Got two on order; they should ship out today. A Samsung and a WD Scorpio.
I'll try to do a comparison for everyone, perhaps even run it against my 4200rpm, 8MB cached 80GB 2.5" Fujitsu.
-Ed
5400rpm, 8MB cached 80GB 2.5" drive comparison...
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
sgtpokey
- *Lifetime Patron*
- Posts: 301
- Joined: Sat Dec 14, 2002 11:29 pm
- Location: Dublin, CA / Liverpool UK
Great, if you have time, can you also compare performance to a relatively stock 3.5" 7200 8mb drive? It would be interesting to see current performance differences so people know what they lose (or do not lose) if they switch to a 2.5" 5400rpm drive on their desktop for the sake of silence.
I keep hearing that the smaller, denser 2.5" drives mean the drive heads have a form advantage in that they don't have to traverse so far. What that means for performance I've never seen from a good source (Come to think of it, I'll quickly check storagereview.com...)
I keep hearing that the smaller, denser 2.5" drives mean the drive heads have a form advantage in that they don't have to traverse so far. What that means for performance I've never seen from a good source (Come to think of it, I'll quickly check storagereview.com...)
-
sgtpokey
- *Lifetime Patron*
- Posts: 301
- Joined: Sat Dec 14, 2002 11:29 pm
- Location: Dublin, CA / Liverpool UK
Oops found some earlier info from a previous SPCR article:
http://www.silentpcreview.com/article207-page1.html
Maybe even just everyday simple single-user things like: Vanilla Windows boot times, game level loading times, or large file transfers... (No HD analysis tools required).
http://www.silentpcreview.com/article207-page1.html
The only problem is the only tests shown are SI Sandra tests. It's a decent benchmark but... well I know you folks don't do the whole gamut of HD tests, but it would help me out (and maybe others) if more test data could be included so we can get a better picture of real-world differences of a 5400 rpm 2.5" vs 3.5" 7200.PERFORMANCE
Generally, most quiet drive users report negligible — if any — real practical difference in perceived desktop system performance between a fast 7200 rpm drive and a slower, quieter 7200 rpm drive. We know from experience that average random access or seek time has the greatest perceivable impact on HDD related desktop performance. Here we are looking at 5400 rpm notebook drives whose stated latency is nearly 2ms slower than typical 7200 rpm drives, and whose rated average seek time is at least 3ms slower than the fastest 7200 rpm drives. You would think that the performance hit would be quite significant; we have not found this to be the case.
Maybe even just everyday simple single-user things like: Vanilla Windows boot times, game level loading times, or large file transfers... (No HD analysis tools required).
-
Edward Ng
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
The initial comparison will be directly between the two new drives--I only bench drives empty (no files). To compare against my MHT2080AT and my SP1614N (which is the best thing I've got, considering what you're asking for; 8MB cached and dual platter/quad head, but 3.5"/7200rpm and also low acoustics), I'd have to then clone each of them off to one of the new drives to empty them out, and then run the same battery of tests.
The problem is I don't have sufficient spare drives to clone off the SP1614N and the MHT2080AT to at the same time so that all four drives in question are empty and can have their tests run concurrently--this puts me at risk of inconsistency in testing. The thing is that my SP1614N is the OS drive for one of my machines, and the MHT2080AT is the OS drive for another one of my machines; I can't just dump the data on them, and I can't just back up only the critical stuff. They must be cloned in entirety.
I'll see what I can do...I have some 6Y060L0s laying around that, hopefully, are large enough for cloning the drives to. I believe the SP1614N is empty enough to be cloned down to a 6Y060L0, but the MHT2080AT is fairly full; I'll have to see how much data I can pull off of it before cloning, because if I can't clear it sufficiently, it's not fitting on a 6Y060L0 (yeah, it's really that tight
).
-Ed
The problem is I don't have sufficient spare drives to clone off the SP1614N and the MHT2080AT to at the same time so that all four drives in question are empty and can have their tests run concurrently--this puts me at risk of inconsistency in testing. The thing is that my SP1614N is the OS drive for one of my machines, and the MHT2080AT is the OS drive for another one of my machines; I can't just dump the data on them, and I can't just back up only the critical stuff. They must be cloned in entirety.
I'll see what I can do...I have some 6Y060L0s laying around that, hopefully, are large enough for cloning the drives to. I believe the SP1614N is empty enough to be cloned down to a 6Y060L0, but the MHT2080AT is fairly full; I'll have to see how much data I can pull off of it before cloning, because if I can't clear it sufficiently, it's not fitting on a 6Y060L0 (yeah, it's really that tight
).-Ed
-
Edward Ng
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
Okay, here are the results so far...
In terms of idle noise, overall volume is very nearly the same (by my guess, less than 2dBA difference between the two drives), but signatures are quite different.
The Samsung MP0804H exhibits a two part idle sound signature; there's a windy-sounding whir noise, which comprises about 90% of the sound output, and then the remaining 10% is a very faint, underlying metallic whine--this is only slightly audible beyond the SPCR standard one meter range; going out a good double to triple that distance. The wind whir portion is mid-high frequency, and carries farther than the whine, going a good several meters when in open air, sitting on some foam on the floor. Keep in mind that noone really utilizes a drive like this in open air, particularly considering how conducive they are to total enclosure dampening methods, being that they make so little heat.
The Western Digital WD800VE also exhibits a two part idle sound signature made up of mostly wind whir with a slight hint of metallic whine, but the whir makes up 95% of the signature, and the whine no more than 5% at most; the metallic whine of this drive isn't audible as far out as the whine from the Samsung, but only by a small margin--maybe a half a meter less. The wind whir is middle frequency, and noticibly lower in pitch than the whir from the Samsung. The whir from this drive carries just as far as the whir from the Samsung.
Personally, I prefer the Western Digital's idle sound signature, as I find higher frequency noises to be more distracting, plus I have a sharp bump in my own hearing sensitivity curve at the high frequency end, moreso than standard A-weighting.
There is no periodic calibration noise exhibited by either drive; seek tests have not yet been performed.
Looking at vibration, going by the, "hand test," both drives exhibit extremely minimal vibration--the closest thing I can come up with as a reference point is like when you're holding a cat and it makes those extremely subtle, almost sub-sonic purrs. The Western Digital is slightly smoother though, in this respect; I'd say its vibration is about 80% as strong as the Samsung's.
When placed on top of some very thin antistatic foam (3/16" thick, and dense--the type that motherboards come shipped on top of inside the antistatic bag) on a rug on the floor, both drives create an extremely subtle, bassy thrum; however, even the slightest attempt at suspending or dampening them completely eliminates this. A 2.5" No-Vibes kit would definitely be 100% effective at preventing the ultra-minor vibration from either of these drives from translating into even the most sensitive aluminum computer case.
Here too, I choose the Western Digital over the Samsung, but to be honest, the vibration is so little that either drive would be perfectly fine.
EDIT: Note that I mean vibration at idle; I have not yet tested for vibration during seeks, an issue the Samsung SP1614N is known for in particular, and is its primary weakness.
Looking at temperature in open air of 25C, the Western Digital beats the Samsung by a nosehair, with a measurement of 29C at the center of the spindle on the top cover for the Scorpio and 30C for the Samsung at the same exact location (measurements were taken multiple times throughout the evening). In both cases, I'd say temperatures are terrific, and I'd hesitate to let idle operating temperature come into play in making my decision when choosing between these two drives, but the WD is one degree cooler.
Again, to reinforce the point, these drives run so cool that there is hardly anything I can conceive of that would cause them to overheat. It should be noted, however, that both drives have tiny breather holes on top that must be left uncovered for the drives to operate properly.
Stay tuned for seek listening tests and performance tests. I'll try to get the seek listening tests done tonight, but I can't make any promises. Performance tests may not happen until some time in the middle of the week, along with temperatures under extended stress. I will also be doing some listening tests with the drives enclosed in SilentDrives.
Btw, I've discovered that Ghost 9.0, part of SystemWorks Premiere 2005, doesn't work the same way as the previous version I used before. I can't seem to find where I go to do a pre-GUI, full disk to disk clone like I did in the past. Has this function gone away? It's making it more difficult for me to do comparisons against my MHT2080AT and SP1614N to not have this functionality available to me...
-Ed
In terms of idle noise, overall volume is very nearly the same (by my guess, less than 2dBA difference between the two drives), but signatures are quite different.
