Hello,
I am about to purchase a new 400 GB hard drive, and in my case low noise levels are a priority. I didn't know about SPCR, and I must say I was extremely glad to discover a resource that provides exactly the information I was looking for. I've read the forum topics, and I really love you guys!
There are some things I did not manage to find though, namely a detailed review of Samsung's HD401LJ, such as this one (about HD400LJ).
And a review of Hitachi's new models, namely the HDT725040VLA360, a 400 GB unit, with a 16 MB buffer and a Serial ATA 3 Gbit/s interface, also known as T7K500.
I am using this article as a reference, because there is no article about the specific model I mentioned.
The choice is clearly pro-Samsung, due to the smaller number of platters. But if you take a look at the new Hitachi's specs, you'll see that the new model uses three platters, rather than five; while the declared noise level when the unit is idle is 3 Bells. With this in mind, Hitachi has the potential to beat Samsung's "older" model.
In my area, the price-delta between these two models is close to zero, so price is not a factor. Has anyone got first hand experience with these units?
I also have other questions. I plan to use the new HDD on my desktop, that runs Linux. Does Hitachi's low power consumption technology rely on any software? Or is it a purely hardware-based solution? What I'm trying to find out is whether I'll have any adventures with trying to set it up in Linux.
My second question is about interfaces. My motherboard is an Asus A7N8X, it does not support SATA 3.0 Gbit/s, so naturally I am expecting that the HDD will not be a hero (performance wise).
How does each of these drives handle backwards compatibility? Are there known issues? Finally, assuming that one drive is more performant than the other (does not matter which), will they both behave in the same way when working with SATA 1.5 Gbps?
Samsung HD401LJ vs Hitachi T7K500, 400 GB
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Hi,
You may be interested by the newer HD501LJ. Also I've made some noise recordings, comparing it to the HD401LJ and the older SP2504C.
You may be interested by the newer HD501LJ. Also I've made some noise recordings, comparing it to the HD401LJ and the older SP2504C.
"There are some things I did not manage to find though, namely a detailed review of Samsung's HD401LJ, such as this one (about HD400LJ)."
Why should there be reviews on both 8MB and 16MB cache variants of the same model & same capacity point? Especially when we are talking about reviews that are limited to noise issues and doesn't even try to evaluate their performance, what difference does different cache sized cause in noise output? In real life, they may in some use environment mean less seeks (equally noisy, but less frequent) but with synthetic benchmarks to create random seeks (where cache size does not matter) I don't think different cache sizes would produce different results.
It's safe to assume HD401LJ noisewise equal to HD400LJ, or possibly slightly better, due to reduced need for seeking. This reduction in number of seeks depends greatly on the software YOU're running and couldn't be reviewed accurately by SPCR.
"And a review of Hitachi's new models, namely the HDT725040VLA360, a 400 GB unit, with a 16 MB buffer and a Serial ATA 3 Gbit/s interface, also known as T7K500.
I am using this (link to 7K400 article) as a reference, because there is no article about the specific model I mentioned."
7K250 (Vancouver III) would be a better reference as T7K500 is a Vancouver model of fifth generation. 7K400 was Kurofune model of first generation. T7K250 didn't have 3-platter variant so 7K250 250GB is actually the closest reference drive to T7K500... and 7K250 sure leaves room for improvement (as an owner of one, I know). Still, two full model generations of evolution, that's a long time in IT business. I have heard there has been some improvement but I don't know if there has been enough of it.
Vancouver LP (later called "Pathfinders") (one-platter variants): 120GXP(?), 180GXP(?), 7K250, 7K80, 7K160
Vancouvers (2 to 3 platters): 120GXP, 180GXP, 7K250, T7K250 (not available with 3 platters), T7K500
Kurofunes (originally 5 platters, with 7K1000 750GB also 4 platter variants): 7K400, 7K500, 7K1000
Pathfinders are identified by "P", Vancouver III with "V", Vancouver IV and V with "D", Kurofunes with "K" in their model names (for example: HDS724040KLAT80 of 7K400 PATA variant).
Pathfinders: those variants with ultra-low profile base cast, but who are not ultra-low profile due to bulging top cover. These drives have "WD" style spider web grooves on the bottom side of the cast. No shiny plates here.
