Maxtor maxline III slow access time

[Handruin]

I bought a new drive and decided to benchmark it for reference. I decided to try Maxtor's maxline III 300GB SATA drive 7L300S0.

I almost fell out of my chair when I saw 18.3 ms access time...so I figured that can't be correct. Well, I've tried two versions of HD Tach and both give me roughly the same number.

My Abit NF7-S has a Silicon Image 3112 SATA raid built into the board, and it's being used as JBOD. I have both ports occupied, one by my 80GB WD (using PATA->SATA converter), and the other port with this new maxtor (native SATA). The WD scores a 13.4 ms, which is more of what I expected out of the maxtor.

So I swapped ports between the drives and reran the test...same result for both drives.

next, I swapped the cables and reran the tests...again, same result.

next, I verified the driver version is up to date on the SI3112 card (1.0.51). Reran the tests, no change.

I reinstalled the nforce2 driver package...reran the test, no change.

I visited maxtor's website and looked for some type of ARM (acoustic management tool) and couldn't find anything...I'm guessing it doesn't have this feature, but I could be wrong.

Tomorrow I'll be pulling the drive out to retest in two other machines to verify if this is an issue. Does anyone have any thoughts as to what is causing this? I did read (unfortunately today) a few comments of these drives having slow access times on some samples. Those comments were from late last year, early this year, so I figured it wasn't going to be a problem. SR had a good run with their second sample of this drive, but I figured they had a pre-released version in their first run that was giving them issues.

So I may not have made the best choice in drives. I wanted a drive larger than what samsung offers (even though I've been buying them outside of this experience) so I chose this maxtor. May not have been the best choice, but any thoughts are appreciated.

 

[Handruin]

I did a retest on my dell poweredge. I used the maxtor as the sole primary drive and installed XP on it from scratch just to rerun the test.

The access time improved significantly between machines. However, there are significant differences in the tests. For one thing, the Dell has a different SATA controller. Which, I might add, won't recognize the full 300GB of space, where as my SiliconImage did. I also only create a 4GB partition on the new XP install. I don't know if this would have any difference.

Another difference is that the Dell is using the maxtor as its boot drive where as on my system, it was just added to the pool of disks.

Perhaps as a solution I need to break down and buy a modern SATA controller to use this drive in my primary system? I feel a little bit better seeing the numbers on the Dell, but I'm still going to try this on another system.

 

[Tannin]

Why would a controller chipset give you bad access times but not bugger up everything else as well?

Nope, I don't believe it Doug. For my money, it's a faulty drive. (Or the model is always incredibly slow, which I don't believe. I'm not the world's biggest Maxtor fan, but surely they aren't that bad.)

I am not familiar with HD Tach, but the usual arrangement is that a benchmark program tests access time within the partition it is testing. Winbench, for example, works like this. So your quite reasonable sounding 11.9ms access times on the test install sound bogus to me. Hey - you subtract 4.2ms for latency of a 7200 RPM drive and you are left with 7.7ms average seek. Since when did mass market drive have 7.7ms seek performance? If it really was a 7.7ms drive, everyone would be raving about them.

The 14.1ms seek you were getting on the other controller sounds more believable - but quite unacceptably slow.

 

[Tannin]

Sorry: I didn't make it clear. From those figures, it looks to me as though the drive is very slow and on the second set of tests it was only testing access times as measured within that tiny partition.

 

[Handruin]

Thanks Tannin for the feedback. Part of human nature doesn't allow us to hear the bad news sometimes. In my case, I'm trying my darnedest to find a reason other than the drive for this issue.

I wasn't sure how much of the physical drive HD Tach would access, but I was hoping 100%. One thing that left me hoping is the 13.1 ms run on the second set of tests. I can believe that number, but you're right, 11.9 ms is unrealistic for this drive. I'm also putting a lot of faith into HD Tach, and it could be blatantly wrong. What other tools can I try to verify my findings? I can't remember if IOmeter does access time.

Is it possible to update firmware on a drive? I looked through maxtor's site and I didn't find anything. I don't believe I've ever updated a hard drive's firmware...

I'll likely try one more run on a new system, and then I'll have to deal with the truth...this drive is slow. If the testing doesn't go well, how in the hell does one convince maxtor that a working drive is defective based on poor access times?

 

[Mercutio]

Is AAM turned on?

 

[Tannin]

Winbench does access times, Doug. No doubt there are a million others too. I can't remember what I used to use, but Winbench will certainly do it. I think Merc is right on the mney: it sounds very like an AAM problem. But what's the story with Maxtor's tech support? Surely they have a utility?

BTW, if Maxtor publish a seek time figure, you are perfectly entitled to demand that their drive achieve it, or get your money back. Same deal as if I sell you a litre of milk and only put 450ml in the glass. If you advertise it, you have to deliver.

