or Connect
AVS › AVS Forum › Video Components › Home Theater Computers › New Seagate NAS HDD's
New Posts  All Forums:Forum Nav:

New Seagate NAS HDD's - Page 3

post #61 of 105
Strong basic logic ^
post #62 of 105
Quote:
Originally Posted by cybrsage View Post

Not to speak for him, but from my point of view if the 5400s were more (or less) reliable than the 7200s the websites dedicated to hard drive reviews would have such information listed. Hard drives are their passion and they would happily collect data about failures. Since no one has any data showing one speed drive is any more reliable than other speed drives we can conclude they are close enough to the same level of reliability to not matter.

Basically, it is like asking which piece of paper weighs more, the one with the drawing done in green crayon or the one with the drawing done in blue crayon. We can find the atomic structure of both crayons and determine which one is actually heavier - but no one is going to actually bother to do that since the difference is so small everyone simply says they both weigh the same.
Maybe I wasn't clear but my gripe was not so much with the claim that there is not a difference in reliability but rather it was with the claim it is statistically proven. So my gripe was not the message but rather the delivery. Replace the word "statistically" with "in my opinion" and two pages of posts would disappear.
post #63 of 105
Here's a question for all: (opinion / speculation wanted)

If you could magically go back and figure out the total amount of failures and the total failure rate of every hard drive made in the last 3 years and complile the data into three categories (7200rpm, under 6000rpm, and all others) which category would you guess had the lowest failure rate ?
post #64 of 105
Thread Starter 
Quote:
Originally Posted by Mfusick View Post

Here's a question for all: (opinion / speculation wanted)

If you could magically go back and figure out the total amount of failures and the total failure rate of every hard drive made in the last 3 years and complile the data into three categories (7200rpm, under 6000rpm, and all others) which category would you guess had the lowest failure rate ?

My guess would be that most drives succumb to mechanical failure so IMO under 6000rpm drives would have the lowest mechanical failure rate. I am assuming that the mechanical components are of the same level of quality between the drives.

David
post #65 of 105
I don't think most drives fail because of mechanical failure of the motor or bearings. I believe I read that head crashes and electronics failures account for majority of failures.

I'm nearly certain of this but I'm scared to state anything other than this is my opinion smile.gif.
post #66 of 105
Thread Starter 
Quote:
Originally Posted by Mfusick View Post

I don't think most drives fail because of mechanical failure of the motor or bearings. I believe I read that head crashes and electronics failures account for majority of failures.

I'm nearly certain of this but I'm scared to state anything other than this is my opinion smile.gif.

I was including head crashes as a mechanical failure.

David
post #67 of 105
Quote:
Originally Posted by amarshonarbangla View Post

Wouldn't know since I am yet to see anyone post a pic of themselves on this forum smile.gif

My pic is of me sounding my Shofar while wearing my Tallit on Shabbat afternoon.
post #68 of 105
Quote:
Originally Posted by wsume99 View Post

Maybe I wasn't clear but my gripe was not so much with the claim that there is not a difference in reliability but rather it was with the claim it is statistically proven. So my gripe was not the message but rather the delivery. Replace the word "statistically" with "in my opinion" and two pages of posts would disappear.

Ah, I had to look back to page one to find what you were talking about. Yeah, neither one is statistically proven to be more reliable than the other...at least not that I have found.
post #69 of 105
Head crashes are typically a failure mode that occurs much sooner than a normal EOL failure such as mechanical wear. I still don't see how head crashes would be less likely with slower spindle speeds. We are still talking near or above 100 rotations per second with minimal clearance. I think a head crash is commonly attributed to the drive experiencing a shock above the design limits of the casing

http://en.wikipedia.org/wiki/Hard_disk_drive_failure

http://en.wikipedia.org/wiki/Head_crash

I've got a couple ST3000DM001 drives that just past their 1yr in use 24/7 day this month cool.gifhttp://lime-technology.com/forum/index.php?PHPSESSID=156ef0a12910bfbbb721f15a804446f4&topic=26568.0
post #70 of 105
There was a couple good links in the limetech post you linked.

I found it interesting some of the Tomshardware links that showed these results:



I've been saying this for about a year and getting flamed for it^ The 7200rpm drive is nearly equal or better in power and heat and noise to a 5400rpm WD drive yet sells for less money and just slays it in performance. The 4TB seagate drops the spindle speed to 5900rpm, maintains a good performance advantage and also opens up a notable advantage on power consumption using half a watt less and adding 1TB of capacity without much extra cost.