Vancouvers: normal "low-profile" base cast. Last generation (T7K500) lost it's shiny metal plate on the bottom, smooth black surface. Older Vancouvers have it (single layered).
Kurofunes: "low-profile" base cast. Spindle motor assembly supported with a screw on top cover (something familiar with ball-bearing drives, but this drive is with fluid bearings). Three-layered shiny metal plate.
What are those plates by the way? There's no reason to add them for thermal dissipation (as base cast is metal too and dissipates heat effectively). Are they present to eliminate some resonant frequencies? Or EMI shielding?
"Has anyone got first hand experience with these units?"
I only have 7K250 and 7K400 so I do not have first hand experience on newer Hitachis. But as APM (Advanced Power Management) feature has probably remained the same, I may give advice how to enable it.
"Does Hitachi's low power consumption technology rely on any software? Or is it a purely hardware-based solution? What I'm trying to find out is whether I'll have any adventures with trying to set it up in Linux."
Purely hardware based. Compeltely OS independent. Hitachi offers a bootable diskette utility to configure it (along with AAM (Automatic Acoustic Management) feature to slow down the seeks).
"My second question is about interfaces. My motherboard is an Asus A7N8X, it does not support SATA 3.0 Gbit/s, so naturally I am expecting that the HDD will not be a hero (performance wise)."
No HDD as of today can saturate 1.5 Gbit/s SATA interface so it doesn't matter what frequency the interface runs at.
"assuming that one drive is more performant than the other (does not matter which), will they both behave in the same way when working with SATA 1.5 Gbps?"
No. Interface limits transfer rates but even if interface was the bottleneck, it wouldn't affect seek speeds and hardware level caching.
"How does each of these drives handle backwards compatibility? Are there known issues?"
Yes, there's some issues of old 1.5Gbit/s controllers that does not support interface speed negotiation and thus are not capable of telling the HDD to use lower interface frequency. Newer 1.5Gbit/s controllers support that negotiation, and 3.0 Gbit/s controllers don't even need it because they can support 3.0 Gbit/s HDD at their maximum interface transfer rate.
There are differences on how HDDs handle forced compatibility to circumvent the lack of negotiation support on controller. The most common one is jumper configuration. This allows HDD to be shipped in 3.0 Gbit/s mode and configured to 1.5 Gbit/s as necessary.
Hitachi uses software configuration. First they shipped the HDDs in 1.5 Gbit/s mode (7K250 and T7K250)... but even though 1.5 Gbit/s doesn't really get saturated, OEMs probably wanted the drives to be shipped in 3.0 Gbit/s mode because OEMs are such a lazy bastards (they don't have the time to configure HDDs with bootable floppies) and OEMs want to advertize features that have zero real value (such as 3.0 Gbit/s interface or "SATA2"). So, they obeyed customers' (=OEM) wishes and shipped them in 3.0 Gbit/s mode, starting from T7K500.
And the result: Hitachi SATA drives need to be configured to lower transfer rate before they work on old controllers. And to configure it, it has to be first recognized by BIOS! So you need another computer (and that one needs to be a new one) to first configure the HDD in compatibility mode with the older one. Pain in the arse. (Also remember that while addon PCI card that support 3.0 Gbit/s may allow you to use the drive in older system, those controllers doesn't necessarily provide low-level access to ATA features of the HDD, thus you can use the drive with addon controller but you can't:
- configure it to compatibility mode (and later move it to motherboard controller) or
- take advantage of APM and/or AAM.
My suggestion, find out what controller your motherboard has (i.e make and model of your southbridge). Then find out about the incompatible controllers from WD's FAQ or some other source and see if there's a match. If interface transfer speed negotiation is not supported by the controller, and you don't happen to have a friend who has a new computer and of course if the retailer has good customer service, you could as them to configure it to 1.5 Gbit/s at the time of purhase. Otherwise forget Hitachi SATA drives.