 

[time]

I had a similar problem with a Samsung drive, but put it down to the severe vibration the drive was exhibiting. The replacement was fine.

From my obervations of different drives in about half a dozen computers, I would say HD Tach results are only approximate, but still reasonable. I achieved expected access times for Samsung P80 SATA, WD800JB, IBM T7K250 and IBM 60GXP. Surprisingly, the benchmark seemed to work across the whole drive rather than just one partition, although I suspected that some variations were due to different partitioning. Ideally, I should have tried different partitions on the same drive to see the effect.

 

[time]

Doug, try the Hitachi Feature tool. It works with all drives.

 

[Handruin]

I was concerned about the AAM (mistakenly called it ARM in my first post). I looked for a way to enable/disable, but maxtor didn't have anything that could be easily found to do this. I'll look again... I tried both tools they offer online, but one of them wouldn't run, and the other had no software for AAM control.

time, I will certainly be trying the Hitachi tool when I get home. I would not have expected another manufacturer's tool to work with maxtor. Thanks for the tip!

If AAM isn't enabled, and the new system I'm building offers no increase in performance, I'll send the drive back. I'd preferably like to return it to newegg for a refund, but it doesn't seem like that's possible. Has anyone returned drives to newegg for a refund? Otherwise, I'll RMA it and hope for a decent return.

 

[Tannin]

If it doesn't meet the published specs, your retailer has no choice. They must refund or replace it. This is true everywhere, it's bound to be true even under US consumer law.

 

[Mercutio]

I don't know about that. The US is very much an anti-consumer place.

If you've noticed, recent HP inkjet printers (yes, Tannin, we know) have started coming with extra-dinky cartridges.

Why?

Because California, one of the only decent places in the United States of Crap, made it illegal to sell "starter" cartridges with something less than a full volume of ink in the tank.

HP's response to that was not the intended "Sell everyone a full tank". Nope, they went with "Sell only starter cartridges", instead.

Back to Handy's problem: I suspect that the manufacturer will tell Doug that whatever 3rd party tool he's using is not using the One True Method for measuring access time. Maxtor will no doubt provide a tool (or maybe not. Maybe they'll make him guess what the right tool is) that will no doubt show his drive to perform within Maxtor spec.

Because that's just how things are.

 

[Bozo]

Newegg used to take back anything if it was under 30 days. Don't know if that's still true.

Bozo :mrgrn:

 

[Tannin]

Then on what basis does contract law operate?

 

[Mercutio]

Newegg used to take back anything if it was under 30 days. Don't know if that's still true.

They won't take back laser printers if you've used any of the included toner, even within the 30 day period.

 

[Fushigi]

Then on what basis does contract law operate?

It would seem that under the current administration, when two parties have a dispute the party with the larger Market Capitalization wins by default. This can be overruled based on party affiliation/donations, closeness to the oil/energy industry, and a few other factors, but seems to hold true by and large.

 

[Tea]

There you go Tannin. Ask a stupid question, get entirely too honest an answer.

BTW, do we count market cap in terms of number of dollars or amount of real-world street cred? If the latter, then Doug the Handruin outcaps Maxtor's marketing department by about 40 to 1.

But they probably only count dollars.

sigh

 

[Buck]

Consumers have rights? :eekers: No one told me.

 

[Handruin]

I'm feeling a bit better tonight:

time, I owe you a beer (or drink of choice) some day. That hitachi features tool was perfect! I love how you can customize the AAM in real-time. I can actually listen to the drive during random seeks and select the level of AAM while it's seeking. Since noise wasn't a factor in purchasing the drive, I've completely disabled it.

I'm pissed that this drive shipped with AAM enabled by default, and no software or info to disable it. I reran my basic tests and the access time has improved quite a bit. It's still no rocket, but much closer to what I'd tolerate from this drive.

I retested using a new machine built yesterday. Prior to disabling the AAM, the drive was scoring a 17.5 ms access. After, it is in the mid 14's as seen below.

 

[time]

:beer:

 

[tazwegion]

Please excuse the ignorance... but what does the acronym AAM actually stand for, and what is it's relationship to the drive, and inherently performance?

I didn't locate any answers on an initial Google search

 

[Handruin]

I believe it stands for Automatic Acoustic Management.

 

[tazwegion]

So... it controls the actual speed of the drive to reduce operating dB? :?:

 

[Tannin]

Not the rotational speed, no. (That would be horribly difficult to do without buggering up the data integrity.) But it controls the acceleration of the read-write heads, making them start and stop gently so as to reduce noise.

 

[tazwegion]

Ah, interesting... thanks for the explaination