SR posts final Raptor BM's, huge improvements

[Pariah]

Well, I was wrong on this one. Looks like WD gave up on the enterprise market with their optimizations. It still gets crushed in the server benchmarks, but it beats everyone including SCSI's best in 2 out of 4 workstation benchmarks. Good job to WD with the final version. What were they thinking sending out those beta drives just days earlier?

SR Database

 

[Idontcare]

FYI - if anyone goes through Pariah's link you must click the Sort right below the SR Office Drivemark 2002 text to see the data.

 

[Sunner]

Damn!

Do a head to head with the Raptor and the Raptor BETA.
The High End Drivemark went from 300 -> 524!

To relate to the previous discussion about WinMark 99, I'd be interested in knowing how WM99 got that 183 MB/Sec number in FP98, considdering SATA1 is limited to 150 MB/Sec, and a 32bit 33 MHz PCI bus is limited to 133 MB/Sec...

 

[mechBgon]

That's more like it. I suppose WD was in the hotseat after the initial preview, and coughed up a new firmware revision ASAP.

 

[Idontcare]

Originally posted by: mechBgon
That's more like it. I suppose WD was in the hotseat after the initial preview, and coughed up a new firmware revision ASAP.

It makes you wonder what WD's strategy was/is.

WD's got smart people, and they would have internally evaluated and benchmarked the Beta model long before unleashing it to the world.

They knew it was going to perform poorly, and it's not like they were trying to beat others to the paper-release 10k rpm SATA/IDE market (or are they?) so why rush it when you know a different firmware will significantly improve performance.

Maybe a good strategy, generate 2x the controversy and double your consumers attention span for your product, while leaving the last thing in their mind to be "so beta sucked, but gamma release made things all better and now the Raptor kicks ass". Course, it kicks the beta raptors ass but still underperforms SCSI, but that ain't the takehome msg that's going to stick, now is it?

 

[mechBgon]

Originally posted by: Idontcare

Originally posted by: mechBgon
That's more like it. I suppose WD was in the hotseat after the initial preview, and coughed up a new firmware revision ASAP.

It makes you wonder what WD's strategy was/is.

WD's got smart people, and they would have internally evaluated and benchmarked the Beta model long before unleashing it to the world.

They knew it was going to perform poorly, and it's not like they were trying to beat others to the paper-release 10k rpm SATA/IDE market (or are they?) so why rush it when you know a different firmware will significantly improve performance.

Maybe a good strategy, generate 2x the controversy and double your consumers attention span for your product, while leaving the last thing in their mind to be "so beta sucked, but gamma release made things all better and now the Raptor kicks ass". Course, it kicks the beta raptors ass but still underperforms SCSI, but that ain't the takehome msg that's going to stick, now is it?

Hehe, I like your cynical way of thinking

 

[BD231]

My Samsung Spinpoint is a frikin piece according to that chart! So seek times with the new 10,000rpm drive should be damn good right?

 

[Pariah]

Originally posted by: mechBgon
That's more like it. I suppose WD was in the hotseat after the initial preview, and coughed up a new firmware revision ASAP.

SR had the final revision before they posted the beta numbers. Eugene over at SR said he wasn't going to post the beta numbers until he found other other sites got beta drives and were going to post previews. With the drives sent out that close together it just doesn't make any sense why they sent out the first batch. I also don't think this is a simple firmware revision. Firmware revisions aren't going to improve STR 7MB/s, there must be mechanical differences in the drive too.

To relate to the previous discussion about WinMark 99, I'd be interested in knowing how WM99 got that 183 MB/Sec number in FP98, considdering SATA1 is limited to 150 MB/Sec, and a 32bit 33 MHz PCI bus is limited to 133 MB/Sec...

The test is bogus. All drives are scoring 150MB/s-200MB/s. It's an example of a HD benchmark running out of system cache.

 

[Idontcare]

Originally posted by: mechBgon

Originally posted by: Idontcare

Originally posted by: mechBgon
That's more like it. I suppose WD was in the hotseat after the initial preview, and coughed up a new firmware revision ASAP.

It makes you wonder what WD's strategy was/is.

WD's got smart people, and they would have internally evaluated and benchmarked the Beta model long before unleashing it to the world.

They knew it was going to perform poorly, and it's not like they were trying to beat others to the paper-release 10k rpm SATA/IDE market (or are they?) so why rush it when you know a different firmware will significantly improve performance.

Maybe a good strategy, generate 2x the controversy and double your consumers attention span for your product, while leaving the last thing in their mind to be "so beta sucked, but gamma release made things all better and now the Raptor kicks ass". Course, it kicks the beta raptors ass but still underperforms SCSI, but that ain't the takehome msg that's going to stick, now is it?

Hehe, I like your cynical way of thinking

I'm in the industry (not HD specifically), I kinda know how we kind of people think.

The product engineers either rushed the Beta model out to reviewers because their Dilbert-like manager felt Seagate was going to scoop WD's 10k rpm SATA/IDE market launch (ok, that's cynical) or they've got a super-saavy market engineer (yes, they get called that) who likes to mix group psychology into the equation for obtaining elevated time-zero mind-share.