Quote:
Performance? Seagate Barracuda
The fastest 3.5” high-capacity desktop hard drive in this round-up is Seagate’s new 3 TB Barracuda (ST3000DM001). It offers almost 200 MB/s sequential read performance, similar write speeds, and it runs relatively coolly thanks to its three-platter design.

I found this interesting:




There is just no point in 5400rpm drives for desktops, PC's or media servers. It's not a good idea IMO. Perhaps if I had a crappy NAS box that was limited in performance capability anyways and the 5400rpm were on sale it might make sense if your also needing TLER for RAID. But then again the Seagates (non NAS) have always been recommended for RAID set ups per the MFG. I am just noticing Seagate is removing this from the non NAS drives now- but has for the past 12 months said they are ok in a RAID set up. I'm guessing it to support the higher price, and increase profits with the NAS line. Bottom line is if your not running a NAS and your not running RAID then your probably going to find a better value in a non NAS specific HDD.
post #71 of 105
I wasn't able to follow this thread today but were the statistics referenced ever posted? I am not trying to be an ass --- I am genuinely interested in seeing the data.
post #72 of 105
@ Mfusick - looking at all the data you posted I don't come the conclusion that 5400 is crap and 7200 rpm is the way to go. Instead I would conclude that selecting a HDD with the minimum number of platters for a given capacity is the goal. Fewer platters means higher density so you'll see better throughput and it also means less rotating mass so power consumption and heat should be lower than a drive with similar spindle speed but more platters. Seagate seems to have pushed 1TB platters to market sooner so right now their products are reviewing much better than drives with identical capacity but more platters.

Also, because all of my content is stored on a media server, I personally don't put much value in drive performance above 100 MB/sec. As long as the drive can sustain those speeds it will more or less saturate a gigabit connection. So based solely on the data you posted above I would select the Hitachi 5k4000 drive for my media server over the Seagate Barracuda (ST3000DM001). Why, because it has sufficient performance to meet my use case while providing lower temps and lower power consumption. Obviously I'd have to consider price and warranty as well before I could determine if the total value of the Hitachi outweighed the Seagate. So my point is that my decision would be based on empirical data not just the fact that one drive was 5400 rpm and the other was 7200 rpm.
post #73 of 105
Quote:
Originally Posted by wsume99 View Post

@ Mfusick - looking at all the data you posted I don't come the conclusion that 5400 is crap and 7200 rpm is the way to go. Instead I would conclude that selecting a HDD with the minimum number of platters for a given capacity is the goal. Fewer platters means higher density so you'll see better throughput and it also means less rotating mass so power consumption and heat should be lower than a drive with similar spindle speed but more platters. Seagate seems to have pushed 1TB platters to market sooner so right now their products are reviewing much better than drives with identical capacity but more platters.

Also, because all of my content is stored on a media server, I personally don't put much value in drive performance above 100 MB/sec. As long as the drive can sustain those speeds it will more or less saturate a gigabit connection. So based solely on the data you posted above I would select the Hitachi 5k4000 drive for my media server over the Seagate Barracuda (ST3000DM001). Why, because it has sufficient performance to meet my use case while providing lower temps and lower power consumption. Obviously I'd have to consider price and warranty as well before I could determine if the total value of the Hitachi outweighed the Seagate. So my point is that my decision would be based on empirical data not just the fact that one drive was 5400 rpm and the other was 7200 rpm.

Excellent take on the subject wsume99.

Another point is that while 5 degrees doesn't seem like much for a single drive it really adds up when you have 10 or 20 drives which is typical in some HTPC servers. The same can be said with a small increase in noise or that fans that are required to cool that 5 degrees per drive.

Some in this thread will cite that they don't want to "wait" on very large data transfers so that getting the fastest drive is of paramount importance. While I appreciate this viewpoint many others will only occasionally be moving very large data in or to their servers. For many more they will just plan on doing this transfer at night or when the server isn't being used. For some (such as myself) I don't really care as I have never had an issue with moving data and playing a movie, for instance.

In any event for a HTPC server that mainly serves movies and media to the Home (the "H" in HTPC) the bottom line is that ANY of these drives are plenty fast for playback many times over with multiple simultaneous streams as you alluded to in your excellent post. We aren't talking business servers in need of the ultimate level of performance here.
post #74 of 105
I understand the points being made about LAN limited speed- but keep in mind two things.

