Well, the new submarine from Seagate, U6, has finally reached our port. This hard disk drive is especially interesting since it features the platters with the highest capacity available today: 40GB (of course, we mean not sample platters, but those used in real mass products). As far as we remember, only two companies announced their HDD models with 40GB platters when the review was in progress. They were Maxtor and Seagate. However, Maxtor drives from 541DX family are thin single-platter HDDs with only one read/write head. Their size is also not that impressive: 10GB, 15GB and 20GB. As for Seagate U6 HDD, it is an absolutely normal, fully fledged storage solution with more than one head on its shoulders.

In fact, there is a very interesting story connected with our Seagate U6 review. About a month ago we managed to get hold of one U6 sample. We ran all the tests and, to tell the truth, obtained very curious results. We were about to offer you a sensational review, but it suddenly turned out that the sample we got was intended for the Set top box and hence had all the corresponding firmware. Of course, we, as well as Seagate guys, were not very enthusiastic about posting a review of the U6 HDD, which didn't meet the PC requirements. So, all we could do was just to put aside an almost ready article until the real U6 drive arrived in Moscow. So, here we are: the true PC U6 HDD is here and you are now reading our review of it. In fact, we can assure you that the performance of the PC and "non-PC" Seagate U6 HDD differs a lot. The non-PC version worked faster in IOMeter, but considerably slower in all other benchmarks. As for the visual differences, there was only one: a line in the Configuration code. On the non-PC HDD you could see "U8G-08", while on the normal U6 - "UXA-01".

  

If we take a quick look at the photos, we will see... Seagate U5 or something looking very much like it. It's true, the new submarine and the U5 are as like as two peas. No wonder since they were made on one and the same building way… The same shape, the same SeaShield. An unsophisticated user may easily mix them up.

On the other hand, however, U6 HDD we had at hand turned out really hard to mix up with anything, since its storage capacity equal to 80GB makes you remember its face really well.

Specification

* - we managed to find a 30GB U6 model on Seagate's web-site. As we learned from Alastair Stewart, Seagate Distribution and sales Manager for Central and Eastern Europe, this HDD is produced only for one single OEM-partner and will never appear in retail stores.
** - "electronically tested" :)

And if you wish to learn more details about the new Seagate U6 HDD, please check this pdf-file.

Well, what can we see from the specs list? First of all, Seagate has finally provided its drives featuring 5,400rpm spindle rotation speed with 2MB cache. Secondly, the rotation speed remained unchanged. And thirdly, all other specs of the new U6 are just the same as those of the predecessor, namely U5 (see the review here).

Testbed and Methods

We tested the newcomer in the following test system:

• Intel Pentium III (Coppermine) 600MHz CPU;
• ASUS CUBX-E mainboard, bios 1007A;
• 2x128MB PC133 SDRAM by Hyundai;
• Matrox Millennium 4MB graphics card;
• Windows 98/Windows 2000 Pro.

In order to check the HDDs performance in different UDMA modes, we had to use the following controllers:

• UDMA33: the controller integrated into i440BX chipset;
• UDMA66: Promise Ultra66 controller;
• UDMA100: Promise Ultra100 controller.

For the Promise controllers we used the drivers ver. 1.60 (build 33).

The disk drives were connected as Master-units to a separate IDE-channel. DMA support in Windows was enabled. We used FAT32 and NTFS file systems to format each of them as one logical drive of the maximum size with the default cluster. All the tests were run 4 times and then the average results were taken for the diagrams. The HDDs didn't rest for cooling down between the tests.

Here are the benchmarks used:

• Windows 98 WinBench 99 1.2
• Adaptec Threadmark 2.0
• Windows 2000 WinBench 99 1.2
• HDTach 2.61
• IOMeter 1999.10.20

We will compare the performance of the new Seagate U6 hard disk drive with that of the pieces featuring 5,400rpm rotation speed and 20GB platter we have already tested before and a new Western Digital WD400AB drive.

Performance

Access Time

Of course, the fastness of heads positioning is not a very important parameter for the HDDs of this type. Not so long ago we had a perfect chance to see that even the drives with the activated Automatic Acoustic Management or AAM (i.e. those switched to a special mode when the heads are positioned slower in order to reduce the noise level) didn't perform much slower in WinBench99 than in case AAM was disabled. As you remember these tests emulate the work with Windows applications that is why we base all our conclusions about the HDD performance in regular Windows conditions on this benchmark set. But, Average Access Time is of great importance for IOMeter and some other tests, so let's start with it.

Well, in fact we have expected to see something like that, because 40GB couldn't fit onto a single platter without any special effort applied. Of course, Seagate had to increase the tracks density and as a result the heads now need more time to be positioned correctly. On the other hand, larger average seek time implies that the hard disk drive we'll be as noiseless as a dew-fall.

Now let's take a look at the data density.

Linear Read Speed

As we have already said, linear read speed lets us evaluate the data density per platter.

At present Seagate is the No 1 here in the IDE field (with its Barracuda ATA III) and SCSI field (with its Cheetah 73LP and Cheetah X15 36LP). When we tested U5 it also showed very high read speed from the platter: 31.6MB/sec. We wonder how the things with per track data density stand by U6 with the 40GB platters.

Hm... U6 falls not only behind its predecessor, but also behind almost all the other testing participants from the previous generation. What's happening? Actually, we discovered a really interesting thing: the platters owe their gigantic size only to the bigger number of cylinders. As you remember, U5 featured 31.100 tracks per inch. By U6 this number has almost doubled up to 58,000!

So, why don't the manufacturers increase the longitudinal data density? The answer to this question is pretty simple: because of the price issue. To reduce the pit size (pit is an information unit on the platter), the read/write heads as well as all the other electronics need to be speeded up. And as you understand, there is nothing for free in this world… The cost of every gigabyte matters a real lot for a hard disk drive with the medium pricing that is why any modifications leading to the cost growth make it no longer that attractive from the pricing point of view. The second important parameter is reliability and the third one appears the rotation speed, which is equal to 5,400rpm in our today's case. To tell the truth, we believe that the today's HDDs perform fast enough to meet the needs of most users. :)

Well, having made a brief introduction we suggest passing over to the benchmarks results obtained for U6 HDD.

As usual, we will start with HDTach 2.61:

HDTach 2.61

As we have already seen above, U6 platter doesn't boast high data density per track that is why this drive didn't win in the read tests. The first prize is given to U5, while U6 rests on a remarkable third position.

Now let's check how the things with writing stand:

Well, here U6 proved the worst of all. The leadership belongs to WD400AB.

HDTach also allows measuring the burst read rate, i.e. we can compare the performance of the HDD DSP as well as the optimization for different UDMA protocols:

As you may notice, Seagate and Fujitsu drives can't boast much and yield to the competitors in all UDMA protocols. Of course, read rate from the cache doesn't determine the HDD fastness, since these values are the maximal rates obtained by checking the read speed for the data packs of the requested size. In real applications the HDD works with packs of different sizes that is why the effective cache use will depend mostly on the firmware algorithms efficiency rather than on the DSP frequency. Anyway, we found this synthetic benchmarks quite interesting and informative as well.

WinBench99 for Windows98

Well, here our hero occupies a very stable third position and shows much better results than U5 HDD.

Of course, the increased cache-buffer told positively on the outcome here (it is exactly what we kept saying all the time). Only WD400AB (see the review here) and Maxtor VL40 managed to leave U6 a bit behind.

And Here U6 suddenly rushed forward having surpassed Maxtor LV40 a little bit. Frankly speaking, we were at a loss, as Seagate budget solutions have very rarely pleased us with fast speeds. As we know now, firmware algorithms play the most important role in achieving high results in WinBench. And we have to admit that Seagate engineers and software developers have done a great job there.

Adaptec Threadmark 2.0

Here U6 proved again quite fast though it failed to win the lead this time. Therefore, the first position belongs to its competitor from Western Digital, also belonging to the newest HDD generation. We would like to point out that Seagate software guys seem to have overcome their dislike of the UDMA33 protocol. At least U6 appears quite successful here.

WinBench99 for Win2000 (FAT32)

If you remember, WD hard disk drives turned out much faster in Win2000. We wonder if it is true for Seagate U6 as well. Let's check it out:

	Linear Read Graphs
	UDMA33	UDMA66	UDMA100
Seagate U6	Graph	Graph	Graph

The results are just brilliant. Seagate U6 is not ahead of all, but even the second prize is quite a valuable thing here. Compared with U5, the new drive boasts much higher performance, which is a really pleasing thing.

In High-End benchmarks U6 also looks very attractive. Just look how much faster it is than the older U5!

WinBench99 for Win2000 (NTFS)

The change of the file system didn't get U6 off the rails: it shows the same second result.

