Originally posted by: ReMeDy{WcS}
I didn't know I was even talking to you.
Well, you did quote my post you know
What they need to do is get this drive to have a 250GB+ capacity before it will be accepted by the performance community.
First post, registered to add to the thread
Anyway, this drive is really perfect for me... I've got an A7N8X Deluxe with a 46GB IBM DeathStar GXP75...somehow it's still alive (1999 model too!).
Anyway it's making nasty grating sounds, and feeling a little slow so i plan to upgrade and get two of these drives to run in RAID0. Not sure how good the controller is on the Asus board, hopefuly it's ok.
I don't want two huge drives (mainly caus of the price), i think about 70-80GB is a nice amount of space So 2x36GB drives will be nice
Would you in the know advise for getting a SATA drive this early, or leaving it for a while?
I don't really see how leaving it would make alot of difference since most of the drives seem to be similar (apart from the 10k ofc!)
JP out!
Anyway, this drive is really perfect for me... I've got an A7N8X Deluxe with a 46GB IBM DeathStar GXP75...somehow it's still alive (1999 model too!).
Anyway it's making nasty grating sounds, and feeling a little slow so i plan to upgrade and get two of these drives to run in RAID0. Not sure how good the controller is on the Asus board, hopefuly it's ok.
I don't want two huge drives (mainly caus of the price), i think about 70-80GB is a nice amount of space
So 2x36GB drives will be niceWould you in the know advise for getting a SATA drive this early, or leaving it for a while?
I don't really see how leaving it would make alot of difference since most of the drives seem to be similar (apart from the 10k ofc!)
JP out!
I agree that Winmark is messed up, but that particular bug is not the fault of WinMark, but of OS's which don't properly support Large ATA drives. HDTach will do the same thing. Once the STR test hits 128GB, for some reason windows tells the drive to go back to the beginning and continue testing until the full capacity is reached. That's why the end of the drive seems to perform like the beginning.
He quoted NFS4.
The performance community has already chosen SCSI, and it isn't 250GB+. Most performance oriented people will sacrifice some capacity for better performance. The final revision of the Raptor, makes it very attractive to people who don't feel like dealing with SCSI for whatever reasons. For WD, I don't know what they are doing with this drive, as they are clearly trying to target the enterprise market, but its performance doesn't put it anywhere near SCSI in that market. What they end up with is a drive with top notch performance in the wrong areas. Now the drive doesn't really have an intended market.
I would really like to see a direct comparison between the performance of this drive and the Maxtor DiamondMax Plus 9 SATA. The review linked-to in the news section shows that the Maxtor is quite a performer. Looking at the HDTach results gives us a maximum speed of ~63 MB/s and a minimum speed of 32 MB/s and an access time of ~12 ms. Except for the access time, it is essentially in the same ballpark as the Raptor.
Anish
Anish
I'd like to answer the question about the intended market, since apparently I'm *in* the intended market.
I'll be building a Prescott/Canterwood PC in a few months, and a pair of these puppies RAIDed with the ICH5 will be my OS/programs/gaming drive, with possibly a WD Caviar 200GB PATA drive for mass storage. No extra cost for the controller makes this an obvious choice, and at about half the price per GB of SCSI drives + controller which would, for gaming and other non-server use, perform similarly. The low space is meaningless since you can get tons of 7200RPM space for mass MP3's or whatever on the cheap. You just need enough 10kRPM space for your OS and other programs to really improve performance; your data can go elsewhere.
So this (the non-beta version anyway) seems to me to be targetted at the enthusiast/high-end/gaming desktop market. Other than the Maxtor Atlas 10k, the only drives that score better in SR's Gaming DriveMark are 15kRPM SCSI drives. Given the optimizations WD made, I think they might be balking at the idea of selling to enterprise and focusing on the enthusiasts/gamers' end of things.
I, for one, have been waiting a long time for 10kRPM to hit ATA and am glad to finally see it happen, thanks in part to SATA and the push to differentiate SATA drives from PATA. I actually somewhat saw this coming--believed the SATA "revolution" would be a catalyst to move ATA technology forward--something which is long past due.
The hard drive is the last big bottleneck in the PC. Until solid state becomes an affordable reality (polymer memory arrays or whatever), I'm all for whatever improvements come to pass.
I'll be building a Prescott/Canterwood PC in a few months, and a pair of these puppies RAIDed with the ICH5 will be my OS/programs/gaming drive, with possibly a WD Caviar 200GB PATA drive for mass storage. No extra cost for the controller makes this an obvious choice, and at about half the price per GB of SCSI drives + controller which would, for gaming and other non-server use, perform similarly. The low space is meaningless since you can get tons of 7200RPM space for mass MP3's or whatever on the cheap. You just need enough 10kRPM space for your OS and other programs to really improve performance; your data can go elsewhere.
So this (the non-beta version anyway) seems to me to be targetted at the enthusiast/high-end/gaming desktop market. Other than the Maxtor Atlas 10k, the only drives that score better in SR's Gaming DriveMark are 15kRPM SCSI drives. Given the optimizations WD made, I think they might be balking at the idea of selling to enterprise and focusing on the enthusiasts/gamers' end of things.