Why should there be reviews on both 8MB and 16MB cache variants of the same model & same capacity point? Especially when we are talking about reviews that are limited to noise issues and doesn't even try to evaluate their performance, what difference does different cache sized cause in noise output? In real life, they may in some use environment mean less seeks (equally noisy, but less frequent) but with synthetic benchmarks to create random seeks (where cache size does not matter) I don't think different cache sizes would produce different results.
It's safe to assume HD401LJ noisewise equal to HD400LJ, or possibly slightly better, due to reduced need for seeking. This reduction in number of seeks depends greatly on the software YOU're running and couldn't be reviewed accurately by SPCR.
"And a review of Hitachi's new models, namely the HDT725040VLA360, a 400 GB unit, with a 16 MB buffer and a Serial ATA 3 Gbit/s interface, also known as T7K500.
I am using this (link to 7K400 article) as a reference, because there is no article about the specific model I mentioned."
7K250 (Vancouver III) would be a better reference as T7K500 is a Vancouver model of fifth generation. 7K400 was Kurofune model of first generation. T7K250 didn't have 3-platter variant so 7K250 250GB is actually the closest reference drive to T7K500... and 7K250 sure leaves room for improvement (as an owner of one, I know). Still, two full model generations of evolution, that's a long time in IT business. I have heard there has been some improvement but I don't know if there has been enough of it.
Vancouver LP (later called "Pathfinders") (one-platter variants): 120GXP(?), 180GXP(?), 7K250, 7K80, 7K160
Vancouvers (2 to 3 platters): 120GXP, 180GXP, 7K250, T7K250 (not available with 3 platters), T7K500
Kurofunes (originally 5 platters, with 7K1000 750GB also 4 platter variants): 7K400, 7K500, 7K1000
Pathfinders are identified by "P", Vancouver III with "V", Vancouver IV and V with "D", Kurofunes with "K" in their model names (for example: HDS724040KLAT80 of 7K400 PATA variant).
Pathfinders: those variants with ultra-low profile base cast, but who are not ultra-low profile due to bulging top cover. These drives have "WD" style spider web grooves on the bottom side of the cast. No shiny plates here.
Vancouvers: normal "low-profile" base cast. Last generation (T7K500) lost it's shiny metal plate on the bottom, smooth black surface. Older Vancouvers have it (single layered).
Kurofunes: "low-profile" base cast. Spindle motor assembly supported with a screw on top cover (something familiar with ball-bearing drives, but this drive is with fluid bearings). Three-layered shiny metal plate.
What are those plates by the way? There's no reason to add them for thermal dissipation (as base cast is metal too and dissipates heat effectively). Are they present to eliminate some resonant frequencies? Or EMI shielding?
"Has anyone got first hand experience with these units?"
I only have 7K250 and 7K400 so I do not have first hand experience on newer Hitachis. But as APM (Advanced Power Management) feature has probably remained the same, I may give advice how to enable it.
"Does Hitachi's low power consumption technology rely on any software? Or is it a purely hardware-based solution? What I'm trying to find out is whether I'll have any adventures with trying to set it up in Linux."
Purely hardware based. Compeltely OS independent. Hitachi offers a bootable diskette utility to configure it (along with AAM (Automatic Acoustic Management) feature to slow down the seeks).
"My second question is about interfaces. My motherboard is an Asus A7N8X, it does not support SATA 3.0 Gbit/s, so naturally I am expecting that the HDD will not be a hero (performance wise)."
No HDD as of today can saturate 1.5 Gbit/s SATA interface so it doesn't matter what frequency the interface runs at.
"assuming that one drive is more performant than the other (does not matter which), will they both behave in the same way when working with SATA 1.5 Gbps?"
No. Interface limits transfer rates but even if interface was the bottleneck, it wouldn't affect seek speeds and hardware level caching.
"How does each of these drives handle backwards compatibility? Are there known issues?"
Yes, there's some issues of old 1.5Gbit/s controllers that does not support interface speed negotiation and thus are not capable of telling the HDD to use lower interface frequency. Newer 1.5Gbit/s controllers support that negotiation, and 3.0 Gbit/s controllers don't even need it because they can support 3.0 Gbit/s HDD at their maximum interface transfer rate.
