WD VelociRaptor: A Triple Crown
Smart Drive Proof25
I removed my VR from the IcePack and installed it in the Smart Drive enclosure for 2.5" SAS drives and it is a great fit.
http://www.gup.co.jp/smartdrive/copper_oversea_01.php
You have to slide the drive into the enclosure so you can't retain the stock thermal pads, but I am not seeing any thermal issues so far.
http://www.gup.co.jp/smartdrive/copper_oversea_01.php
You have to slide the drive into the enclosure so you can't retain the stock thermal pads, but I am not seeing any thermal issues so far.
Re: Smart Drive Proof25
What stock thermal pads are you talking about?sopwith wrote:You have to slide the drive into the enclosure so you can't retain the stock thermal pads
How does the VR sound like when in the Smart Drive Proof25? Is the whine totally gone?
Its this enclosure, right?
Im asking because im not satisfied with the whining coming from my VR even though is enclosed in a 2,5"SQD enclosed in a 3,5"SQD and suspended/laying on foam.
I know the SPCR review says the VR is quiet. Mine isn't. Thats for sure.jaganath wrote:can this really be the quietest desktop drive if even with such massive overkill its noise cannot be damped?im not satisfied with the whining coming from my VR even though is enclosed in a 2,5"SQD enclosed in a 3,5"SQD and suspended/laying on foam.
A very high pitch whine is still audible.
My HD501LJ was much quieter. It was in the 3.5"SQD. AAM not even set.
Seeks were barely audible from 30cm away. Motor/spindle noise was inaudible.
-
- Friend of SPCR
- Posts: 356
- Joined: Wed Jan 16, 2008 8:56 pm
- Location: Council Bluffs, Iowa
- Contact:
Allan's exactly right. The SPCR review sample is the quietest desktop drive ever, but there's some significant sample variance and from what I've seen, the SPCR sample is the only Velociraptor with this trait.
I mentioned this once in another Velociraptor-related thread, but I'll mention it here as well: Perhaps I'm untrusting, but I think there's a very good chance the SPCR review sample was hand-picked by Western Digital, and I would love to see an addendum review using a retail drive. I understand that SPCR reviews carry the disclaimer that sample variance can't be anticipated, but the variance we're seeing here is far beyond what I'd normally expect. If the SPCR review is using a one-of-a-kind drive, carefully chosen for its properties, then it's not even close to representing reality and not particularly useful.
I mentioned this once in another Velociraptor-related thread, but I'll mention it here as well: Perhaps I'm untrusting, but I think there's a very good chance the SPCR review sample was hand-picked by Western Digital, and I would love to see an addendum review using a retail drive. I understand that SPCR reviews carry the disclaimer that sample variance can't be anticipated, but the variance we're seeing here is far beyond what I'd normally expect. If the SPCR review is using a one-of-a-kind drive, carefully chosen for its properties, then it's not even close to representing reality and not particularly useful.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
OK, I took another look/listen at the VR sample on hand. It was tested in the lab that's become the anechoic chamber... and that room was already the quietest. The converted kitchen that houses the PSU tester is much more "live" acoustically -- it's much less damped, there's a thin rug over vinyl tiles, and there are many melamine-covered surfaces (counter tops, cupboard doors, etc.). There's also a pretty big window which lets in more external noise even when it's closed.
In this room, the bare VR sounds considerably louder than I recall, and a quick & dirty recording makes it sound a bit noisier than the original recording. However, other drives such as the WD640 and the Samsung F1 series also sound noiser, but I'd put them all in about the same class subjectively. The VR still measures lower in SPL because it has less low freq output. It also sounds higher pitched, which is as it should be as its primary tone is at 167Hz instead of 120Hz, and all the harmonic overtones are at proportionately higher freqs. There's a bit more of the high frequency noise that's more prominent when it's it is in its heatsink/bracket. Again, I think this is due to the room being livelier.
I don't believe this was hand picked. They could not spend the time / effort to differentiate sonically between drives, and the screws looked untouched before I removed them. Also, many readers are much younger than me, and perhaps you/they have more high freq sensitivity. I have very good hearing but I would not be surprised if my high freq hearing is worse -- I've had many more decades to abuse it. So maye you/they hear the same thing I do but are bothered more by it.
