2 tb caviar se16 stands out as it is the "only high capicity drive spcr feels comfortable reccomending"
a raid array like that is going to be some what loud no matter what. what case do you have these in btw?
-
- Posts: 606
- Joined: Wed Oct 13, 2004 4:03 pm
- Location: Albany, GA USA
Do a forum search for Samsung HD501LJ. This drive has not been reviewed yet by SPCR but members here seem to prefer it to the WD drives. I have two of these drives suspended in an Antec Solo and the drives are extremely quiet and run cool. I just got a promo code from Newegg (EMC529DHD50) which results in a price of $109.99.
A shop in Belgium is claiming new 750 GB & 1 TB Samsung 3.5" drives are on their way (HD752LJ & HD102UJ). Probably 250 GB platters from a new T250 Series. samsung.com already shows single-platter HD250HJ and HD200HJ S250 Series.
the case i have is a Chenbro SR106 (i think, could be sr107) it has 8 hot-swap sata drive bays (cooled by 2*120mm fans - currently papst 4412FGL - moving to scythe 120mm E model fans)
currently it looks like either more 5000KS (is the AAKS any cooler?)
or the samsung 501LJ - do you know which of these runs cooler?
i take it the current higher capacity drives aren't any good for silence? (750GB Seagate, 1TB Hitachi)
currently it looks like either more 5000KS (is the AAKS any cooler?)
or the samsung 501LJ - do you know which of these runs cooler?
i take it the current higher capacity drives aren't any good for silence? (750GB Seagate, 1TB Hitachi)
WDs are among the coolest of high-capacity HDDs. Hitachi is cooler than WD, though, but only if the HDDs are in light use.
I cannot comment on noise, because there's no (noisewise) reputable reviews of T7K500 or 7K1000. Anandtech said 7K1000 was the most silent HDD ever reviewed (AAM enabled) but I think that may be a little exagerated. StorageReview said it's quite noisy but only tested it with AAM disabled (fast seek mode). Also SR uses the numerical dB(A) values as main basis of their conclusion... and they make their measurement ridiculously close to the HDD because rest of their benchmarking system is NOT quiet.
If you cannot decouple all the HDDs properly (with rubber straps) to 5.25" bay, then you ought to forget all Samsungs as they have a tendency to vibrate more than HDDs of other manufacturers. That is on average - sample variance do exist. Samsungs are also hotter than WDs and Hitachis. Seagate is also hot. WD could appear hotter than a Seagate, though, because some WDs have faulty temperature sensors. You'll have to compare actual temperature to your other HDDs, that report themselves to run cool when polling for SMART values, with some external sensor (like your hand) to determine if they run hot in real-life.
WD5000KS ---> WD5000AAKS might or might not be an improvement. I don't think it'd be big enough improvement to justify the cost... unless someone is ready to purchase your current HDD at a price close to that of new HDDs.
WD5000KS ---> T7K500... gray area. Only a handful of people here on SPCR discussion boards have actually tried T7K500. If I don't remember incorrectly, comments have been mostly positive. Low vibration, not too noisy seeks, etc. One thing is confirmed though: it's the coolest 500-gigger at low load. (While in theory, measuring power consumption alone isn't enough to determine operational temperature, in practice all same form-factor HDDs hava almost same thermal resistance characteristic: highly conductive metallica base cast, thin metal plate top cover, PCB on down side, base cast painted (improves heat loss by radiation), etc.)
WD5000KS ---> 7K1000... gray area. The main reason to consider this would be to reduce total number of HDDs or raise the maximum capacity that is possible to use in a system. It has quite a bit price premium... but so do 750-giggers. 500-giggers are very low at price/GB. Despite having more platters, 7K1000 uses less power than 500...750GB competition of other manufacturers. 7K1000: 7.9W, WD5000YS: 8.6W, Barracuda 750GB: 9.4W. Peak power consumption during start-up from +12V rail is also notoriosly low on all Kurofune series Hitachis (7K400, 7K500, 7K1000) at around 15...16W while Seagates are pretty much the opposite (up to 30W). (WD and Samsung are somewhere in between.)
...also I'd probably start the transition to new HDD from those Seagates, not those WDs. Especially if higher "temperatures" is caused by faulty sensors or if they are a result of packing them closer together than Seagates are packed. How high temperatures are we talking about, btw?
And what are those other HDDs? "currently i'm running with a mix of 400GB drives a couple of WD (500GBs) and a lot of Seagate 7200.8 / 7200.9 drives.."
Also those 7200.8 and 7200.9... are those 400GB as well? Or are those Seagates the only 400-giggers you have? If they are why did you mention them separately (i.e mix of 400GB drives and WDs and Seagates)?
I cannot comment on noise, because there's no (noisewise) reputable reviews of T7K500 or 7K1000. Anandtech said 7K1000 was the most silent HDD ever reviewed (AAM enabled) but I think that may be a little exagerated. StorageReview said it's quite noisy but only tested it with AAM disabled (fast seek mode). Also SR uses the numerical dB(A) values as main basis of their conclusion... and they make their measurement ridiculously close to the HDD because rest of their benchmarking system is NOT quiet.