And here U6 seems to have relaxed and rolled back to the third position. Anyway, we wouldn't regard this performance difference as too grave, since it actually lies within the acceptable measuring error. So, we could probably call it the second winner of the second prize :)

Intel IOMeter

Well, now we will try to figure out how the large platter like that of U6 may tell on the IOMeter results. We suspect that the influence will be fatal for the drive, but let's start from the very beginning:

So, the numbers look very much like those of the IBM DTLA 305020 in the silent mode... let's check different patterns before drawing any final conclusions:

It looks as if Seagate U6 felt best of all in WorkStation pattern and worst of all in DataBase one. The difference between WorkStation and FileServer patterns lies with the size of the data block, which is larger in case of WorkStation pattern. But the most important thing is that the data block size is constant, which reduces the firmware workload. As for DataBase, it has more write operations (33%), so that the requests need more time to get processed.

We would also like to point out that the performance grows really insignificantly as the workload increases, which is very likely to signify that the HDD firmware isn't optimized for this type of workload.

Well, and now we have just to compare the performance of U6 HDD with that shown by other testing participants. We will compare the Sum Total value, which is calculated taking into consideration the results obtained with Light, Moderate and Heavy loads (16, 64 and 256 requests):

Yeah, good performance in IOMeter requires very few cylinders and higher spindle rotation speed. Since Seagate U6 can't boast anything of those, it performed quite poorly in this benchmark.

Heat and Noise

We measured the heating with the external thermal diode connected to the mainboard. IOMeter takes over 2.5 hours to run just once and the HDD works quite tensely all this time that's why we may regard it as normal thermal conditions for the drive. We put the HDDs into a PC case, without any extra coolers installed. The room temperature was around 22 degrees Centigrade. As soon as the test came to an end we restarted the system and took the temperature from the system BIOS (1007A).

Seagate U6 HDD heated up to 35^oC, which is a pretty good result for a dual-platter drive with 5,400rpm spindle rotation speed.

As far as the noise level goes, we didn't notice any differences between U5 and U6 hard disk drives.

Conclusion

Well, this is how our idea of the average HDD storage capacity changes. About a year ago, 20GB seemed to me the most optimal capacity, and today we are holding in our hands an ordinary, relatively cheap IDE hard drive four times as big.

This HDD didn't set any records, however, it nevertheless managed to outperform the previous generation drives with 20GB platters almost everywhere except IOMeter tests. What does it mean then? It means that those Seagate's competitors, which can't boast larger platters so far, will have to think hard what to do next. In the first place this problem will arise for IBM, since Samsung and Western Digital have already introduced 30Gb platters, and Maxtor is starting to put into life its new 40GB platters. IBM has already made up its mind to discontinue the DTLA 3050xx family by the end of the year. As for Fujitsu, it looks as if we will not see any new desktop HDD solutions from then any more. As we have told in our news, the company decided to leave the desktop HDD business and to focus on the mobile and server drives.

But enough for the future plans, let's return to today, and namely to our today's hero - Seagate U6 drive.

Judging by the results obtained in the benchmarks we have carried out for it, we can say that Seagate managed to optimize the drive firmware very well for work with Windows applications. Unfortunately, it resulted into dramatic performance drop in IOMeter tests. But frankly speaking, there is hardly anyone "extraordinary" enough to use U6 for FileServer system that is why we don't think that this optimization can be regarded as too harmful. Anyway, the drive has been never designed to break any performance records. Its advantages are simple construction, low cost per gigabyte and high reliability. Actually, these parameters are exactly the ones to determine the drive's popularity. Seagate U5 HDDs turned out the most well sold HDD and now we've got the impression that U6 has every chance to prove even more successful.

Of course, it's a pity that the introduction of 40Gb platters didn't appear a turning point for the HDD performance boost, but it is just a normal evolution. If we revise the entire HDD history, we will see that the speed and the storage capacity didn't grow linearly. The evolution of each HDD family can be presented as a stepping scale. Seagate is now going through a horizontal step (increasing the data density). Do you think the users will be against twice as large HDD for the same money? We don't think so. As we have already told you, the customers buy MEGAbytes, MEGAhertz and MEGAtons :)

And as soon as Seagate engineers manage to squeeze higher performance out of these 40GB platters without increasing the prices dramatically, Seagate will launch another drive. Then there will be another step, and the next one…

Anyway, happy sailing, U6!