First, Drives slow down and the higher performance drive would likely remain above your LAN speed preventing this from becoming a bottle neck when the drive was old and full. Not the case with a 5400rpm that would drop below the gigabit LAN speed if full, and reading or writing data to the slower area of the platter. You would then have a bottle neck, often under 100MB/sec. With some 5400rpm drives it could be as low as 50MB/sec. If you don't want to take my word for it there is statsitics, independent data, and professional 3rd party PC hardware tech review sites that publish the test results showing I am right. Just because a 5400rpm drive can do faster than a LAN speed- does not mean it always will. It's likely at times it will not. I've personally experienced this myself.

Second,
This is an extension of above really for my second point^. Not only do hard drives slow down, but other performance limiting factors happen. For instance if your running a software raid set up in your media server or doing "pooling" or "parity" there is often an extra process that must happen that might chip away a little at your performance. There is a good thread here: http://www.avsforum.com/t/1476704/slow-write-speeds-on-flexraid-raid-f-snapshot-raid/0_100#post_23442384
I see many people asking why do they get slower write or read speeds from their servers all the time around here and the common denominator is always 5400rpm drives. I am not advocating spending tons of money or always having to go the high performance route - but when you can get better performance for a lower price it makes perfect sense. It's been that way for a while now yet there is so many proponents of 5400rpm drives still even though the 7200rpm models are available at superior performance, and lower prices. My personal belief is they are more reliable too but that is not a fact I can prove so I will leave it alone.

For my personal server my hardware it pretty modest. I use a $60 CPU, an $80 mobo , $80 Sata card, $25 DDR3 ram, $60 PSU, and $99 hard drives. I've paid as little as $80 for a 3TB 7200.14 Seagate and never more than $109. I picked the best components for my budget (IMO) and I've always valued performance and value (price) which is why I have chosen the parts I did. My goal was to get the best possible performance at my budget. My server is cheaper than most I'd guess. I am proud it's so affordable and capable.

I see others making mistakes or spending extra $ on components that are lower performance and I scratch my head without understanding. If I question it I usually first get some BS about how someone thinks it's more reliable, then I get flamed. It's the same story every time. This thread being no different. I understand my posting style is a cause for much of this and accept it. Call it a flaw I have. But never once does anyone stand up and show me where I am wrong - or point out a serious real world benefit to buying a more expensive lower performance 5400rpm HDD.

Here's what I know:

If you do the math on energy savings there is almost zero real world benefit with a 5400rpm hard drive. The real cost savings are pennies and meaningless considering you would never recoup the extra you paid for the more expensive 5400rpm drive even if the drive lasted in excess of 5 years. The idea of energy savings (and heat and noise to some effect) is mostly marketing propaganda. The temps, the noise, and the energy consumed is pretty close between a modern 7200rpm and a 5400rpm drive. A much more important factor is the price, the reliability, and the performance. In these three areas the 7200 usually wins or is equal.

So - while I am clearly wrong when I get overly agressive with my words trying to make my point clear by overstating my opinion as fact in some cases, I don't think my opinion is wrong about 5400rpm drives. That's why 95% of the defense of 5400rpm drives is often personal attacks at my credibility and posting style, and very little data or statistics or scientific facts to support 5400rpm drives as a good choice. It flows both ways. I have yet to see anyone, anywhere- post anything meaningful about how a 5400RPM is a better choice. Perhaps on the TLER issue, but even then you can use a 7200.14 in a NAS box or a desktop raid set up. I'd upgrade if I was doing serious server hardware RAID but that's beyond the scope of this forum I think.
Quote:
Originally Posted by assassin View Post

Excellent take on the subject wsume99.


Good points ? Yes. But the idea that he proposes about 5400rpm drives exceeding LAN in the real world is a bit optimistic. In my personal experience a WD RED or GREEN will slow down after time and once full to speeds under gigabit LAN.
I'd be very curious to see your Flexraid server with a WD GREEN drive that was 80% full copy paste to an SSD over your LAN. I'm guessing your under my 100MB/sec+ My experience was near 60MB/sec which is far more common and likely. Removing this bottle neck without spending any more $ seems to make sense if you were buying today. I know you and I have built servers around the same time and back when we did it- that was how you did it and what you had to work with. I've since upgraded mine, replaced my 5400rpm drives and I can say that at least for me personally- There is a really big difference.

Just my .02cents.



Quote:
Originally Posted by assassin View Post


Another point is that while 5 degrees doesn't seem like much for a single drive it really adds up when you have 10 or 20 drives which is typical in some HTPC servers. The same can be said with a small increase in noise or that fans that are required to cool that 5 degrees per drive.


But there is not a 5 degree difference. The results are nearly the same. A 7200.14 has the same surface temps as a 3TB GREEN drive, the same noise (minus the clicking of the green wink.gif ) It's not a big deal at all. The difference if any is not significant. Each generation of HDD's comes out is a little bit cooler than the last- and a little bit more energy efficient. The 4TB seagate adds a platter and an extra 1TB of capacity with nearly the same energy and heat, and in some cases even improves in these areas. A RED drive is more modern than a GREEN drive and is also just a tad bit cooler I believe. But - at the end of the day none are significantly different enough to make it a concern. As new models are introduced the older models just get exposed as less competitive and this will never change- but we have reach a point where most modern drives today are acceptable for this to be a non issue in all but the most extreme circumstances.

I own the same case as you. (ANTEC 300) and I am certain even a budget $39 case like that is more than up to the task of keeping quiet and keeping cool a multitude of HDD's. Most server cases are designed to do this and if they are not then your choosing the wrong case for a server IMO. My Norco case certainly can keep 20 high performance hard drives at full bore cool- and with my optional 120mm silent fan blade it's as quiet as my Antec300 (which is quiet with fans on low or medium).

There is no run away temps or noise like your suggesting with any properly designed build if you choose 7200rpm drives. It's not an issue.
Quote:
Some in this thread will cite that they don't want to "wait" on very large data transfers so that getting the fastest drive is of paramount importance. While I appreciate this viewpoint many others will only occasionally be moving very large data in or to their servers. For many more they will just plan on doing this transfer at night or when the server isn't being used. For some (such as myself) I don't really care as I have never had an issue with moving data and playing a movie, for instance.

In any event for a HTPC server that mainly serves movies and media to the Home (the "H" in HTPC) the bottom line is that ANY of these drives are plenty fast for playback many times over with multiple simultaneous streams as you alluded to in your excellent post. We aren't talking business servers in need of the ultimate level of performance here.


I can give you this point. I'm willing to accept you don't mind waiting twice a long for a transfer or that the transfer speed is less important to you. But keep in mind I am not advocating high performance. Rather I am advocating for maximum value. Maximum value to me is getting the most possible performance and capability for a given cost or a given budget. I'm advocating getting more and paying less. Perhaps if the extra performance cost more- your argument about not needing it might make more sense. But when the higher performance costs less it's really hard to argue against that. So in this regard- spending more on a 5400rpm drive make less sense.
post #75 of 105
Actually, 106F is 11 degrees warmer than 95F, both being Hitachi Deskstar 4TB drives.
post #76 of 105
(I think we all know who this question is directed to so no need to explicity state it.)

What is a better value - ST3000DM001 or ST4000DM000?

I don't want to hear about what the prices were at one time or may be in the future. I did a very quick search and I see the 3TB going for $130 and the 4TB going for $170 - both running about $43/TB. So based on today's price, which is the better value (in your opinion)?

EDIT: Apparently my googlefu is lacking and I missed the fact that the ST3000DM001 is an egg shocker for $110 today. So that changes the calculus a bit. So the question is what is the better value the 3TB @ $37/TB or the 4TB @ $43/TB?
Edited by wsume99 - 6/19/13 at 9:45am
post #77 of 105
Quote:
Originally Posted by cybrsage View Post

Actually, 106F is 11 degrees warmer than 95F, both being Hitachi Deskstar 4TB drives.

True. But your looking at Hitachi drives comparing the coolest to the hottest. In both cases these Hitachi drives sell for a lot more than a WD 5400rpm GREEN/RED or a Seagate Barracuda so I just ignore them based on price.

I was looking at the 7200rpm Seagate at 99 degrees versus the WD GREEN 3TB at 99 degrees in much of my thinking. Thanks for pointing this out to me, as it reminded me to point out what I was looking at.
post #78 of 105
Quote:
Originally Posted by wsume99 View Post

(I think we all know who this question is directed to so no need to explicity state it.)

What is a better value - ST3000DM001 or ST4000DM000?

I don't want to hear about what the prices were at one time or may be in the future. I did a very quick search and I see the 3TB going for $130 and the 4TB going for $170 - both running about $43/TB. So based on today's price, which is the better value (in your opinion)?