If you cannot decouple all the HDDs properly (with rubber straps) to 5.25" bay, then you ought to forget all Samsungs as they have a tendency to vibrate more than HDDs of other manufacturers. That is on average - sample variance do exist. Samsungs are also hotter than WDs and Hitachis. Seagate is also hot. WD could appear hotter than a Seagate, though, because some WDs have faulty temperature sensors. You'll have to compare actual temperature to your other HDDs, that report themselves to run cool when polling for SMART values, with some external sensor (like your hand) to determine if they run hot in real-life.
WD5000KS ---> WD5000AAKS might or might not be an improvement. I don't think it'd be big enough improvement to justify the cost... unless someone is ready to purchase your current HDD at a price close to that of new HDDs.
WD5000KS ---> T7K500... gray area. Only a handful of people here on SPCR discussion boards have actually tried T7K500. If I don't remember incorrectly, comments have been mostly positive. Low vibration, not too noisy seeks, etc. One thing is confirmed though: it's the coolest 500-gigger at low load. (While in theory, measuring power consumption alone isn't enough to determine operational temperature, in practice all same form-factor HDDs hava almost same thermal resistance characteristic: highly conductive metallica base cast, thin metal plate top cover, PCB on down side, base cast painted (improves heat loss by radiation), etc.)
WD5000KS ---> 7K1000... gray area. The main reason to consider this would be to reduce total number of HDDs or raise the maximum capacity that is possible to use in a system. It has quite a bit price premium... but so do 750-giggers. 500-giggers are very low at price/GB. Despite having more platters, 7K1000 uses less power than 500...750GB competition of other manufacturers. 7K1000: 7.9W, WD5000YS: 8.6W, Barracuda 750GB: 9.4W. Peak power consumption during start-up from +12V rail is also notoriosly low on all Kurofune series Hitachis (7K400, 7K500, 7K1000) at around 15...16W while Seagates are pretty much the opposite (up to 30W). (WD and Samsung are somewhere in between.)
...also I'd probably start the transition to new HDD from those Seagates, not those WDs. Especially if higher "temperatures" is caused by faulty sensors or if they are a result of packing them closer together than Seagates are packed. How high temperatures are we talking about, btw?
And what are those other HDDs? "currently i'm running with a mix of 400GB drives a couple of WD (500GBs) and a lot of Seagate 7200.8 / 7200.9 drives.."
Also those 7200.8 and 7200.9... are those 400GB as well? Or are those Seagates the only 400-giggers you have? If they are why did you mention them separately (i.e mix of 400GB drives and WDs and Seagates)?
-
- Posts: 1386
- Joined: Fri Jun 18, 2004 6:53 pm
Thanks for the info. It's good to know Samsung is working on the next gen drives. S250 series is already on Samsung website, however that's an S series with only 8MB buffer, I would like to see new performance series T250 drives as well.alfred wrote:A shop in Belgium is claiming new 750 GB & 1 TB Samsung 3.5" drives are on their way (HD752LJ & HD102UJ). Probably 250 GB platters from a new T250 Series. samsung.com already shows single-platter HD250HJ and HD200HJ S250 Series.
-
- Posts: 606
- Joined: Wed Oct 13, 2004 4:03 pm
- Location: Albany, GA USA
-
- Posts: 220
- Joined: Wed Mar 15, 2006 10:16 am
- Location: U.S.A.
Holy Crap Batman!
Sorry for shouting but...
The new Samsung F1 1TB drive apparently uses 3 x 334GB platters!
I read the news in this thread over at Storage Review so thanks to those guys over there.
Some links from the aforementioned thread:
F1 pdf file.
PowerPoint presentation by a Samsung Exec confirming the news. It's in Chinese so any help with translation would be appreciated.
In conclusion,
Samsung is to release a 3 platter, 1 terabyte drive with 32mb cache for under 250 Euros. Do you think Seagate, Hitachi, and Western Digital are getting nervous?
edit: maybe the OP can change the thread title? This deserves its own thread!
Sorry for shouting but...
The new Samsung F1 1TB drive apparently uses 3 x 334GB platters!
I read the news in this thread over at Storage Review so thanks to those guys over there.
Some links from the aforementioned thread:
F1 pdf file.
PowerPoint presentation by a Samsung Exec confirming the news. It's in Chinese so any help with translation would be appreciated.
In conclusion,
Samsung is to release a 3 platter, 1 terabyte drive with 32mb cache for under 250 Euros. Do you think Seagate, Hitachi, and Western Digital are getting nervous?
edit: maybe the OP can change the thread title? This deserves its own thread!
I guess I should elaborate my previous comments: on idle T133 uses less power than WD5000xx (but the difference is only about 1 watt), when seeking (AAM enabled) they use the same amount of power. Thus I thought AAxx would likely use less than T133. I don't know about T166 if it's cooler than T133. And of course AAxx might not be cool running - it was merely an assumption of typical energy consumption/platter figures from past WD products which have been better in that category than Samsungs (most 3-platter variants of previous generation was cooler than T133). T1xx vs WD5000xx is in favour of Samsung due to less platters.