I, for one, have been waiting a long time for 10kRPM to hit ATA and am glad to finally see it happen, thanks in part to SATA and the push to differentiate SATA drives from PATA. I actually somewhat saw this coming--believed the SATA "revolution" would be a catalyst to move ATA technology forward--something which is long past due.
The hard drive is the last big bottleneck in the PC. Until solid state becomes an affordable reality (polymer memory arrays or whatever), I'm all for whatever improvements come to pass.
The real question is, how big is that market?
WD sure as **** can't rely on such a small market to find development of a 10K RPM S-ATA series of drives.
Im not asking because I don't see a place for it from the consumers point of view, Im asking where in belongs in the business world,
None. But then, how long did it take for 7200RPM hard drives to finally make it into OEM systems? Quite a while, if my memory serves me. I was quite amazed to see most OEM systems using 5400RPM *years* after I started using 7200RPM's in self-builds without a second thought.
THIS drive will probably NEVER be in an OEM system (except through end-user upgrades). By the time OEM's bother putting 10k's in their systems, we'll have at least one, probably more, new generations of the technology.
Well its a nice drive but not all that impressive. When you can buy a Maxtor 10K IV for a lil over $200 for the same capacity it kinda pales in comparison. The drive will probably command around $175, so while its performance is really good its still a poor man's scsi while not being all that much cheaper. You still need to buy an adapter so while scsi adapters are more, a scsi system is flat out faster than any ide setup. If someone is going to pay that kind of money for a drive with that kind of capacity they might as well get a great scsi drive like 10K IV which will mop the floor with the Raptor. In the land of 10K drives it might be a Raptor but the 10K IV is a T-Rex
Random Variable
Lifer
- Aug 10, 2001
- 10,420
- 2
- 0
WTF? It posted a higher score than the 15K drives in the SR High-End Drivemark! That's insane! :Q
Hey Xentropy, guess what is used to store data on CD/DVD media? You've got your aluminum stamped variety, but the ones you and I buy to burn (R or RW) use Pthalocyanines, a form of polymer arrays...
Anybody who's actually read the SR article is likely to go along with their analysis- the drive is aimed at entry level servers, where the hot swap features of sata come into play. This is also an extremely beta drive- single platter and some obvious firmware updates in the pipe.
And the advantage over scsi is obvious- price. SATA is in its infancy, so the prices are still relatively high. That will change very quickly, as more boards include it, and as SATA raid controllers become available. Price out 4 scsi 36gig drives and raid controller, do the same for an ata system of similar capacity- there's no reason to believe that sata will be more expensive than ata in the long run.... Multiply out those savings thru a big organization, and factor in the serviceability nightmare that is scsi, and it looks like a winner to me...
And the advantage over scsi is obvious- price. SATA is in its infancy, so the prices are still relatively high. That will change very quickly, as more boards include it, and as SATA raid controllers become available. Price out 4 scsi 36gig drives and raid controller, do the same for an ata system of similar capacity- there's no reason to believe that sata will be more expensive than ata in the long run.... Multiply out those savings thru a big organization, and factor in the serviceability nightmare that is scsi, and it looks like a winner to me...
it is all in the firmware - i wonder what would happen if they tweaked a 15k drive for workstation use instead of server use
In the mean time, raptor will spank all but 3 drives currently on the market in typical desktop/gaming applications at approximately half the total costs of those other 4 drives. Within a month or so, SATA will be a standard integrated feature and the target retail price of the raptor is $150, if my memroy serves me right. Street price may well be lower... The cheapest of those 3 scsi drives, Maxtor Atlas 10k III, roughly matches the raptors performance in destop/gaming applications and has a current street price of 200ish. It's conceivable that the raptors street price may fall close to low 100s. Once you figure in the price of a good scsci adapter you have to fork over at least 2-3 times the cost of the raptor to get a scsci setup that matches and potentially exceed the raptors performance in non server applications.
I agree though that the raptor may not be the best entry lvl enterprise drive...
Originally posted by: Idontcare
Hey Xentropy, guess what is used to store data on CD/DVD media? You've got your aluminum stamped variety, but the ones you and I buy to burn (R or RW) use Pthalocyanines, a form of polymer arrays...
I'm aware of that, but, as you're probably aware, "polymer memory" is a very different thing. I'm too lazy to find some links at the moment (it's late here; just use Google), but the basic idea is spinning polymer on wafers which can store bits via the electrical conductivity of the polymer. In theory, it'd be about a quarter the speed of flash, but be fully nonvolitile (no electricity whatsoever to keep the data indefinitely), and *extremely* cheap per bit. The technology is also fully and easily "stackable" with existing silicon; rumor has it one of Intel's plans for the technology if they pull it off is to put a 1GB "L3 cache" on their processors. Even if it's slow by usual memory standards, it's still extremely fast by the standards of current permanent storage, and its nonvolitility lends itself to believe the next step would easily be to build solid state hard drives out of the technology.
Granted, such applications are probably a decade away. But, then again, technology tends to advance more quickly than most people expect.
Looking at the benches, it sure looks to me like WD tweaked the firmware for workstation tasks rather than server tasks.