I'll get Larry to give it a close listen and he can say what he hears with his young ears.
In this room, the bare VR sounds considerably louder than I recall, and a quick & dirty recording makes it sound a bit noisier than the original recording. However, other drives such as the WD640 and the Samsung F1 series also sound noiser, but I'd put them all in about the same class subjectively. The VR still measures lower in SPL because it has less low freq output. It also sounds higher pitched, which is as it should be as its primary tone is at 167Hz instead of 120Hz, and all the harmonic overtones are at proportionately higher freqs. There's a bit more of the high frequency noise that's more prominent when it's it is in its heatsink/bracket. Again, I think this is due to the room being livelier.
I don't believe this was hand picked. They could not spend the time / effort to differentiate sonically between drives, and the screws looked untouched before I removed them. Also, many readers are much younger than me, and perhaps you/they have more high freq sensitivity. I have very good hearing but I would not be surprised if my high freq hearing is worse -- I've had many more decades to abuse it. So maye you/they hear the same thing I do but are bothered more by it.
I'll get Larry to give it a close listen and he can say what he hears with his young ears.
VelociRaptor sheds the IcePack for enterprise.
Wonder if/when they'll be available to the great unwashed...
Wonder if/when they'll be available to the great unwashed...
-
- Posts: 2198
- Joined: Thu Feb 10, 2005 11:20 am
- Location: TN, USA
I think we have an answer for the noise difference between review samples and the retail velociraptor:
VelociRaptor - Beta vs. Final version
Performance has improved in the latest revision but noise has increased in step with the performance improvement.
VelociRaptor - Beta vs. Final version
Notice the increase in idle noise and Peak power consumption.WD's 10,000 RPM drive loses a bit of ground in the Office DriveMark 2006 but delivers improvements in all other single-user tests as well as an across-the-board increase when it comes to server-oriented performance.
Performance has improved in the latest revision but noise has increased in step with the performance improvement.
I thought I had the perfect plan - bare drive in 2.5" Scythe Quiet Drive and then put that in the P182 lower HDD bay. Well this POS Scythe enclosure uses non standard screws, so there is no way to connect it to the drive bay.
Anyone know if the No Vibes enclosure has this problem?
I think I'm just going to switch everything over to an Antec Solo and use the suspension mounting system.
Anyone know if the No Vibes enclosure has this problem?
I think I'm just going to switch everything over to an Antec Solo and use the suspension mounting system.
I asked this directly from WD themselves, and now I got a reply:MikeC wrote:regarding latency: I'm simply citing what's on WD's website. Their press briefing only highlights the improved seek times, doesn't mention latency. All the other specs appear to be substantially improved.lm wrote:...
How come the average latency can be so much worse? Isn't it supposed to be a linear function of the spindle speed?
Seek time tells you how long it takes for the drive head to move to the correct position, and then the latency tells how long it takes for the spindle to spin to the correct position afaik.
For every 10krpm drive it should therefore be 3ms average (half rotatation), 6ms max (full rotation)?
And true enough, that link takes you to the corrected specs.Western Digital Customer Service and Support wrote: That was a fault on the website about WD VelociRaptorâ„¢ specification. The true Average Latency is 3.00 ms (nominal).
Kindly follow the URL below for more information :
http://www.wdc.com/en/products/products ... 1050099587
MikeC: You might want to update your article to have the true latency number.
Nice drive.....but when you put the samsung 1tb f1 into the equation theres no contest.....in my eyes anyway. The velocirapter is the most quiet and most mb/s between the two hdds.....but only by a very small margin. Then things get rocky for the velocirapter. The velocirapter is twice the price of the f1. The velocirapter is less than a third of the capacity of the f1.
Regarding drive noise
Regarding drive noise, I can confirm that removing the IcePack reduces the high pitch noise significantly.
For me, the whine went from "definitely irritating" to "almost inaudible" when I removed the drive from the IcePack.
