Are WD JB drives 60GB/platter yet?

[time]

Does anyone know?

I seem to recall LiamC being unhappy with the 40GB version as well?

 

[Buck]

If you have a WD1800JB or WD2000JB, then it is from a 60 GB per platter family (66 GB for the 2000).

 

[e_dawg]

Apparently, there is a 50-50 chance of getting a 1200JB with 60 GB platters.

http://forums.storagereview.net/viewtopic.php?p=91445#91445

 

[LiamC]

According to WD's site, 600JB's are still 40GB/platter.

I was unhappy that WD were advertising them as 60GB/platter. They
would not return emails and eventually changed the specs on their site to reflect a 40GB/platter model.

After talking to people directly in the U.S. WD apologised and replaced my 600JB with a 60GB/platter 1800JB.



Now all I have to do is figure out what to do with the space, I couldn't fill a 60G drive...

 

[e_dawg]

You could always e-mail me a few gigs

 

[Handruin]

I read through a bunch of that and I still don't know how to tell if my new WD1200JB is 60 or 40 GB per platter.

 

[e_dawg]

All you need to do is run HDTach or ZDBench. Once you know the beginning/ending STR values, you can say unequivocally whether it is a 40 or 60 GB/platter model.

 

[Handruin]

I can't use HDTach because it cost money. The free version won't run under XP. I don't have ZDBench, can I use atto?

 

[e_dawg]

You can use Atto as well, although HDTach IMO is the best utility for a quick check of the low level performance metrics of drives (remeber to use Advanced size check).

 

[Handruin]

What transfer range should I be looking for in HDTach for the 60GB per platter? I'll post a screen shot in a sec.

 

[e_dawg]

29-49 MB/s for the 40 GB/platter
32-54 MB/s for the 60 GB/platter

 

[Handruin]

OK here is what I got:

1st run after reboot (F@H was running oops)

2nd run no additional reboot (F@H stopped)

While I was at it I ran the rest of my drives for comparison:

18GB atlas 10KIII
http://www.storageforum.net/img/content/hdtach3-atlas10KIII.jpg
http://www.storageforum.net/img/content/hdtach4-atlas10KIII.jpg

9GB atlas 10KII
http://www.storageforum.net/img/content/hdtach5-atlas10KII.jpg

Seagate Barracuda 18XL
http://www.storageforum.net/img/content/hdtach6-cuda.jpg

WD Expert 18GB
http://www.storageforum.net/img/content/hdtach7-expert.jpg

Surprisingly the drive with the worst access is the WD1200JB

 

[Handruin]

FWIW, I purchased this drive a couple of weeks ago. I'll have to look up the date of manufacture.

 

[e_dawg]

What you have there is the 40 GB/platter version. Here is my 40 GB/platter WD600BB:

 

[Handruin]

Oh well. :-? I like the drive regardless.

Any thoughts as to why my CPU utilization is so high for EIDE? My SCSI utilization is half the amount of my EIDE.

 

[e_dawg]

Doug, a tip for you if you don't mind... save your screenshots as GIF or PNG instead of JPEG, for they will be clearer and the resultant files will be smaller.

 

[e_dawg]

Any thoughts as to why my CPU utilization is so high for EIDE? My SCSI utilization is half the amount of my EIDE.

SCSI CPU utilization should be half of IDE. That's why people pay more for SCSI Well, not really, but it is one advantage.

What are your system specs like? Are you running an antiviral? IDE drivers?

 

[Handruin]

At the time of the test I had shut down all normal running task such as firewall, trillian, rivatuner, F@H (after first run I shut it down) and it was after a reboot.

My specs are:

WinXP Pro SP1
MSI K7T Pro2a (VIA KT133 chipset / VT82C686B ATA100)
Althlon T-bird 1200 MHz @ 1200
Crucial 768MB PC 133 CL2
Onboard EIDE ATA100 (no added drivers, default Windows XP pro SP1. Probably my issue?)
GF3 Ti 200
Tekram DC-390U3W U160 SCSI (running XP Default Driver)
SB Live
linksys 10/100

 

[blakerwry]

I think it's just the combination of controller you're using and your CPU. You could try running in ata66 mode (I've seen benches that show it uses less CPU time while providing the same benchmark speeds)


fot what it's worth, here are read/write benches of both an 800JB and a 1200BB

800JB:

1200BB:

As you can see, they are both 40GB/platter.

 

[blakerwry]

Those were both done on a promise FastTrak100/Ultra100 with a palomino 1700+ and Soyo Dragon+ (kt266a/8233).

 

[Fushigi]

If you have a WD1800JB or WD2000JB, then it is from a 60 GB per platter family (66 GB for the 2000).

Apparently, the 160GB SE also uses 60GB platters: http://www.storageforum.net/forum/viewtopic.php?t=1603

- Fushigi

 

[e_dawg]

MSI K7T Pro2a (VIA KT133 chipset / VT82C686B ATA100)
Onboard EIDE ATA100 (no added drivers, default Windows XP pro SP1. Probably my issue?)

Those were both done on a promise FastTrak100/Ultra100 with a palomino 1700+ and Soyo Dragon+ (kt266a/8233).

Bwahahaha.... VIA strikes again... My HDTach test was done on a Palomino 1800 and an nForce board. Is it really surprising that I have lower CPU utilization?

 

[blakerwry]

MSI K7T Pro2a (VIA KT133 chipset / VT82C686B ATA100)
Onboard EIDE ATA100 (no added drivers, default Windows XP pro SP1. Probably my issue?)

Those were both done on a promise FastTrak100/Ultra100 with a palomino 1700+ and Soyo Dragon+ (kt266a/8233).

Bwahahaha.... VIA strikes again... My HDTach test was done on a Palomino 1800 and an nForce board. Is it really surprising that I have lower CPU utilization?

you're so cruel

 

[Handruin]

I never noticed this problem because my machine has been mainly SCSI. This WD1200JB is the first IDE drive I've purchased in almost 4 years.

 

[e_dawg]

If you've never noticed this problem before then it really isn't a problem, right? Don't let these benchmark figures bias you into thinking your system is slower than it really is. If you continue to use you system without noticing any CPU utilization problems, then don't worry about it.

 

[Handruin]

I never noticed because I rarely used ATA storage. Now that I have a nicer hard drive, I use it more. Truth is, I don't notice the CPU utilization. I was concerned that I should be upgrading my drivers or something to fix this, but if there isn't a fix...so be it, I'll just upgrade my machine at some point in time.

 

[blakerwry]

Handruin, I thought maybe it was just the 686b southbridge, but I tried my 1200BB with my 8233 southbridge that came with a kt266a board... I get the same results as you... 57-58% CPU utilization. I tried both microsoft's and VIA's drivers with no difference. I confirmed this with a 1GB file copy from the 1200BB to another folder on the 1200BB.

My CPU is just 66mHz faster than yours I believe (palomino 1700+ runs at 1466 and 2/3rds mHz)...


Of course these benchmarks are not indicitive of the actual real life performance between the different controllers. Although I could guess that while doing both CPU intensive and disk intensive activities the NForce 2 people would have a faster/smoother experience then us VIA folk.

 

[blakerwry]

Just out of curiosity I wanted to bench my HDD in UDMA66 mode, but to my surprise, neither WD's or IBM's tool would do the trick to knock the drive down to UDMA mode 4 or 3... not even UDMA mode 2 (although I could do this through cabling)

I tried 2 mobos and 3 controllers without success. (440bx, ultra100, and via 8233)

Perhaps I can compare my fastTrak66 with my fastTrak 100... but the different BIOSes and different drivers along with the different electronics in the controllers themselves would mean that the data is not really comparable and thus I would not be able to tell if there is any more CPU usage by going from ata100->ata66 or vise versa.

 

[Buck]

Blake, what version of Data Lifeguard Tools are you using and was this tried through the EIDE 0 or 1 connectors on the motherboard?

 

[Buck]

Just as a side note, WD does have 80 GB per platter JB drives out. Presently they can be found as either a WD400JB or WD800JB.

 

[blakerwry]

Blake, what version of Data Lifeguard Tools are you using and was this tried through the EIDE 0 or 1 connectors on the motherboard?

This was the UDMA utility stripped out of the datalifeguard tools... i dont know what version

the data lifeguard tools have never detected my drive when it is set to slave... even so, tonight, I tried and as usual it failed to even detect the drive when it was slave and my DVD was master...

I tried:

Via 8233
secondary slave : not detected
Primary Master  : Unsuccessful


Promise Ultra100
Primary Master  : Unsuccessful

440BX
Primary Master  : Unsuccessful

The drive was detected and was shown to be compatible with UDMA mode 2, 4, and 5 (labelled 33, 66, 100) and was shown to be operating at 100... everytime I tried to set it to a different mode i would get an error saying that my attempt was unsuccessful.

Incidentally i tried IBM's(actually hitachi's) drive feature tool and it did not even act like it would support changing the UDMA mode on the WD drive.

 

[Jan Kivar]

To the original topic: I started a thread at SR about the issue. Seems that the best way to be certain is to either use WinBench99 and check the transfer rates, or to check the product/model code. If it's "WD1200JB-00DUT0" (this relates only to 1200JB BTW, but that's really the problematic one...), it has 60 GB platters. WD confirmed it to me.

Quite frankly I'd give small amount of STR to get faster seek times. Compare 1200JB's & 2000JB's "read service times" at SR's database and You'll know what I mean. (OK, maybe there is other variables than platter capacity that inflicts the difference...)

Cheers,

Jan

 

[Buck]

Sorry if this sounds elementary Blake, but I must suggest: Connect the drive to the Primary Master connector on the 440BX as you've already tried. Make sure it is the only device on the data cable, and make sure that there are no jumper shunts being used to jumper the drive. Boot to the Data Lifeguard Tools version 10 diskette or a bootable DOS diskette with the extracted dlgudma file, and run the software. Do not run the software from within Windows.

Please don't attempt to use the IBM tool for this purpose on the WD drive. IBM tools are for IBM drives. WD tools are for WD drives.

 

[blakerwry]

good call buck, i simply downloaded a new version of the Data lifeguard tools and it worked.. didn't even need to move the drive.

 

[blakerwry]

hmm.. didnt have any effect on CPU usage.. all it managed to do was lower the burst transfer rate to ~56MB/sec

 

[Buck]

I'm glad it finally worked Blake.