I think they are equal. Two days ago the 3TB was $109 at new egg. Coupon code posted in the other thread. Last week the 4TB was $149. Per TB I think the Seagates are pretty close usually- both about the same per TB. All drives go on sale so that is up to the buyer to find.

Based on the same cost per TB I'd say both are equally great value, and clearly a better value compared to Hitachi or WD.

Personally- I would get a 3TB 7200.14 because I value the speed and it costs nothing extra. But I have a 20 bay server case with hot swaps and adequate cooling, and I also have a 3TB parity drive. I am not personally limited in drive bays, but my 3TB parity drive would need to be upgraded to 4TB to allow for 4TB data drives in my server. So- Personally I have a few reasons why the 3TB is a better choice for me that doesn't necessarily mean that the 4TB would not be a better choice for someone else. The 4TB seagate is 4 platters, uses 5 watts, and only 5900rpm. The 3TB seagate is 7200rpm, uses 3 platters, and uses 5.2 watts. There is not much difference between them. If your looking for a happy medium between 5400rpm and 7200rpm I'd tend to think the 5900rpm 4TB Seagate is exactly that. It's a better performer than the 5400rpm WD or Hitachi, and it's cheaper, cooler and more energy efficient. It's not as fast at the 7200rpm Seagate 3TB- but it's more than fast enough to make that not a huge deal to many.

I think the 7200rpm is better for a high performance /high value server and it's better as a desktop drive. I think the 4TB 5900rpm Seagate is better for a low power media server where performance is not demanded and cost savings is.

In the end I think it will simply come down to how much storage people want per sata port and per drive. If your looking to maximize capacity and minimize the number of drives in your system clearly the biggest difference and advantage is the 4TB is 4TB and the 3TB is only 3TB. I would think choosing the 4TB for this reason is a wise choice.

Since I have 20 bays in my Norco and the cost per TB is the same- Personally I'd get the 3TB 7200.14 and enjoy 200MB/sec + speeds; The 3TB Seagate performance wise is operating in a territory other hard drives can't go.
post #79 of 105
Quote:
Originally Posted by Mfusick View Post

True. But your looking at Hitachi drives comparing the coolest to the hottest. In both cases these Hitachi drives sell for a lot more than a WD 5400rpm GREEN/RED or a Seagate Barracuda so I just ignore them based on price.

I was looking at the 7200rpm Seagate at 99 degrees versus the WD GREEN 3TB at 99 degrees in much of my thinking. Thanks for pointing this out to me, as it reminded me to point out what I was looking at.

My takeaway was to avoid Hitachi...but then I am still leery of the exploding DeathStar... smile.gif
post #80 of 105
Quote:
Originally Posted by wsume99 View Post

(I think we all know who this question is directed to so no need to explicity state it.)

What is a better value - ST3000DM001 or ST4000DM000?

EDIT: Apparently my googlefu is lacking and I missed the fact that the ST3000DM001 is an egg shocker for $110 today. So that changes the calculus a bit. So the question is what is the better value the 3TB @ $37/TB or the 4TB @ $43/TB?

Right smile.gif

I caught your edit too late.

But it's not like the 4TB does not go on sale either. I was assuming general pricing in my thoughts. If one was on sale and the other was not- well then.. That makes the choice a little easier. Money always matters. But if you really wanted one more than the other I don't think $5 per TB is a deal breaker. Extra $15 might be worth it in this case because someone might just rather want or need 4TB. If you only had three HDD ports or HDD bays... you can get 12TB with 3 of the 4TB drives and only 9TB with 3 of the 3TB drives.

12TB > 9TB biggrin.gif

But if you had 20 bays like me available and only use 12 of them so far, I'd take the faster drive with a lower cost per TB.
post #81 of 105
I have been buying the Seagate 4TB drives simply for the extra storage space they provide in the same physical space. I only have 10 SATA ports in my PC (8 on the mobo and two on an PCIe card) so I have to worry about density. I lose three SATA ports to the OS SSD, the Recorded TV HDD, and the BR / HD-DVD combo drive, so that leaves only seven free ports. One of those is used up by the parity drive, leaving only six ports for storage. Filling all six with 3TB drives gives me 18TB of storage while using 4TB drives gives me 24TB of storage. Right now I have three 4TB drives and two 3TB drives leaving one free port. Once that is filled I will have to replace drives which means not much extra storage space added.

