Sign in with
Sign up | Sign in
Your question

Two harddrives...Do i need to Raid??

Last response: in Storage
Share
September 24, 2006 4:13:33 PM

im planning on building my own PC, for the first time --- I ordered Two 250GB Seagate SATA Hard Drives...

Reason i got the 2nd hard drive was mainly to put all my Music, Videos, Movies etc into that Drive....and use my 1st hard drive for my OS, Applications, Games.

I keep reading though people who do RAID & Partions etc...

Since i am using two Hard drives should i be doing a RAID or Partion? Is it okay to just use each as a seperate hard drive without losing performance?

any type of reply will be appreciated. thanks

More about : harddrives raid

September 24, 2006 4:37:20 PM

u could use raid

but for ur intentions, i don't thnk u should

u can certainly use the 2 hdd's not in raid
September 24, 2006 4:47:04 PM

Raid-0 all the way dude. Those who say don't, have probably never had it.

It's twice as fast, and you get to keep all your HD space.
Related resources
September 24, 2006 5:12:14 PM

No raid for you. There are many downfalls to raid. I suggest you stick with a non raid setup if you are beginning.

BTW I do have a raid 0 with 2 raptor drives. That being said I also have a basement full of servers for all my data.
a b G Storage
September 24, 2006 5:15:37 PM

RAID 0 will let your PC boot, load programs and load games a little faster. RAID 0 is the topic of a lot of discussion whether it actually makes THAT much difference over a fast single drive.
I personally use RAID 0 on my system, would never go back to a single drive again, I think it makes a noticable difference.

Just be warned, if either one of the drives fail running RAID 0, you lose EVERYTHING.

Have I ever had a drive fail using RAID 0? No
Have I ever known someone one who had RAID 0 fail? Yes.
No matter what kind of drives you run in any configuration, should you back up your important files? Yes!
September 24, 2006 9:30:23 PM

It's not twice as fast. The only place it's fastER is in synthetic benchmarks. In single-user applications the gain is minimal to non-existent. If you are into posting and comparing HDTach numbers, get raid-0. If you aren't, and you aren't running a server, then use a single raptor (or some combo for whatever you need).
September 24, 2006 9:56:58 PM

Quote:
It's not twice as fast. The only place it's fastER is in synthetic benchmarks. In single-user applications the gain is minimal to non-existent. If you are into posting and comparing HDTach numbers, get raid-0. If you aren't, and you aren't running a server, then use a single raptor (or some combo for whatever you need).


Thats bullcrap. My PC's noticeably faster if I raid-0 my two raptors, compared to using them as single drives.
September 24, 2006 10:45:28 PM

Use RAID 0 on your two 250GB hard drives and get another hard drive to store all your music, vids, docs, etc. in, that is, if your budget permits. If not, then don't use RAID anymore. Just let them run as 2 separate hard drives.
September 24, 2006 11:31:16 PM

I assume you're considering RAID0 to get more performance. Well, from my experience, using RAID0 indeed gives you a great increase in performance, and I am all for it. Actually, I must say that at first I was quite skeptic, because technically it only gives you better sequential speed, while access times remain the same. But in fact, in day to day use I was surprised what a boost you get from it.

However, and this is very important: for your mental health, **DO NOT** put anything on a RAID0 disk that is not backed up!! Beyond having x2 the risk of loosing the info (it is enough that one of the drives crashes) it is also much more complicated - and expensive - to recover a RAID0 disk.

As I see it, these are your options:
1. Have the two in RAID0. In this case, you *must* have another HD in the system for backing up your data, or otherwise back up to DVD in a regular basis.

2. Forget about RAID. In this case you can use partitioning to gain a bit of speed. Put at the beginning of the first HD a 20-30GB partition just for Windows+Apps, and at the begining of the second HD a partition just for your swap file (size it to about x2 your RAM). You benefit here because the partitions are both at the fastest areas of the HDs, and because you can simultaneously r/w Windows ans swap files. In fact, I would also have another partition at the begining of the second drive - say right after the swap parition - for temporary files. You can move all your TEMP directories to there and hopefully gain a bit more speed there.

3. This one is a bit more involved: most RAID controllers now allow having different RAID configurations on different parts of the same HD. What you can do is allocate a 20GB partition at the beginning of each drive, and have only these two partitions in RAID0, while the remaining parts of the drives are non-RAID. This will give you one 40GB RAID0 partition for Windows, apps, swap and tempfiles, while your remaining partitions (you should be left with two 200GB partitions now) will contain your files and other important data. Of course you can play around with these things further, and allocate additional partitions for different types of files etc. For instance, you can have your downloads, which I assume are disposable, on another pair of RAID0 parititions.

To summarize: RAID0 is best for system and temporary files, where you gain the most speed while having little risk of loosing data. As to your data - always back up your important stuff, and especially anything on RAID0 you care about. If you keep parts of these HDs non-RAID, you can use one HD to backup the other, for instance.

Good luck!
September 25, 2006 12:49:55 AM

Quote:
It's not twice as fast. The only place it's fastER is in synthetic benchmarks. In single-user applications the gain is minimal to non-existent. If you are into posting and comparing HDTach numbers, get raid-0. If you aren't, and you aren't running a server, then use a single raptor (or some combo for whatever you need).


Perfect example. This dude has never had raid. He really wishes he did, and this makes him angry. therefore he tells you it sucks.
September 25, 2006 1:40:21 AM

You could do the RAID 0, get good performance and buy a cheap USB external or NAS to back up your important stuff to.

- OR -

You could set the drives up as RAID 1 and have fault tolerance.

But, if you do the two separate drives as you suggest then the really important stuff like photos or purchased music should be regularly copied to the second drive.
September 25, 2006 5:35:45 AM

Quote:
It's not twice as fast. The only place it's fastER is in synthetic benchmarks. In single-user applications the gain is minimal to non-existent. If you are into posting and comparing HDTach numbers, get raid-0. If you aren't, and you aren't running a server, then use a single raptor (or some combo for whatever you need).


Thats bullcrap. My PC's noticeably faster if I raid-0 my two raptors, compared to using them as single drives.
I say that two raptors in raid-0 is NOT twice as fast as a single raptor for single-user use (games, etc.). You say that is "bullcrap". Ok, prove it's bullcrap. Show me numbers that place the raid set-up as twice as fast. And, based on the comment which you dismissed, I'm not talking HDTach numbers, I'm talking actual single-user benchmarks. Or, just show a substantial improvement, if not a doubling of performance.

And again, with real-world single-user benchmarks, not synthetic I/O measures.
September 25, 2006 5:38:27 AM

Quote:
It's not twice as fast. The only place it's fastER is in synthetic benchmarks. In single-user applications the gain is minimal to non-existent. If you are into posting and comparing HDTach numbers, get raid-0. If you aren't, and you aren't running a server, then use a single raptor (or some combo for whatever you need).


Perfect example. This dude has never had raid. He really wishes he did, and this makes him angry. therefore he tells you it sucks.

:roll:
September 25, 2006 5:41:58 AM

http://www.storagereview.com/articles/200406/20040625TC...

RAID helps multi-user applications far more than it does single-user scenarios. The enthusiasm of the power user community combined with the marketing apparatus of firms catering to such crowds has led to an extraordinarily erroneous belief that striping data across two or more drives yields significant performance benefits for the majority of non-server uses. This could not be farther from the truth! Non-server use, even in heavy multitasking situations, generates lower-depth, highly-localized access patterns where read-ahead and write-back strategies dominate. Theory has told those willing to listen that striping does not yield significant performance benefits. Some time ago, a controlled, empirical test backed what theory suggested. Doubts still lingered- irrationally, many believed that results would somehow be different if the array was based off of an SATA or SCSI interface. As shown above, the results are the same. Save your time, money and data- leave RAID for the servers!
September 25, 2006 5:45:26 AM

http://www.anandtech.com/printarticle.aspx?i=2101

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.

If you do insist on getting two drives, you are much better off putting them into a RAID-1 array to have a live backup of your data. The performance hit of RAID-1 is just as negligible as the performance gains of RAID-0, but the improvement in reliability is worthwhile...unless you're extremely unlucky and both of your drives die at the exact same time.