There are differences on how HDDs handle forced compatibility to circumvent the lack of negotiation support on controller. The most common one is jumper configuration. This allows HDD to be shipped in 3.0 Gbit/s mode and configured to 1.5 Gbit/s as necessary.
Hitachi uses software configuration. First they shipped the HDDs in 1.5 Gbit/s mode (7K250 and T7K250)... but even though 1.5 Gbit/s doesn't really get saturated, OEMs probably wanted the drives to be shipped in 3.0 Gbit/s mode because OEMs are such a lazy bastards (they don't have the time to configure HDDs with bootable floppies) and OEMs want to advertize features that have zero real value (such as 3.0 Gbit/s interface or "SATA2"). So, they obeyed customers' (=OEM) wishes and shipped them in 3.0 Gbit/s mode, starting from T7K500.
And the result: Hitachi SATA drives need to be configured to lower transfer rate before they work on old controllers. And to configure it, it has to be first recognized by BIOS! So you need another computer (and that one needs to be a new one) to first configure the HDD in compatibility mode with the older one. Pain in the arse. (Also remember that while addon PCI card that support 3.0 Gbit/s may allow you to use the drive in older system, those controllers doesn't necessarily provide low-level access to ATA features of the HDD, thus you can use the drive with addon controller but you can't:
- configure it to compatibility mode (and later move it to motherboard controller) or
- take advantage of APM and/or AAM.
My suggestion, find out what controller your motherboard has (i.e make and model of your southbridge). Then find out about the incompatible controllers from WD's FAQ or some other source and see if there's a match. If interface transfer speed negotiation is not supported by the controller, and you don't happen to have a friend who has a new computer and of course if the retailer has good customer service, you could as them to configure it to 1.5 Gbit/s at the time of purhase. Otherwise forget Hitachi SATA drives.
Alfred and whiic, thank you for the feedback.
Of course, it is possible that the difference is there because of imprecise measurement, or it only applies to the particular samples that were verified, while other products of the same line can behave differently.
Your points are very convincing, and I admire your way of getting a message across. I am pretty much inclined towards Samsung's HD401LJ, the only 'kernel of doubt' is
In theory this should not be a big deal, but it is possible that they've changed other components besides the cache (without making the changes public). If you take a look at the graphs posted by Alfred, you'll see that there is a visible difference between the noise levels; and the audio recordings confirm that. So the difference is not only visible, but also audible.Why should there be reviews on both 8MB and 16MB cache variants of the same model & same capacity point?
Of course, it is possible that the difference is there because of imprecise measurement, or it only applies to the particular samples that were verified, while other products of the same line can behave differently.
Your points are very convincing, and I admire your way of getting a message across. I am pretty much inclined towards Samsung's HD401LJ, the only 'kernel of doubt' is
That's exactly what I had in mind, but I could not find such an elegant way to express that idea.Still, two full model generations of evolution, that's a long time in IT business.
"In theory this should not be a big deal, but it is possible that they've changed other components besides the cache (without making the changes public)."
Everything is "possible" but I don't think it's likely. Sure, you can take a HD401LJ and compare it against HD400LJ and you may even find slight (or even obvious) differences between them. But that doesn't decessarily mean it's a difference between different cache variants... it may just be that newly reviewed HD401LJ could be of later manufacture than the HD400LJ, and thus be different revision of the model. Reviewing a newer sample of HD400LJ against older HD400LJ could produce the same result.
Because of this, I would not see HD401LJ review necessary and even if such a review was done, I would probably consider any differences to be caused by different manufacturing dates, revisions, batches, etc. (i.e pretty much anything but not the size of cache).
Everything is "possible" but I don't think it's likely. Sure, you can take a HD401LJ and compare it against HD400LJ and you may even find slight (or even obvious) differences between them. But that doesn't decessarily mean it's a difference between different cache variants... it may just be that newly reviewed HD401LJ could be of later manufacture than the HD400LJ, and thus be different revision of the model. Reviewing a newer sample of HD400LJ against older HD400LJ could produce the same result.
Because of this, I would not see HD401LJ review necessary and even if such a review was done, I would probably consider any differences to be caused by different manufacturing dates, revisions, batches, etc. (i.e pretty much anything but not the size of cache).