So like msfusick said, you have to take your personal storage considerations into account. From my view, the cost of replacing a HDD with a higher capacity one means the cost per TB is enormous, so I should buy the largest available now instead of doing that.
post #82 of 105
Quote:
Originally Posted by cybrsage View Post



My takeaway was to avoid Hitachi...but then I am still leery of the exploding DeathStar... smile.gif

I can't help myself. I have to do it. An oldie but a goodie:




*Note: No hate on Hitachi. I've seen statistics showing Hitachi to be the most reliable drives in enterprise, and I also I own some and love them. One of my favorite drives is in my desktop and is a 2TB Hitachi. I believe Hitachi makes some very good hard drives these days. But the deathstar will be famous forever in the laurels of HDD history. eek.gif
post #83 of 105
So based on your last few posts would you be willing to admit that your previous comments were not entirely correct?
Quote:
Originally Posted by Mfusick View Post

There is just no point in 5400rpm drives for desktops, PC's or media servers.

Wouldn't it be true to say that there are valid reasons to use 5400 rpm drives? Just like there are valid reasons to use 7200 rpm drives. For that matter I pay zero attention to spindle speed when making a selection. I look at how the drive performs not what the spindle speed is. Again, my gripe is with your insistence that 5400 rpm drives should basically be de facto eliminated from consideration when making a HDD purchase.
post #84 of 105
Quote:
Originally Posted by cybrsage View Post

I have been buying the Seagate 4TB drives simply for the extra storage space they provide in the same physical space. I only have 10 SATA ports in my PC (8 on the mobo and two on an PCIe card) so I have to worry about density. I lose three SATA ports to the OS SSD, the Recorded TV HDD, and the BR / HD-DVD combo drive, so that leaves only seven free ports. One of those is used up by the parity drive, leaving only six ports for storage. Filling all six with 3TB drives gives me 18TB of storage while using 4TB drives gives me 24TB of storage. Right now I have three 4TB drives and two 3TB drives leaving one free port. Once that is filled I will have to replace drives which means not much extra storage space added.

So like msfusick said, you have to take your personal storage considerations into account. From my view, the cost of replacing a HDD with a higher capacity one means the cost per TB is enormous, so I should buy the largest available now instead of doing that.

Perfect real world example^

I'd say filling that next port with a 4TB would make perfect sense in your situation.

It is is worth noting that a 5900rpm Seagate 4TB performance is a good deal better than a 5400rpm WD RED or GREEN drive so while the performance might not be as good as the fastest consumer hard drive available today, it's quite a bit better than the average and certainly more than acceptable in performance with read and write speeds north of gigabit LAN capabilities. I don't see much issue with the 4TB Seagate being 5900rpm because it maintains nice performance advantage over the 5200rpm drives. The 4TB Seagate seems to balance everything you want in a HDD very well- making nearly a universal best buy for most. I just personally like the 7200.14 3TB models because they kick ass and take names.
post #85 of 105
Quote:
Originally Posted by wsume99 View Post

So based on your last few posts would you be willing to admit that your previous comments were not entirely correct?
Quote:
Originally Posted by Mfusick View Post

There is just no point in 5400rpm drives for desktops, PC's or media servers.

Wouldn't it be true to say that there are valid reasons to use 5400 rpm drives? Just like there are valid reasons to use 7200 rpm drives. For that matter I pay zero attention to spindle speed when making a selection. I look at how the drive performs not what the spindle speed is. Again, my gripe is with your insistence that 5400 rpm drives should basically be de facto eliminated from consideration when making a HDD purchase.

No. I would not admit it tongue.gif

There is a difference between a 5900rpm and a 5200rpm drive. Specifically it's 700rpm, and a measurable drop in performance.

Just like there is a difference between 7200rpm and 5900rpm. Specifically it's 1300rpm and a measurable drop in performance.

Where the issue lies for me is that 5900rpm seems to balance or at least maintain a bit more performance than the 5200rpm drives do, making it a much better choice in between the two and a better balance IMO. I believe that 5200rpm drives give up too much in actual performance and provide not enough real world benefits to account for it.

Taking a look at the current WD 5200rpm line up over last 12 months for example you notice a bunch of things that stand out and support my opinion. First- There's a serious lack of 4TB availability. That alone puts the 4TB seagate on an island all unto it's own. Most 5400rpm are 2TB and 3TB making them a better comparison to the 7200.14's at the same size. -Difference being the 7200.14's handily romp the 5200rpms on performance and typically sell for less cost. They are nearly the same in energy, heat and noise too. Paying extra for less is a concept my brain won't easily accept.

I don't count a 5900rpm 4TB seagate with 150MB/sec performance and low cost per TB, Low energy, and low Heat/noise in my consideration and opinion about 5200rpm drives in general. I don't consider them the same and they don't behave the same. The 5900rpm Seagate is superior in most cases and accounts- making it clearly a better choice. I take much less offense and issue given this reality and I don't see any good reason not to like or consider a 4TB 5900rpm Seagate.

But make no mistake- If I had a choice between a 7200rpm 4TB and a 5900rpm 4TB at the same cost- and the same warranty I'd take the 7200rpm every time. Since 7200rpm 4TB are don't exist yet it's not a problem though. If you want a 4TB right now you really only have 1 choice that's highly affordable. Other 4TB drives are much more expensive that disqualifies them based on higher prices.
post #86 of 105
Quote:
Originally Posted by assassin View Post

I wasn't able to follow this thread today but were the statistics referenced ever posted? I am not trying to be an ass --- I am genuinely interested in seeing the data.

Not sure, but I hypothesized in the following post why there woudn't likely ever be any such statistics.
Quote:
Originally Posted by 4th-horseman View Post

Unless any manufacturer decided to take large portions (enough to be statistically relevant) of their product and run them till they die there would never be any data on failure modes

How many times do you call WD, Seagate, HGST (Hitachi, IBM, tongue.gif) when you experience a drive failure to tell them what kind of failure it was? You'd have to send it to them for analysis for them to begin gathering data on why yours and all their other consumers failures occured. I can't see that happening

Also in a different place mentioned how I can't begin to understand the correlation between reliability and spindle speed. Before even starting to mine data to support the claim, if someone could indicate a basis for this claim that makes any sort of sense I'd be relieved. Surely it's not just more rotations = more failures right? Way too simple and could easily be debunked by something like a high-rpm refinery grade pump or compressor vs an el-cheapo lower-rpm pump/compressor (like A/C in a budget model car or cheapest possible ANSI sewer pump, etc). # of rotations vs life expectancy just isn't a correlatable comparison

Quote:
Originally Posted by wsume99 View Post

that changes the calculus a bit
That was algebra btw, and unlike Assassin I am trying to be an ass tongue.gif
post #87 of 105
Quote:
Originally Posted by 4th-horseman View Post

Also in a different place mentioned how I can't begin to understand the correlation between reliability and spindle speed. Before even starting to mine data to support the claim, if someone could indicate a basis for this claim that makes any sort of sense I'd be relieved.

The short explanation is that faster-spinning drives are closer to the noise than slower spinning ones.

The longer explanation is that reading (and writing) the drive involves a signal to noise ratio. The faster the drive spins, the lower the SNR. This is obvious if you give it a moment of thought. Why not just spin the drives at 10K or 15K rpm for higher performance? Besides requiring more power, the answer is that the SNR is lower at higher speeds. Once the SNR gets too low, you have a lot more sectors failing to read (or write) properly that cannot be fixed even with the ECC the drives use. So, as everything about a drive deteriorates with age and abuse, the faster-spinning drives are more likely to have problems before the slower spinning drives, since the faster ones are operating closer to the noise to begin with.
Edited by jim2100 - 6/21/13 at 9:33am
post #88 of 105
Quote:
Originally Posted by jim2100 View Post

Quote:
Originally Posted by 4th-horseman View Post

Also in a different place mentioned how I can't begin to understand the correlation between reliability and spindle speed. Before even starting to mine data to support the claim, if someone could indicate a basis for this claim that makes any sort of sense I'd be relieved.

The short explanation is that faster-spinning drives are closer to the noise than slower spinning ones.

The longer explanation is that reading (and writing) the drive involves a signal to noise ratio. The faster the drive spins, the lower the SNR. This is obvious if you give it a moment of thought. Why not just spin the drives at 10K or 15K rpm for higher performance? Besides requiring more power, the answer is that the SNR is lower at higher speeds. Once the SNR gets too low, you have a lot more sectors failing to read (or write) properly that cannot be fixed even with the ECC the drives use. So, as everything about a drive deteriorates with age and abuse, the faster-spinning drives are more likely to have problems before the slower spinning drives, since the faster ones are operating closer to the noise to begin with.


This might be true in the old days of HDD's and even in HDD's in general- but today's standards it's not a big deal. It's not like 7200rpm is some new ultra high spindle speed, it's been around a really long time and I'm not sure the technology is all the complicated these days. It's really been the standard HDD speed for a long time now. WD has long made 10k rpm drives (like velociraptors) as their answer to high performance drives - and those drives are consistently some of the most reliable HDD's in the industry. The endurance of them is well known- and WD even backs it with a 1.2million MTBF rating and a 5 year warranty. The drives are tested and proven in enterprise duty too.

Slowing down to 5400rpm to keep things working or prevent errors these days would suggest the HDD and the technology is crap. It;s not. The lower spindle speed was to reduce heat, and noise and temps. It was introduced when drives were noisier and hotter and more power hungry. These days- the newest 7200rpm drives are quite competitive with 5400rpm on those factors- and often better in performance. Once you factor in the cost is lower in many cases for 7200rpm drives then the choice to by a 5400rpm is not well supported. There is not as much benefit from 5400rpm these days- and the performance penalty is still very real.
post #89 of 105
You really do not know what you are talking about (not unusual for you, I have noticed).

The 10K and 15K rpm drives are lower platter density and/or 2.5" platters, so you are comparing apples to oranges.

I am comparing identical drives except for whether they spin faster or not. And in that case, what I have stated is an obvious consequence of how HDDs work. It has nothing to do with whether the technology is "crap". Whether the HDD technology is the best or below average, the SNR is lower when you spin it faster. It may be a small effect, but it is there.

Getting 20 or 30% faster sequential speeds (and slightly lower average seek times) at the cost of a slightly lower reliability, higher heat load and slightly higher power usage may be a reasonable tradeoff for some usage cases, but the tradeoff is real. You don't get something for nothing. Most people with media servers care more about maximizing reliability and minimizing heat and power than they care about a little extra performance.
Edited by jim2100 - 6/21/13 at 12:37pm
post #90 of 105
Quote:
Originally Posted by jim2100 View Post

You really do not know what you are talking about (not unusual for you, I have noticed).

The 10K and 15K rpm drives are lower platter density and/or 2.5" platters, so you are comparing apples to oranges.

I am comparing identical drives except for whether they spin faster or not. And in that case, what I have stated is an obvious consequence of how HDDs work. It has nothing to do with whether the technology is "crap". Whether the HDD technology is the best or below average, the SNR is lower when you spin it faster.

So your trying to tell me that a 7200rpm drive with 1TB platters is significantly different in realiability than a 5400rpm with the same 1TB platter density ?

Example:

WD GREEN 3TB versus a 7200.14 Seagate Barracuda.

????


I don't believe that for a split second. There is no data proving either.

7200rpm is not a fast hard drive speed. It's a standard speed. I don't think it's a major reliability concern, Hard drives have been this speed for 20 years. I can remember when the platter density was 20GB, 40GB, 80GB, 160GB, 250GB, 500GB... etc,..

Last couple generations have been 500GB, 666GB, 800GB, and now 1TB. The old 7200rpm Seagate used 5 platters of 600GB to hit 3TB size. Today's version uses 3 platters of 1TB. That's why the newest drives are faster- but I have not seen any significant change in reliability in the last 15 years I've been buying hard drives.


I have 40GB drives from 10 years ago that actually still work- (well it just died actually)

Platter density has been increasing with each new generation of hard drives, but reliability has not changed all that much. If your theory was true MFG would make a 1TB drive instead of a 3TB/4TB drive that is 5 times as reliable and market it with a 5 or 10 year warranty with the promise it's very reliable. I am sure consumers would jump on that. The reason they don't do it is because the reliability is a crap shoot with all HDD's and they all last about the same (consumer drives)

How about a Seagate 5900rpm 4TB versus say a WD 3TB 5400rpm ?? You think the 5900rpm Seagate is less because it spins a little faster?

Spindle speed does not determine HDD reliability and there is no intelligent correlation to made between them. It's a nice theory that on the surface might look ok- But there is just too much other stuff going on that helps determine HDD reliability. Spindle speed or anything related to it is not the primary cause of HDD failures- and therefore it makes sense it's not a major factor in reliability too.

Of coarse this is my opinion wink.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Home Theater Computers
AVS › AVS Forum › Video Components › Home Theater Computers › New Seagate NAS HDD's