When Intel introduced ICH5, and now with ICH6, they effectively brought RAID to the mainstream, pushing many users finally to bite the bullet and buy two hard drives for "added performance". While we applaud Intel for bringing the technology to the mainstream, we'd caution users out there to think twice before buying two expensive Raptors or any other drive for performance reasons. Your system will most likely run just as fast with only one drive, but if you have the spare cash, a bit more reliability and peace of mind may be worth setting up a RAID-1 array.

Bottom line: RAID-0 arrays will win you just about any benchmark, but they'll deliver virtually nothing more than that for real world desktop performance. That's just the cold hard truth.
September 26, 2006 9:20:49 AM

Have you ever tried it? If you had, you wouldnt be posting this trash. I notice a SIGNIFIGANT speed up of loading in games, esp. BF2 and EQ2. MMO's such as WoW and eq benefit greatly from the extra speed as the streaming nature of the game demands that things be loaded as they come into view.

You sir, as just as bad as BaronBS.
September 26, 2006 11:04:47 AM

The whole argument with RAID 0 and not RAID 0 seems to be as follows;

People who think RAID 0 is faster have set up a RAID 0 array, and they think that RAID 0 is faster.

People who think that RAID 0 is slower have got a non-RAID 0 solution, and they think that their non-RAID 0 is faster (or at least "just as fast").

Basically, if you think it's going to be faster, you're going to see an improvement, even if it's just a perception. It's the way things work - go and ask a scientist about how the placebo effect works, for example (a medical example of basically the same thing).

Nowhere have I seen anyone on any forum do any kind of double-blind testing of which is faster, RAID 0 or non-RAID 0.

The only place we can find said information is from 3rd party websites, people such as Anandtech and Storagereview. Both of these websites have said that RAID 0 is pointless for desktop use.

I would be interested in seeing any independant reviews, tested in direct comparison to a single disk, which would claim that RAID 0 is significantly better in desktop use. I've never seen one.

Anyone seen one?
September 26, 2006 11:29:19 AM

My advice:

1) If you need to install drivers anyway to take full advantage of your SATA controller, consider installing RAID 0 and doing backups
2) If you just want your PC to work and it has an IDE compatibility mode, don't bother.

If this is your first build don't go out of your way to make it harder to upgrade maintain, or backup. When you RAID 0 it, you will be depending on that particular raid controller, and depending on your backup programs to be able to read it, anytime you need to do boot time recovery, you are banking that your tools can read it.

If you are willing to accept that risk, yes it is nice to have one large device which is usually faster.

Do be aware, Whether you choose to RAID or not, you should consider backing up your data right away. Right after install/setup is a nice time, since you won't have to do it again, if a game screws it all up, or new DX install or whatnot.

Just my two cents.

John
September 26, 2006 11:48:57 AM

It's up to you to raid or not. But there's little to no increase in performance. I do have raid 0 with 2x74Gb Raptors mainly for gaming. It's about 5 or more seconds faster in loading my games but not much. I just did it for I needed more space for os and all my games and to try it. It's very bad in terms of performance-to-cost ratio I tell you but it's a little faster on loading my games nonetheless. For you, don't even, just use one for os and game and the other for backups. Is my raid worth it, well no not really. :) 
September 26, 2006 12:18:32 PM

Quote:
Reason i got the 2nd hard drive was mainly to put all my Music, Videos, Movies etc into that Drive....and use my 1st hard drive for my OS, Applications, Games.


If this is why you forked out the money for the second hard drive then set it up for that purpose. As you can tell from the other post it is a very touchy subject.
RAID0 is great if you need games or apps to load faster, how much faster depends on your drives. (just like Chuckshissle said he only sees 5 seconds of improvement)
The downside is the lack of data protection, unless you backup your data regularly.
The only advice I can give you is to experiment with the various options until you find one you like and are happy with. Think about what is more important having 500GB of space or 250GB?

Use this as a chance to gain some knowledge and experience.
a b G Storage
September 26, 2006 12:29:22 PM

Quote:
http://www.anandtech.com/printarticle.aspx?i=2101

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.

If you do insist on getting two drives, you are much better off putting them into a RAID-1 array to have a live backup of your data. The performance hit of RAID-1 is just as negligible as the performance gains of RAID-0, but the improvement in reliability is worthwhile...unless you're extremely unlucky and both of your drives die at the exact same time.

When Intel introduced ICH5, and now with ICH6, they effectively brought RAID to the mainstream, pushing many users finally to bite the bullet and buy two hard drives for "added performance". While we applaud Intel for bringing the technology to the mainstream, we'd caution users out there to think twice before buying two expensive Raptors or any other drive for performance reasons. Your system will most likely run just as fast with only one drive, but if you have the spare cash, a bit more reliability and peace of mind may be worth setting up a RAID-1 array.

Bottom line: RAID-0 arrays will win you just about any benchmark, but they'll deliver virtually nothing more than that for real world desktop performance. That's just the cold hard truth.


I run RAID 0 and I disagree completely.
I would never, never, ever for any reason go back to a single drive. I have never done any benchmarking on my drives. I can tell the difference just using the machine, and it's a pretty big difference.
September 26, 2006 1:19:54 PM

>I run RAID 0 and I disagree completely

This is exactly what I mean. You can't point to anything saying from an objective POV that your RAID 0 system is faster. Equally, no-one can point to anything saying from an objective POV that a single-drive system is faster.

Subjectively, you think it's better. That's great. Subjectively lots of people think a single drive is just as good. That's also great. Sorry to be so relentlessly postmodern, but hey, if it feels good for you, that's great! :) 
September 26, 2006 2:18:52 PM

i run 5 hard drives
3 are the exact same model (250 gb sata1 seagate barracudas)
1 is a slighlty older 80gb diamond max
1 new 300gb sata2 seagae barracuda

i run my os off one of the 250's the other two are set up in raid 0 the older 80gb contains my music files whilst the newer drive is used to backup everything important (music videos documents drivers savegames etc..) in terms of real world usage (i dont really bother benching my hard drives) i would say that in terms of write times (when transfering large videos (vob's) the order is (quickest first)

1.)raid 0 setup
2.) 300gb backup drive
3 250gb drive
4.) older 80gb drive

now i know this is the order you would expect but im just trying to proove that raid 0 is noticably quicker and before you ask no i dont time them so dont have any figures. but i must say that it is a joy to use the raid 0.
now in terms of access times i would say the the 300gb drive is far more responsive when simply browsing thru folders than any of the others

1 300gb backup
2.) 250gb/raid0 setup
3 old 80gb

please do not try to say that raid 0 speeds are just perceived improvements based upon us wishing they were better when that is clearly not the case


to Nedkt
Have you ever used Raid??
and for what did you use it?


back on topic tho

use seperatly
until you have a backup medium i would keep your media collection seperate from everything and then run everything off the other drive
(apart from maybe the swap file but thats another discussion)
once you get happy with this setup then consider going raid

oh and i have had a failed raid 0 never quite found out the reason for it (wasnt a failed drive) so i rebuilt the array reformated ran a few data recovery apps and managed to get back 60% of my data( god knows how, i was very very Lucky). which is what prompted me to get the backup.

finally a bit of nostalgia for you
i remember the days when it was rare to even see a d:\ on your computer it used to be just a A:\ and a C:\ but now with multiple hard drives two cd drives my phone my pda. my usb card reader even my sodding printer P:\ can be acheived with out even trying. anyone know what happens after z:\?
September 26, 2006 2:54:12 PM

>after z:

You really don't want to go there... Once Windows has hit Z: it goes back to A: and starts overwriting old drives. All kinds of problems.





...




I'm lying really. I don't know what happens ;) 
September 26, 2006 3:39:56 PM

You sir are misinformed. If you think about what you're saying you will realize just how stupid you are. Game loading times are severely bottlenecked by the CPU performance and its ability to decompress the files and load them into memory. MaximumPC did a benchmark involving 2 Raptors in RAID 0 and Doom 3 loading times because readers were curious. And guess what they found, loading times acctually increased (by .5 seconds or so), not decreased because the HDD's are not the bottleneck.

Think about the circumstances people are talking about. When doing large file transfers, dealing with mass photoshop files, and constant data transfers RAID 0 is faster. I used it for 2 years or so, and it really wasn't worth it to me. I have the same basic use for hard drives as the OP. Just dealing with music and video and documents, RAID 0 offers no benefits only increased risk because you risk 2 drives worth of information instead of one.

Think before you speak. Put things in context and you will avoid sounding like Baron....
September 26, 2006 3:46:41 PM

I have an example of single user desktop practial application where RAID0 has had a noticable and measurable difference. I recently built computers for myself and for a friend of mine. The computer had support for ultra320 SCSI raid, SATA150 raid, and IDE raid. Our system builds were EXACTLY the same, except for i had chosen to go with a seagate cheetah, a 15K RPM Ultra320 scsi drive for those who dont know, and my friend chose to go with 2 WD Raptors in RAID0. We each claimed each's system would be faster, so of course we had to test it. We placed our computers side to side, and low and behold, the Raptor RAID0 beat out my 15K cheetah in all cases, by a noticable measureable difference! Ever since then, i have reccomended to all my enthusiast customers to purchase a dual raptor RAID0 configuration, with a large caviar drive on the side to back up their information.
September 26, 2006 3:50:12 PM

Uh, of course the seagate's going to lose out to the Raptors - the Seagate is heavily optimised for multi-user server applications and the Raptor is desktop optimised.

If you'd compared *one* Raptor the the Seagate it'd've outperformed it for almost all desktop applications - where it's access time and latency that matter, not burst or max transfer rate.

You're comparing apples to oranges there. A fairer and more representative test would've been a RAID 0 array in one machine and a single disk in the other - that way you could see the difference properly.

We're back to the Qualitative evidence, not quantitative! Subjective, not objective!
September 26, 2006 3:52:28 PM

It shouldn't beat it in all cases. The cheetah should beat it in seek times due to the high RPM and lower rotational latency. However, what did you use to test the hard drives? if you are using synthetics then you are getting skewed results. In a tradtional enviornment a single user will not notice enough diffrence to justify the increased risk unless you are in one of the 3 enviornments I identified. Those aren't the only place you will notice a diffrence, but they are the main areas where you will benefit from the performence.

My main point is don't for 1 second think RAID 0 acctually helps in gaming. I have played FPS's MMORPG's Strategy games and more and RAID 0 doesn't do jack for you.

Off to public speaking... I shall return
September 26, 2006 3:53:55 PM

It's not as easy as that when doing comparisons between SCSI drive and desktop drive. They exist in totally different worlds. You compared apple to oranges.
Explain here regarding to firmware.
September 26, 2006 4:38:37 PM

Quote:
We're back to the Qualitative evidence, not quantitative! Subjective, not objective!


Don't even get me started on that stuff... crazy auditing has got me all wound up about evidence and persuasive not conclusive, GRRR. lol 8O
September 26, 2006 7:38:32 PM

THIS WHOLE THREAD =

"Sunk cost fallacy!"
"Science!"
"Sunk cost fallacy!"
"Science!"
"Sunk cost fallacy!"
"Science!"
Ad Nauseum!

People! For chrissakes! You got schnookered by a greasy salesman! The people who sell harddrives are laughing all the way to the bank! They figured out how to get a sucker to pay double: with one word, faster, you can get the adolescent male homo sapien to simultaneously disregard his critical thinking skills and intellectual honesty and buy whatever you put in front of him!

Human behavior is economic behavior. The businessman wants to double his profits, and the consumer wants to believe he made a wise investment. The former provides a happy lie, and the latter happily swallows it.

Only, because it is not genuine knowledge - there is not a shred of scientific rigor to it - his freakishly large monkey brain is aware, at least subconsciously, that he does not know. So he tries to enlist other people to swallow the lie, hoping like an addict to appease his craving for self-deception with more self-deception, in this case called group reinforcement. Hence the ferocity of his position in this thread, nomatter how impossible, nomatter how indefensible, nomatter a mountain of evidence to the contrary.

We (should) all be aware there's a long list of products that prey on the ignorant. Especially in this industry for some reason. The Pentium: Expensive Edition, discrete Physics cards, BluRay, anything that has the name Dell on it, let alone PSU's with poor amperage, HDD's that fail all the time, batteries that explode, air and liquid cooling systems that don't... the list goes on and on. I mean, they're still selling discrete AGP video cards.

Add RAID0 to the list.
September 26, 2006 8:27:29 PM

Careful who you reply to, people in this forum take that seriously. I know you're not slamming me, it is just a friendly tip.

I concur on a general level what you are saying, but I do have to step in with regards to BluRay's concept, not BluRay itself. I am a HD-DVD proponent, but there is a growing need for large capacity DVD's. There isn't a big need yet, but as we move towards mainstream HD, including 720p and 1080p, DVD's cannot hold their own. DVD's are recorded in 480p, tack on DTS surround sound, a long movie and your disc is reletively full. Note I say reletively, there is of course some space, but not much. If you toss 1080p into the mix you more than double the image detail and your overall image size jumps through the roof.

I am not saying that the HD-DVD/BluRay's should be jumped on right now, because they shouldn't. The players and the movies themselves are way too expense to justify the cost. especially when a BluRay player costs more than a HDTV.

Like any product there is a specific segment of the market with which the product fits, RAID 0 included but a better example is RAID 5. The joke is on people when they use the product outside of its intended use, in our case RAID 0. RAID 0 was never intended for gamers, never intended to speed up Windows. RAID 0 was designed for large file transfers, slow hard drives, etc. A computer is only as fast as its bottleneck and the bottleneck itself in a computer is constantly moving, it just depends on what you are doing. In gaming it can be a CPU or a GPU depending on the game and setup, in transfers and image/video editing it is typically the hard drive. So, addressing the bottleneck in context is how you speed up a computer, but sweeping generalizations like this thread which say RAID 0 for everyone is a poor judgment.

People in general, and in this forum, forget to place things in context, and context makes all the difference. Wusy is very good a putting things in context and knows way more than I, which is why I take his advice. These goobers who, just because they use something, think that it is the best thing for everyone are the reason that such premises like RAID 0 for everyone exist.
October 3, 2006 1:17:42 PM

You know what I think, he should RAID 0 if he wants to and only if he wants to. I mean, I would RAID 0 my system if I had the right hardware just so I know what it's like to run a RAID 0 config. Do whatever you like, man. We're all behind you.

I'd RAID 0 my car if it's possible. Maybe then I could squeeze out a little more performance out of that old junk heap. And while I'm at it, I'd probably RAID 0 my fridge too. Maybe it'd make it run cooler or something. Then I'd RAID 0 my dog...
October 3, 2006 4:13:09 PM

Quote:


People who think that RAID 0 is slower have got a non-RAID 0 solution, and they think that their non-RAID 0 is faster (or at least "just as fast").


You really said that? LOL man where do these people come from? yep people out there think non-Raid is faster, yep look at you, living on what planet?
October 3, 2006 4:18:37 PM

Quote:
Nowhere have I seen anyone on any forum do any kind of double-blind testing of which is faster, RAID 0 or non-RAID 0.


ROFL, wow, double blind testing of RAID 0 to a single drive.

Sorry about making two posts but i just didn't think there would be any more stupid comments left in that post...little did I know.

Well MORON anything scientific requires DATA, perception alone IS NOT DATA.

If you are testing a drug to see its effictiveness, it is reported using DATA.

OPEN UP A PDR, look at every single drug in there, omg look what you'll find, DATA SUPPORTING ITS EFFECTIVENESS. Data like prostate size decreased 2cm, bowl movements increased 1-2 additional times a day.

DATA, not perception......like uhhh yea umm i think that propecia is making my hair grow back, DATA like no follicle loss in the bitemporal lobe. Or in your terms, perception would be uhhh yea dude this computer seems slower than the one i used two seconds ago, maybe its cause i used it after the one i used two seconds ago and like i yawned in between and like the yawning made me tired which made me think things were going slower or like maybe really it was faster but i just throught it was slower cause like i dont have any real data to use.

how people like you exist in society, i'll never understand.
October 3, 2006 4:21:45 PM

Quote:
No numbers=mind tricks



LOL, I spend two posts trying to say what lottie says in 4 words and an equal sign, haha
October 3, 2006 9:17:22 PM

Hey shaba;

My point, which you so expertly missed, was that you *need* proof to make a case that RAID 0 is faster than non-RAID 0.

Did you read the rest of my post? I believe the same things you do. I believe that you need data before you can make a decision. I believe that the standard measurements of STR, BTR, Random reads and writes, seek time, etc, are all well and good, and indeed these are *objective* measurements, but that there is definitely a place for *subjective* measurements as well.

You need both to make a conclusive statement about how RAID 0 is "faster". Why? Because the term "faster" when it applies to Hard Drives does not just mean "transfers data faster", it does not just mean "has a lower latency", it doesn't just mean anything - it's a conglomeration of different factors, all having a bearing on the whole.

Some people will use a machine and see that their game loads that bit faster, and they'll say "this is a faster machine". Some people will use a machine and see that this or that bit of random data gets accessed faster, and they'll say "this is a faster machine". According to raw benchmark data, either could be the "faster" one, depending what you're testing for.

Subjective measurements are required.

But, since you're "Mr "Data, not perception"", please show me some objective proof that RAID 0 is "faster" than non-RAID 0, making certain that your proof satisfies all possible meanings of the word "faster".

Or are you just the troll I think you are?
October 3, 2006 9:33:47 PM

wow look at you, that was beautiful...you realized what you said was moronic so you take this philosophical approach to try and sound like you're still right because of the exact definition of the word faster? lol

everyone knows that raid 0 is going to perform better/FASTER in just about 99% of benchmarks so long as there is no bottleneck, what everyone doesn't seem to realize is that the actual amount that its FASTER is dependent on your use.

and as benchmark after benchmark shows, average usage yields about a 10% performance increase or 10% FASTER monkey boy.

anyways I'm done with this, grow a brain u fu*kin idiot cause you really need to learn when you're wrong and just suck it up or you're going to spend the rest of your life caught up behind your computer screen asking people if they feel like the computer they used was faster or slower with your double blind experiments lololol
October 4, 2006 2:11:56 AM

Quote:
Some people will use a machine and see that their game loads that bit faster, and they'll say "this is a faster machine".


Let's try this again:

Quote:
Game loading times are severely bottlenecked by the CPU performance and its ability to decompress the files and load them into memory. MaximumPC did a benchmark involving 2 Raptors in RAID 0 and Doom 3 loading times because readers were curious. And guess what they found, loading times acctually increased (by .5 seconds or so), not decreased because the HDD's are not the bottleneck.


Now reevaluate what you said. I could not find the actual link on the maximumpc website, but I found someone who posted the results of the test here

Read that, then come tell me RAID 0 decreases gaming load times.

Quote:
Some people will use a machine and see that this or that bit of random data gets accessed faster, and they'll say "this is a faster machine". According to raw benchmark data, either could be the "faster" one, depending what you're testing for.


Well you see the problem with that statement is it is amibigous at best. Is data acccessed under RAID 0 faster? Not per se, the actual rotational latency and seek times are the same, therefore not faster. However, the transfer rates from disk to host are faster because under RAID 0 is alternating receiving strip blocks of a predetermined size, so theoreticly the hard drive transfer speed doubles. Note: theoretical increase.
October 4, 2006 8:24:49 AM

Troll. Thought so.
October 4, 2006 8:29:54 AM

You're missing the point too, 'Fly...

My point is that you can't categorically say "this is faster", or "this is faster", because there's so much dependant on perception as well as theoretical bandwidth rates.

My personal opinion is that RAID 0 does nothing for a single user (it's a godsend for servers, though, that's why I use it at work).

It *may* decrease loading times for *certain* games in *certain* situations with *certain* hardware configurations, but in general the latency of sending stuff through the RAID controller, and the extra CPU overhead, slows everything down.

Basically, I agree with everything you said. Please re-read my posts and you'll see that at no point did I ever state that *I* thought RAID 0 was faster...
October 4, 2006 12:16:57 PM

Point taken. You were saying there is room for subjective measurement, and I tend to stay away from it because it leads to biased decisions, in general. So when you said "some people ... feel thier system is faster" I was just trying to clarify that it doesn't. You are right that you didn't say "I believe it makes a single user system faster." I just hate subjective measurements because they aren't comparable so how are you supposed to compare two products? If after all objective means of measurement two things come out equal... then you have to fall back on subjective measurements, my point was to establish that I hate them lol. :wink:

Alright, now that we are on the same page im off to Jack in the box, hmmm good breakfest.
!