The Samsung MP0804H exhibits a two part idle sound signature; there's a windy-sounding whir noise, which comprises about 90% of the sound output, and then the remaining 10% is a very faint, underlying metallic whine--this is only slightly audible beyond the SPCR standard one meter range; going out a good double to triple that distance. The wind whir portion is mid-high frequency, and carries farther than the whine, going a good several meters when in open air, sitting on some foam on the floor. Keep in mind that noone really utilizes a drive like this in open air, particularly considering how conducive they are to total enclosure dampening methods, being that they make so little heat.
The Western Digital WD800VE also exhibits a two part idle sound signature made up of mostly wind whir with a slight hint of metallic whine, but the whir makes up 95% of the signature, and the whine no more than 5% at most; the metallic whine of this drive isn't audible as far out as the whine from the Samsung, but only by a small margin--maybe a half a meter less. The wind whir is middle frequency, and noticibly lower in pitch than the whir from the Samsung. The whir from this drive carries just as far as the whir from the Samsung.
Personally, I prefer the Western Digital's idle sound signature, as I find higher frequency noises to be more distracting, plus I have a sharp bump in my own hearing sensitivity curve at the high frequency end, moreso than standard A-weighting.
There is no periodic calibration noise exhibited by either drive; seek tests have not yet been performed.
Looking at vibration, going by the, "hand test," both drives exhibit extremely minimal vibration--the closest thing I can come up with as a reference point is like when you're holding a cat and it makes those extremely subtle, almost sub-sonic purrs. The Western Digital is slightly smoother though, in this respect; I'd say its vibration is about 80% as strong as the Samsung's.
When placed on top of some very thin antistatic foam (3/16" thick, and dense--the type that motherboards come shipped on top of inside the antistatic bag) on a rug on the floor, both drives create an extremely subtle, bassy thrum; however, even the slightest attempt at suspending or dampening them completely eliminates this. A 2.5" No-Vibes kit would definitely be 100% effective at preventing the ultra-minor vibration from either of these drives from translating into even the most sensitive aluminum computer case.
Here too, I choose the Western Digital over the Samsung, but to be honest, the vibration is so little that either drive would be perfectly fine.
EDIT: Note that I mean vibration at idle; I have not yet tested for vibration during seeks, an issue the Samsung SP1614N is known for in particular, and is its primary weakness.
Looking at temperature in open air of 25C, the Western Digital beats the Samsung by a nosehair, with a measurement of 29C at the center of the spindle on the top cover for the Scorpio and 30C for the Samsung at the same exact location (measurements were taken multiple times throughout the evening). In both cases, I'd say temperatures are terrific, and I'd hesitate to let idle operating temperature come into play in making my decision when choosing between these two drives, but the WD is one degree cooler.
Again, to reinforce the point, these drives run so cool that there is hardly anything I can conceive of that would cause them to overheat. It should be noted, however, that both drives have tiny breather holes on top that must be left uncovered for the drives to operate properly.
Stay tuned for seek listening tests and performance tests. I'll try to get the seek listening tests done tonight, but I can't make any promises. Performance tests may not happen until some time in the middle of the week, along with temperatures under extended stress. I will also be doing some listening tests with the drives enclosed in SilentDrives.
Btw, I've discovered that Ghost 9.0, part of SystemWorks Premiere 2005, doesn't work the same way as the previous version I used before. I can't seem to find where I go to do a pre-GUI, full disk to disk clone like I did in the past. Has this function gone away? It's making it more difficult for me to do comparisons against my MHT2080AT and SP1614N to not have this functionality available to me...
-Ed
Last edited by Edward Ng on Sun Mar 06, 2005 10:17 am, edited 1 time in total.
-
madman2003
- Posts: 182
- Joined: Sat Jan 29, 2005 9:02 am
- Location: Planet earth
Having had to use Ghost 9, I can confirm the DOS imaging isn't supported anymore. In fact, Ghost 9 seems to be more of a copy of Drive Image (which they bought, so makes sense). Perhaps in the next version it'll be a "new" feature? The worst part is you can't, IME, make actual ghost extension files anymore, so using the old boot disks on new images is out of the question. I'd stick to Ghost 2003 unless you need the new features of 9Edward Ng wrote: Btw, I've discovered that Ghost 9.0, part of SystemWorks Premiere 2005, doesn't work the same way as the previous version I used before. I can't seem to find where I go to do a pre-GUI, full disk to disk clone like I did in the past. Has this function gone away? It's making it more difficult for me to do comparisons against my MHT2080AT and SP1614N to not have this functionality available to me...