Now I just hope I won't need the warranty for the next couple of years.
For me, the whine went from "definitely irritating" to "almost inaudible" when I removed the drive from the IcePack.
Now I just hope I won't need the warranty for the next couple of years.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
It really depends what your priorities are. As an OS drive, the speed / silence of the velocirapter w/o icepack is unmatched except by the best SSDs, specifically the Intel $600 (best wholesale price in 1000 lots) 80GB X25m. In comparison to that, it looks cheap. The F1 doesn't even get close especially for random access speed. 300GB is plenty for a single drive in a high performance desktop PC. You don't need 1tb.drjunk wrote:Nice drive.....but when you put the samsung 1tb f1 into the equation theres no contest.....in my eyes anyway. The velocirapter is the most quiet and most mb/s between the two hdds.....but only by a very small margin. Then things get rocky for the velocirapter. The velocirapter is twice the price of the f1. The velocirapter is less than a third of the capacity of the f1.
My experience with the WD3000GLFS
Hi -- I just bought two WD3000GLFS's to use in RAID0, and figured I'd outline my experience and mini-review.
I was more than a bit worried about noise, and was apprehensive at first -- I only installed one of the two drives, thinking that if I needed to return them, at least I'll only pay the restocking fee on one of the two drives.
I plugged it in, booted the computer with the case's side panel off, and... well at first I thought the drive wasn't being powered. I couldn't hear it at all over my other components (I'll list them below, for whoever might care).
I booted into Vista (on my other drives, two 7500AAKS), then formatted the Velociraptor and copied a big file onto it. I can hear the seek noise but it's really not loud at all. It's on the same level as the WD7500AAKS, just a bit of a different tone ("sharper", if that makes any sense). I definitely wouldn't say the WD3000GLFS is louder by any means.
So I installed the 2nd Velociraptor, changed my BIOS settings to enable RAID, created a RAID0 array on the new drives (16k stripes), and installed Vista on the array. I then installed a bunch of Vista updates (including SP1, which took 14 minutes to install, including the reboot), and installed some drivers (motherboard stuff, nVidia drivers).
I guess I'm writing this in case someone is on the fence about getting the WD3000GLFS from a noise point of view, and is soaking up people's reviews and opinions (as I was) before making a decision. So for what it's worth, I'm extremely satisfied with mine, and find them to be very quiet drive. It's not like "the noise is okay, the performance makes up for it" -- they're quiet drives.
Regarding the RAID0 setup, I haven't done real-world tests yet aside from installing Vista and its patches + SP1, but my subjective impression of that was very positive. A synthetic benchmark (HD Tach) gave me results with 2x WD3000GLFS that are slightly above the results I was getting with 3x WD7500AAKS in RAID0. I don't have the numbers with me, but I believe I'm getting 7ms seek, 240MB/s average and 370MB/s peek. I can double-check when I get home if anyone is interested.
For the record, the case I'm using is the Silverstone Temjin TJ10, which has two drive cages that have rubber lining where the drive touches them, but the cages themselves aren't insulated against noise or vibration. There is a vent on both sides of the HD cages, but again the noise levels coming from the Velociraptors are very, very acceptable.
Finally, here's the list of the noise-generating components I'm using on that computer -- might help some people evaluate what I mean by "my drives are silent", which of course is subjective and relative to my environment and other components:
- PSU: Ultra X3 1000W
- Motherboard: Asus Maximus Formula (didn't install the optional chipset fan)
- Video: 8800GT
- Fans: 6x 120mm (1x on CPU, 1x on rear, 2x on top, 1x between HD cages, 1x in middle of case (feature of the Temjin TJ10 case))
- Drives: 3x WD7500AAKS, 2xWD3000GLFS
I'll answer any questions you guys might have in regards to noise and/or performance. Just remember that my Velociraptors are in RAID0 so I might not be able to give you discrete performance results for a single drive.
I was more than a bit worried about noise, and was apprehensive at first -- I only installed one of the two drives, thinking that if I needed to return them, at least I'll only pay the restocking fee on one of the two drives.