Either way, we're all left scratching our heads thinking "ok...that's nice, but when the hell is it going to be faster and cheaper and available standard in my next Dell inspirion"

 

[GoSharks]

so much for the entry level server target market...

 

[Sunner]

I still have to wonder who it's targeted at...

If it would be available in the same sizes as other 10K RPM drives, it would be a different matter, but 36 GB only?
That would work for alot of server applications since servers often don't need all that much HD space, but it's server performance is pretty lackluster for a 10K drive.
Enthusiasts will require more space, sure some people might buy the Raptor and a 120 GB 7200 RPM drive, but that's not a very large market.

Major OEM's wont exactly jump on it either, S-ATA is too new, the first 10K RPM, too small.

The performance is certainly very impressive now, but I still fail to see who it's targeting...

Oh and Pariah, out of curiosity, do you have some sort of realtime monitor that checks SR's database for updates or something?

 

[NFS4]

I might actually give a rat's a$$ if it had a capacity greater than 36GB. Otherwise, I won't give it even a second glance.

 

[T9D]

..

 

[Pariah]

Oh and Pariah, out of curiosity, do you have some sort of realtime monitor that checks SR's database for updates or something?

No, I stole some Google pigeons, and they do all the searching for me. Actually, Eugene who runs the site often posts when benchmarks are available on their messageboard.

 

[Remedy]

if it had a capacity greater than 36GB. Otherwise, I

I don't think you're the target WD and the rest of the manufactures actually aim for with drives like this. NLE, Mid level NAS applicances etc etc.

 

[XBoxLPU]

Originally posted by: NFS4
I might actually give a rat's a$$ if it had a capacity greater than 36GB. Otherwise, I won't give it even a second glance.

Using a 10K SCSI 36gb and IMO that is all I need for my programs and OS. I have UT2003, CC:Generals, BF1942, and AOM. And over 20 + programs installed. Still have 24gb free space

 

[Deuce300]

I'm definitely in the market for a Raptor. I've been planning on overhauling my main rig with a T'Bred/NF7-S combo and want the best HDD performance possible. These new numbers (from what is mentioned above since I can't access storagereview.com right now, I guess its down) have once again reinforced my faith in WD. I currently use 4 WD JB model drives in my server for storage (which have worked flawlessly). This bad boy will work just fine in my main rig since I never use up more than 20-30gb anyway. I was so disappointed in the initial reviews, thank goodness they came through on the final release.

 

[NFS4]

Originally posted by: ReMeDy{WcS}

if it had a capacity greater than 36GB. Otherwise, I

I don't think you're the target WD and the rest of the manufactures actually aim for with drives like this. NLE, Mid level NAS applicances etc etc.

It still makes little to no sense. Serial-ATA is only included on very few motherboards these days, and when they are included it's from 3rd party chipsets that aren't exactly "sponge worthy."

Intel's ICH5 will add native Serial-ATA support, but even then boards with ICH5 won't exactly be mainstream and that's still months away.

This HD is answering questions that no one asked. If you want server performance, get a SCSI drive. If you want excellent performance and large capacity, get a regular 7200RPM drive.

I don't remember anyone asking "I need a Serial-ATA hard drive with performance not as good as SCSI, slightly better than 7200RPM IDE drives, and with far less storage capacity than regular IDE drives"

 

[Sunner]

Originally posted by: ReMeDy{WcS}

if it had a capacity greater than 36GB. Otherwise, I

I don't think you're the target WD and the rest of the manufactures actually aim for with drives like this. NLE, Mid level NAS applicances etc etc.

I never said so, I asked just who their target group is.

 

[mechBgon]

There is also the noise factor. *zzzzzing!*

 

[Deuce300]

I definitely agree that the target group for this drive is VERY limited. It definitely fits the bill for me, but for most other applications it isn't much of an improvement from what has been mentioned already. We all know that the "latest and greatest" isn't always the best option in many cases. We saw that way back from the first 7.2k IDE drives too, remember? This drive has a limited segment right now but will pave the way for greater things to come and offers performance improvements to a few who are looking to get a bit of a boost right now...

 

[kuk]

Here's the head to head comparison between beta and final. I wonder if the noise levels have been reduced, as there is no indication in the table.

 

[glugglug]

If noone made 10K RPM SATA drives, there would be zero performance advantage for SATA vs. PATA, and that would inhibit their ability to force the newly overhyped standard onto the market.

 

[zephyrprime]

To relate to the previous discussion about WinMark 99, I'd be interested in knowing how WM99 got that 183 MB/Sec number in FP98, considdering SATA1 is limited to 150 MB/Sec, and a 32bit 33 MHz PCI bus is limited to 133 MB/Sec...

Winmark is messed up. I saw some other diskmark (which is part of winmark) benchmark the other day for a drive and it said the transfer rate for the beginning and end of the media was the same! It's just too old and is probably being fooled by windows big file cache or something. Maybe the file it makes for testing is actually smaller than the size of windows' disk cache on the machine the test was run on.

 

[Samus]

Originally posted by: xerosleep
I never use more than 5 gigs so it's all good to me

haha, yea, ditto.