Even IF WD did use 1 watt more than T166, that small difference in power consumption cannot create bigger than 6 deg C difference in operational temperature. I still think it's a faulty temperature sensor playing tricks on you. It's not that uncommon on WDs. Some batches are all faulty. My WD3200JB's sensor acts more like a workloadmeter than thermometer: when I read or write anything from/to disk, temperature jumps up 10...20 deg C from idle values. And I have another first-hand example of sensor inaccuracy (and this time constantly inaccurate): my 5-platter 7200rpm 7K400 is "cooler" than 4-platter 5400rpm MaXLine II... if I believed the temperature sensor of that MaXLine. Since I have confirmed the accuracy of 7K400's sensor by using external temperature sensor, that leaves my thinking Maxtor's sensor is showing approximately 5 degrees over the real temperature.
Even IF WD did use 1 watt more than T166, that small difference in power consumption cannot create bigger than 6 deg C difference in operational temperature. I still think it's a faulty temperature sensor playing tricks on you. It's not that uncommon on WDs. Some batches are all faulty. My WD3200JB's sensor acts more like a workloadmeter than thermometer: when I read or write anything from/to disk, temperature jumps up 10...20 deg C from idle values. And I have another first-hand example of sensor inaccuracy (and this time constantly inaccurate): my 5-platter 7200rpm 7K400 is "cooler" than 4-platter 5400rpm MaXLine II... if I believed the temperature sensor of that MaXLine. Since I have confirmed the accuracy of 7K400's sensor by using external temperature sensor, that leaves my thinking Maxtor's sensor is showing approximately 5 degrees over the real temperature.
i would also concur with alfred
i have 2* 4 hot swap enclosures with the same cooling (120mm fan)
the seagate 400GBs hover around 38-40 ish degrees...
and the WDs in the same enclosure are 45ish
they seem to suffer from the enclosed space a lot more than the seagates...
just my observation of my setup.
i have 2* 4 hot swap enclosures with the same cooling (120mm fan)
the seagate 400GBs hover around 38-40 ish degrees...
and the WDs in the same enclosure are 45ish
they seem to suffer from the enclosed space a lot more than the seagates...
just my observation of my setup.
Sounds more like that WD's temperature sensor is positioned to some hot spot.whiic wrote:Even IF WD did use 1 watt more than T166, that small difference in power consumption cannot create bigger than 6 deg C difference in operational temperature. I still think it's a faulty temperature sensor playing tricks on you. It's not that uncommon on WDs. Some batches are all faulty. My WD3200JB's sensor acts more like a workloadmeter than thermometer: when I read or write anything from/to disk, temperature jumps up 10...20 deg C from idle values.
Sure; if temperature was closely related to power consumption, we would already know it.
I've got two WD5000AAKS and eight HD501LJ drives; when used in free air in similar conditions (i.e. when benchmarking, evaluating noise and testing temperatures) the Westerns feel much hotter to the hand.
whiic, I very much appreciate your extensive knowledge and detailed posts about hard disks, but sometimes I've got the feeling you don't like Samsung hard drives very much; I'm confident you would change your mind if you came to buy a T166 or S166.
I've got two WD5000AAKS and eight HD501LJ drives; when used in free air in similar conditions (i.e. when benchmarking, evaluating noise and testing temperatures) the Westerns feel much hotter to the hand.
whiic, I very much appreciate your extensive knowledge and detailed posts about hard disks, but sometimes I've got the feeling you don't like Samsung hard drives very much; I'm confident you would change your mind if you came to buy a T166 or S166.
"I'm confident you would change your mind if you came to buy a T166 or S166."
Maybe, but one might not always have the possibility to decouple it, if using for example external enclosures or Macintosh computer.
And I don't have the need for such "small" capacity HDDs. I might give a try with HD103UJ, though... when it comes available, which might take several months.
If people have actually confirmed by touching that the SMART temperatures aren't in conflict with reality, then I have to accept WD has lost it's second position as manufacturer of cool HDDs (even in terms of heat/platter). Thus Hitachi leads heat/platter and (for now) also heat at specified capacity points (500, 750 and 1000). When Samsungs HD103UJ comes available in a few months, I'm relatively certain it's take the position of coolest 1000-gigger around. I hope that time for availability doesn't take too long...
Maybe, but one might not always have the possibility to decouple it, if using for example external enclosures or Macintosh computer.
And I don't have the need for such "small" capacity HDDs. I might give a try with HD103UJ, though... when it comes available, which might take several months.
If people have actually confirmed by touching that the SMART temperatures aren't in conflict with reality, then I have to accept WD has lost it's second position as manufacturer of cool HDDs (even in terms of heat/platter). Thus Hitachi leads heat/platter and (for now) also heat at specified capacity points (500, 750 and 1000). When Samsungs HD103UJ comes available in a few months, I'm relatively certain it's take the position of coolest 1000-gigger around. I hope that time for availability doesn't take too long...