I plugged it in, booted the computer with the case's side panel off, and... well at first I thought the drive wasn't being powered. I couldn't hear it at all over my other components (I'll list them below, for whoever might care).
I booted into Vista (on my other drives, two 7500AAKS), then formatted the Velociraptor and copied a big file onto it. I can hear the seek noise but it's really not loud at all. It's on the same level as the WD7500AAKS, just a bit of a different tone ("sharper", if that makes any sense). I definitely wouldn't say the WD3000GLFS is louder by any means.
So I installed the 2nd Velociraptor, changed my BIOS settings to enable RAID, created a RAID0 array on the new drives (16k stripes), and installed Vista on the array. I then installed a bunch of Vista updates (including SP1, which took 14 minutes to install, including the reboot), and installed some drivers (motherboard stuff, nVidia drivers).
I guess I'm writing this in case someone is on the fence about getting the WD3000GLFS from a noise point of view, and is soaking up people's reviews and opinions (as I was) before making a decision. So for what it's worth, I'm extremely satisfied with mine, and find them to be very quiet drive. It's not like "the noise is okay, the performance makes up for it" -- they're quiet drives.
Regarding the RAID0 setup, I haven't done real-world tests yet aside from installing Vista and its patches + SP1, but my subjective impression of that was very positive. A synthetic benchmark (HD Tach) gave me results with 2x WD3000GLFS that are slightly above the results I was getting with 3x WD7500AAKS in RAID0. I don't have the numbers with me, but I believe I'm getting 7ms seek, 240MB/s average and 370MB/s peek. I can double-check when I get home if anyone is interested.
For the record, the case I'm using is the Silverstone Temjin TJ10, which has two drive cages that have rubber lining where the drive touches them, but the cages themselves aren't insulated against noise or vibration. There is a vent on both sides of the HD cages, but again the noise levels coming from the Velociraptors are very, very acceptable.
Finally, here's the list of the noise-generating components I'm using on that computer -- might help some people evaluate what I mean by "my drives are silent", which of course is subjective and relative to my environment and other components:
- PSU: Ultra X3 1000W
- Motherboard: Asus Maximus Formula (didn't install the optional chipset fan)
- Video: 8800GT
- Fans: 6x 120mm (1x on CPU, 1x on rear, 2x on top, 1x between HD cages, 1x in middle of case (feature of the Temjin TJ10 case))
- Drives: 3x WD7500AAKS, 2xWD3000GLFS
I'll answer any questions you guys might have in regards to noise and/or performance. Just remember that my Velociraptors are in RAID0 so I might not be able to give you discrete performance results for a single drive.
Quick update: the two Velociraptors in RAID0 (16k) give me 212MB/s average read and 449MB/s burst.
I just ran a HD Tach test on the Velociraptor array followed by a test of one of my standalone WD7500AAKS, and although the sound is different, I couldn't tell you one of them was louder or more annoying than the other. They're both drives that I can barely hear.
I just ran a HD Tach test on the Velociraptor array followed by a test of one of my standalone WD7500AAKS, and although the sound is different, I couldn't tell you one of them was louder or more annoying than the other. They're both drives that I can barely hear.
Fair enough I did build this PC this February with the intent of having a quiet PC, and it is quiet (as far as my ears are concerned). All fans, including the one in the power supply, are quiet. I tweaked the case / CPU fan speeds down to their "sweet spot". I can hear air being displaced, but I think I did pretty good with quieting the whole thing and picking quiet components.ck8-04 wrote:Sorry, but an Ultra X3 1kW PSU? 6 120mm fans? That's above the threshold for what I could consider quiet.
@MikeC: is there a chance that SPCR would review the newest version of the Velociraptor, namely HLFS, and update the existing article?
We SPCR people would like to have a definitive answer on the noise vs. performance question.
And since now I can afford to buy a VR for my next build, I'd love to see this question answered.
IMO tests could include placing the VR in a 2.5" SQD and then in a 3.5" SQD, to make it pretty much silent.
We SPCR people would like to have a definitive answer on the noise vs. performance question.
And since now I can afford to buy a VR for my next build, I'd love to see this question answered.
IMO tests could include placing the VR in a 2.5" SQD and then in a 3.5" SQD, to make it pretty much silent.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
The 3 drive samples I refer to are 3 "different versions" of the VR 300GB, including the HLFS. They all sound the same and perform the same, there's no question of that. The bare drives are extremely quiet in idle, without any whine, somewhat audible in seek. Just how audible will depend on many extraneous factors -- mounting details, case, room acoustics, your own hearing.Tzupy wrote:@MikeC: is there a chance that SPCR would review the newest version of the Velociraptor, namely HLFS, and update the existing article?
We SPCR people would like to have a definitive answer on the noise vs. performance question.
And since now I can afford to buy a VR for my next build, I'd love to see this question answered.
IMO tests could include placing the VR in a 2.5" SQD and then in a 3.5" SQD, to make it pretty much silent.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
The 3 variants:
1) the original - GLFS
2) a second version with different style IcePack (heatsink frame) - HLFS
3) bare drive - BLFS
The bare drive is available in 150/300gb capacities, and it is the BLFS model. http://www.wdc.com/en/products/products.asp?driveid=496
I know that the 150gb model is hard to find because they're being snapped up by corporate buyers, but you might have to hunt for the 300gb version too. Some US sellers are listed here.
1) the original - GLFS
2) a second version with different style IcePack (heatsink frame) - HLFS
3) bare drive - BLFS
The bare drive is available in 150/300gb capacities, and it is the BLFS model. http://www.wdc.com/en/products/products.asp?driveid=496
I know that the 150gb model is hard to find because they're being snapped up by corporate buyers, but you might have to hunt for the 300gb version too. Some US sellers are listed here.
These Velociraptors sound awesome!
I'm wanting to use two 150GB Velociraptors in RAID0 inside a CM Cosmos 1000 for my next build. The Cosmos 1000 uses silicone mounting points for the HDDs.
I don't want to use the IcePack due to the acoustic problems, so I'll be mounting the bare 2.5" HDDs.
Would it be feasible to use a couple of simple 2.5" HDD to 3.5" brackets to fit them into the HDD bays?
The silicone mounting points in the case should be enough to absorb any vibration and stop it from effecting the case.
What are your thoughts? Has anyone tried this?
Camnz
I'm wanting to use two 150GB Velociraptors in RAID0 inside a CM Cosmos 1000 for my next build. The Cosmos 1000 uses silicone mounting points for the HDDs.
I don't want to use the IcePack due to the acoustic problems, so I'll be mounting the bare 2.5" HDDs.
Would it be feasible to use a couple of simple 2.5" HDD to 3.5" brackets to fit them into the HDD bays?
The silicone mounting points in the case should be enough to absorb any vibration and stop it from effecting the case.
What are your thoughts? Has anyone tried this?
Camnz
Unless you really need the speed from RAID 0, I'd go with a single VelociRaptor and wait for reliable SSDs to hit the market. Oh wait, that's what I have donecamnz wrote:I'm wanting to use two 150GB Velociraptors in RAID0 inside a CM Cosmos 1000 for my next build.
I bought a 300GB VelociRaptor, carved a 10GB partition off the disk for my OS, and use the rest as /home - I have a pair of WD10EADS green 1TB drives as bulk storage. When reliable SSDs hit the market at an affordable price, I'll move my OS onto an SSD and keep the 300GB as a high speed data drive for VMs etc (my primary usage).
bgiddins wrote:Unless you really need the speed from RAID 0, I'd go with a single VelociRaptor and wait for reliable SSDs to hit the market. Oh wait, that's what I have donecamnz wrote:I'm wanting to use two 150GB Velociraptors in RAID0 inside a CM Cosmos 1000 for my next build.
I bought a 300GB VelociRaptor, carved a 10GB partition off the disk for my OS, and use the rest as /home - I have a pair of WD10EADS green 1TB drives as bulk storage. When reliable SSDs hit the market at an affordable price, I'll move my OS onto an SSD and keep the 300GB as a high speed data drive for VMs etc (my primary usage).
Yeah, I do want the speed of RAID0, and SSDs aren't really an option yet.
But the main point of my post was regarding the mounting of the bare VelociRaptor with a 2.5" HDD to 3.5" bracket. I'm looking at something simple like this inside my CM Cosmos 1000 - http://djb.co.nz/store/index.php?main_p ... s_id=20783
Has anyone tried this?
Thanks heaps,
Camnz
I asked a similar question in this post (similar looking adapter too):
viewtopic.php?t=48741&start=30
The consensus was that suspending the drives was recommended. I'm still thinking about the 2.5" adapter one day - my drive doesn't need the icepack, when running OS and a VM it never moves above ambient temps.
viewtopic.php?t=48741&start=30
The consensus was that suspending the drives was recommended. I'm still thinking about the 2.5" adapter one day - my drive doesn't need the icepack, when running OS and a VM it never moves above ambient temps.
-
- Posts: 2198
- Joined: Thu Feb 10, 2005 11:20 am
- Location: TN, USA
In most use RAID 0 and a single drive are not noticeably different in terms of speed.camnz wrote:Yeah, I do want the speed of RAID0, and SSDs aren't really an option yet.
http://www.codinghorror.com/blog/archives/000335.html goes into it from many angles or you could just go with the AnandTech summary:
You may already know this and if so ignore this post but I'm replying on the chance you don't and for the benefit of lurkers that read but don't post.If you haven't gotten the hint by now, we'll spell it out for you: there is no place and no need for a RAID-0 array on a desktop computer. The real world performance increases are negligible at best and the reduction in reliability, thanks to a halving of the mean time between failure, makes RAID-0 far from worth it on the desktop.
There are some exceptions, especially if you are running a particular application that itself benefits considerably from a striped array, and obviously, our comments do not apply to server-class IO of any sort. But for the vast majority of desktop users and gamers alike, save your money and stay away from RAID-0.
viewtopic.php?p=388987 has some more thoughts on raid levels if anyone is interested.
Just a note on AAM - I found that you can check/modify AAM in Linux with the following command:
sudo hdparm -M /dev/sda
...or whatever your device is.
My WD3000GLFS had a level of "254", which equates to "fast" - I set this to "quiet" with the following:
sudo hdparm -M 128 /dev/sda
I run a few VMs constantly, and there's always a slight chatter of seek noise coming from the PC - it was almost like flicking a switch, the PC is FAR quieter to my ears. I shut down and restarted 3 VMs concurrently, which always produces a cacophony of noise (I usually start them separately to avoid thrashing the HDD) - there's still obvious noise, but I would rate it about 50% of what it was.
First impressions are that this is a recommended tweak to the drive settings. Next one will be deciding what to do about the icepack - still thinking about suspension.
sudo hdparm -M /dev/sda
...or whatever your device is.
My WD3000GLFS had a level of "254", which equates to "fast" - I set this to "quiet" with the following:
sudo hdparm -M 128 /dev/sda
I run a few VMs constantly, and there's always a slight chatter of seek noise coming from the PC - it was almost like flicking a switch, the PC is FAR quieter to my ears. I shut down and restarted 3 VMs concurrently, which always produces a cacophony of noise (I usually start them separately to avoid thrashing the HDD) - there's still obvious noise, but I would rate it about 50% of what it was.
First impressions are that this is a recommended tweak to the drive settings. Next one will be deciding what to do about the icepack - still thinking about suspension.
SATA I compatibility of the Velociraptor
[not quite a necro resurrection]
My old root drive Samsung's seem to be having intermittent problems, and thus threatening a failure soon. However, my system (originally built by ARM, RIP) is now several years old and only supports SATA I.
Do the Velociraptors have jumpers to backwards support SATA I, or is only usable for SATA II? I can not find any information in reviews or teh WD data sheets.
My old root drive Samsung's seem to be having intermittent problems, and thus threatening a failure soon. However, my system (originally built by ARM, RIP) is now several years old and only supports SATA I.
Do the Velociraptors have jumpers to backwards support SATA I, or is only usable for SATA II? I can not find any information in reviews or teh WD data sheets.