-Ed
Re: 5400rpm, 8MB cached 80GB 2.5" drive comparison...
I'm very intrested in comparsion with 4200rpm, 8MB cached 80GB 2.5" Fujitsu, because i want to replace my notebook HDD with a new one, hoping that the new one will be quieter an cooler.Edward Ng wrote: A Samsung and a WD Scorpio.
I'll try to do a comparison for everyone, perhaps even run it against my 4200rpm, 8MB cached 80GB 2.5" Fujitsu.
[my thread: http://forums.silentpcreview.com/viewto ... highlight=].
Can you do that comparsion? I don't really care about preformance (it is web/mail notebook) - i just want cool & quiet hdd. Is 4200rpm significantly cooler than 5400rpm? 4200rpm vibrate less than 5400rpm?
Can you measure it, please - no fancy test "hand/finger" test will be ok.
If 4200 and 5400 are equall (more or less) in heat/vibration category - I'll buy 5400.
I don't mind little preformance gain after all
-
madman2003
- Posts: 182
- Joined: Sat Jan 29, 2005 9:02 am
- Location: Planet earth
-
Edward Ng
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
Not sure if you know about financial industry operation, but Monday was the quarterly coupon roll out (interest payments on CDS trades) and I've been up past my chin in work for the last three weeks settling payments and since Monday, reconciling payments that were missing, went the wrong way, to the wrong fund etc. etc. etc. It's left me barely any time to sleep, let alone eat or whatnot (i.e. review hard drives). I happen to be able to log in here and check on things every now and then randomly because sometimes it's sit and wait--send off e-mails to counterparties, then wait for a response, and it appears a lot of the counterparties are still letting the dust settle.madman2003 wrote:When will there be some kind of info on the seek test?
-Ed
-
Edward Ng
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
Unfortunately, I've not yet the opportunity to even connect anything to the IDE interface. Heck, they're still sitting there, collecting dust, hooked up to my fanless listening test power supply.sthayashi wrote:Do you have subjective opinions about the drive performances? I'd personally trust you if you told me that one of them 'felt' faster than the other.
-Ed
Xbit reviews the scorpio
-
madman2003
- Posts: 182
- Joined: Sat Jan 29, 2005 9:02 am
- Location: Planet earth
-
madman2003
- Posts: 182
- Joined: Sat Jan 29, 2005 9:02 am
- Location: Planet earth
-
Edward Ng
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
I just built an HTPC for a friend with a pair of these in SilentDrives (not RAIDed). Am currently running Windows XP MCE 2005 setup; there's no noticeable noise whatsoever from them, but as I said they're enveloped in SilentDrives. Given their remarkably low thermal output and higher thermal tolerance, I see absolutely no reason whatsoever not to completely isolate them.madman2003 wrote:You're using the drive now? Working well?
Windows setup file copying is taking a rather long time, but Windows XP MCE is spread across two CDs, so it cannot be compared to installing regular Windows XP Home/Pro. The machine is 3200+ Winchester with a gig of PQI Turbo PC3200 on a Gigabyte GA-K8N Ultra-9 (fanless from the factory) with a CNPS7700-Cu that I did a fan swap on (took out the stock Zalman unit and installed a de-hubbed Nexus 120). The case is a Lian-Li V1000 with the stock fans replaced with Nexus 120s, and the PSU is a Super Silencer 300 with the stock fan swapped for a Panaflo L1A. Graphics card is a Gigabyte GV-NX66T128VP (also fanless from the factory). Other than the aforementioned hard drives, there are a pair of Plextor SATA 16X dual layer DVD burners and a D-Link PCI 108G 802.11A/B/G Wi-Fi adapter. Keyboard and mouse are the new Logitech MX3100 (which I like so much I bought myself two sets of them!).
My friend will be attaching an Apple widescreen LCD to the DVI port and a widescreen HDTV to the component output. We're still deciding on a sound card. For the remote, we're skipping the stock MCE unit and awaiting the release of Logitech's Harmony 880. I've got several Zalman FanMate IIs here that I may or may not implement, depending on if I feel the need for them (all the fans are plugged into the mainboard and are speed regulated by it, so they're already partially undervolted).
-Ed
EDIT: Btw, I set up a fixed-size page file on the second drive, removed all page data from the first drive, made a Temp folder on the second drive and reassigned all Temp folder references to the Temp folder on the second drive.