As for the prices of RAID controllers, true ATA RAID controllers(that is, not semi-software controllers like the promise/highpoint controllers found on many mobos) aren't exactly cheap.
Have a look at 3Ware's controllers for example.
And what's the "serviceability nightmare that is scsi" exactly?
it is all in the firmware - i wonder what would happen if they tweaked a 15k drive for workstation use instead of server use
GOSHARKS is spot on. One thing avid readers of SR have learned in the past couple years is that firmware optimization can be more important than spindle speed and access time to single user access patterns (that's 99% of us, BTW). So important that, in fact, WD's 10K ATA drive can beat top 15k SCSI drives in 2 out of 4 SR workstation benches.
Food for thought:
The raptor is 16% faster than the next fastest ATA drive (WD2000JB) in SR's office benchmark.
The raptor is 23% faster than the next fastest ATA drive in SR's high-end benchmark.
The raptor is 35% faster than the next fastest ATA drive in SR's bootup benchmark.
The raptor is 14% faster than the next fastest ATA drive in SR's gaming benchmark.
It's not my place to question others' perceived need for hundreds of gigs of storage, but 36GB is plenty for me. I'm currently happy with a single 18GB X15-36LP.
Please note that I've been talking about workstation/single-user usage. Don't confuse that with server performance; two different animals.
Updated Link: SR Head-To-Head Comparison Beta vs Production vs 2000JB.
OMG I'm soooooooo happy. I soooo TOLD you guys that they'd pull it together for their production drive.
Thorin
OMG I'm soooooooo happy. I soooo TOLD you guys that they'd pull it together for their production drive.Thorin
dullard
Elite Member
- May 21, 2001
- 25,945
- 4,535
- 126
Well Pariah, you knew I would have to jump in on this one. After our nice discussion last week, it is nice to see you admit that an ATA drive can be good performing for non-server uses.
As for the target audience I think it is enormous. Very, very few people currently need more than 36 GB. Think of all the people out there who don't edit videos or download MP3s. Virtually none of them need hundreds of gigabytes of space. I bet you all the money in the world that the top selling drive at major OEMs isn't in the 100 GB range - that is simply since most people don't need that capacity.
However there are lots of us who want an inexpensive way of getting lots of speed. SCSI is too expensive, but this drive seems to give as good or better performance on many home uses while the total cost should be significantly lower. Put in any SCSI drive at Dell and you are talking $500 minimum ($600 for the 15K drives), but when they get around to using this drive it will likely be $300 at Dell. I realize that there is a big Dell markup, but that is what typical home users will see. Near SCSI performance at half the cost - looks like a no brainer for the people without large movie collections on their hard drives.
As for the target audience I think it is enormous. Very, very few people currently need more than 36 GB. Think of all the people out there who don't edit videos or download MP3s. Virtually none of them need hundreds of gigabytes of space. I bet you all the money in the world that the top selling drive at major OEMs isn't in the 100 GB range - that is simply since most people don't need that capacity.
However there are lots of us who want an inexpensive way of getting lots of speed. SCSI is too expensive, but this drive seems to give as good or better performance on many home uses while the total cost should be significantly lower. Put in any SCSI drive at Dell and you are talking $500 minimum ($600 for the 15K drives), but when they get around to using this drive it will likely be $300 at Dell. I realize that there is a big Dell markup, but that is what typical home users will see. Near SCSI performance at half the cost - looks like a no brainer for the people without large movie collections on their hard drives.
This is the harddrive for performance enthusiasts. It has been a long time coming. SCSI drives have been at 15K for a long time now. I don't even mind it being only 36GB; I can just this drive for my OS/Apps and my 120GB WD as my storage drive.
I'm thinking that this would be a good drive for the OS and major apps that are speed sensitive, and that other apps can be stored on a more conventional, larger drive. 36GB really is quite a bit when you think about it. All of my apps and projects can easily fit. Games are another matter...
That's a good question. It's targetting the server market (officially), but even comparing price/performance, it doesn't impress compared to something like the Atlas 10K-IV or even Seagate Cheetah 10K.6. I imagine that many enthusiasts will pick one up, but as Eugene of SR said, the small enterprise market is MUCH larger than the computer enthusiast market.
Remember that this is WD's first try, and that for some reason the drive doesn't support TCQ. Not bad for a first generation SATA drive using a first generation 10KRPM mechanism from a company that hasn't made enterprise products for several years.
Samsung Spinpoint drives are considered the most reliable IDE drives in the SR forums. There's a user (Tannin) there which has sold huge numbers of nearly every drive since MFM was all the rage, and he says he's had exactly one return for a Samsung drive, and that under the circumstances it appeared to be intentionally damaged. (And even then, a quick format revealed only one sector was bad--the drive still worked). Older Samsung drives are reportedly quite terrible though. He has a very interesting (if you're a geek) webpage that has a good history of hard drives, including a few models I never knew existed. Single-drive RAID0 anyone?.
Of course, as you stated, the performance is terrible.
Actually they didn't. SR posted the results and then WD sent production units ASAP. (I would have too with the scores that thing was churning up!)
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter