Modifying the Seagate 1.5TB Hard Drive: Unleash the Hidden Performance Within
Introduction
As many of you know, recently Seagate released one of the largest consumer drives on the market: the Seagate Barracuda 7200.11 ST31500341AS . The drive’s capacity is 1.5 TB (1500 GB) or approximately 1.397 TB. The drive specifications are decent and Seagate itself announced that this drive will be able to handle 120 MB/sec sustained transfer rate. All of us know that these rates will not be across the whole drive and were most likely obtained under the best possible conditions. That being said, we still can not overlook the fact that a 1.5 TB drive’s speeds place it directly in the Velociraptor territory. After discovering this astonishing speeds people normally bust out their Western Digital Velociraptor drives and start benchmarking them for comparison. So, what do they get? Velociraptor obviously wins in all categories – max speed, min speed, average speed, access time, you name it and it wins it. At this point they label the Seagate drive as second best and close the case. Of course thats the point, the Seagate drive was not meant to compete with the Velociraptor which is a performance drive and is instead targeted towards good performance and high storage capacity. So end of story right? . . . What if I told you there was a way to modify the 1.5TB Seagate to unleash the hidden beast within. . .
I decided to take the matters a step further. The WD Velociraptor is the largest “Raptor” class drive available on the market with the capacity of 300 GB. Despite the fact that it is the largest out of the lot, you still have a 300 GB drive compete against a drive that is basically five times its size. Normally, the giant will not stand a fighting chance, (look up David and Goliath if you don’t believe me) so I decided to even the fight out a little. What kind of speeds would you get if you were to take the large drive and downsize to the size of the Velociraptor? I decided to find out.
Prep for Surgery
Though really in this case we are not going to open up the drive, flash the firmware, or do cutting of any kind.
What we did to the 1.5 TB hard drive was to shrink it. There are a 5 easy steps you have to follow in order to accomplish that.
1. Download SeaTools DOS version from the Seagate website.
2. Burn the ISO onto a blank CD using any standard CD burner.
3. Put the burned CD into the CD-Drive and restart your computer.
4. The CD will boot automatically into the software’s graphical interface at which point you must click “ACCEPT” to the licence agreement.
5. Click the drive labeled “ST31500341AS” by left clicking the name.
|
|
 |
6. Go to the “Advanced Tools” tab and select the “Set Capacity Manually” option.
7. Once you have done that, a blank dialog should appear. Enter the amount of LBAs (Logical Block Addresses) you want your “new” drive to have and click “Continue”. The magic number for the 300 GB mark is 589080586. For reference the max number that you can enter into this field is 2930277167 which equals 1500.302 GB.
 |
 |
8. The confirmation message should look like the screenshot below. Once you achieved the confirmation, you are done. You can take out the CD and press the reset button or click “CTRL + ALT + DEL” to restart your computer.
**Note: The “Set Capacity Manually” only accepts LBA number so if you put something like “300 GB”, the software will set the size to zero and you will have to shut down the computer, restart and put the correct number into the Box. Also, after the capacity has been set, the software can not alter the size unless you completely Shut Down your computer, a simple Reset will not work.
After you set the size less than the maximum capacity, the remaining space will be invisible to the system and thus rendered useless. (If you downsize the 1500 GB drive to 300 GB, the other 1200 GB would be invisible) The drive will effectively become 300 GB in size.
Now I can hear you all screaming “What happened to 1.2TB worth of my drive space?” As I mentioned the drive is now a 300GB drive for all purposes. Look at it this way, the 300GB Western Digital Velociraptor is approximately ~$229 while the price of the Seagate 1.5TB drive is approximately ~$119 at the time of this writing. Now what you end up with is a drive that is higher in performance in all regards except seek times at a lower cost per GB. We are all about getting more for less around here. The added benefit is that you can always go back and reclaim that 1.2TB at any time, try to expand the Velociraptor at a later date (good luck).
I think there are too many variables to draw conclusions on which dive manufacturer is better eg no. of platters, perpendicular recording, density, transit etc etc. Some say seagate great and some no. I remember an article a while back published by google which detailed the drives that had a high fail rate, though they kept this part secret from the public. My best experiences have been with Hitachi and Fujitsu.
How bout you guys?
The method that your doing is called Short Stroking its some what common on some enterprise file servers where speed is more important then size.
Please update the test with just one 300gb partition created and compare with ‘disk’ created with this method and Velo.
You know, something was bugging me about this and I think I just realized what it is.
All this is doing is tricking the benchmarks into reporting higher numbers. Forcing the drive to only use the first 300Gb isn’t going to make it perform any differently than if the data is properly defragmented and is already at the beginning of the each platter. Hard drive benchmarks test transfer rates at various parts of the drive, so the average speeds are influenced by parts of the drive that you may never use. By limiting the area the drive uses, you’re just limiting the area that the benchmarks will test, which will make the test results much higher, without actually changing anything at all.
I have a hunch that this will bring little to no actual performance improvement outside of benchmarks.
Yes, it is making the drive only see the first 300 gb, but yes, it does make the drive perform better. We have tested in both lab set benchmark environents, as well as real world video editing, gaming, and every day use. Yes it is faster, yes the seek time is a bit slower than the velociraptor, but its also half the price…
Sorry to say, but you are wrong. It is _not_ just fooling the benchmarks. It is changing the useable portion of the disk so that only the outermost tracks are used. Since RPMs remain the same regardless of where you read the track, the angular velocity of the outside tracks is higher. Thus, most bits pass by the head in the same amount of time (as compared to an inner track).
Seek times should also be slightly improved as you effectively can access more data in the same amount of time. Of course, you will not be able to match the seek times of a drive with higher RPMs, but will should see some improvement.
Theoretically, limiting the drive to say, 100GB should even offer a bit more improvement – but only as long as it is the outermost tracks that are used.
What about the disk temperature? Won’t it go higher after the mod?
Shouldn’t We haven’t seen any temp differences that called our attention…
Pretty simple, just make your Operating system/lots of data altering partitions in first 300G, and data storage like mp3′s and so to inner-circle partitions that requires little altering/speed.
Not sure what’s the situation nowaydays, but some disks i’ve seen had end faster. aka starting from inside track, so testing is smart, unless you want to install os to slowest partition.
having single big partition IS tempting, but atleast in windows use, oh well… it’s not an option really.
What you are doing is just using the fastest part of the disc. On any disc it gets slower as you get into the center as the data density is reduced, your just avoiding that. Disc speeds tail off, your just chiopping of the tail, albiet a long tail.
you can get the same effect with regards to an operating systems level of usage from the disc by partitioning.
Run your 300gb setup using some realworld application benchmarks, then create normal 1.5tb affarid disc an da 300gb partition at the start and rinse and repeat, notice its pretty much the same.
only aspect your highlighting is how HD benchmark tools benchmark discs and not partitons, your just partitioning at a lower level and indeed loosing the rest
.
Good partitioning and OS/data placment across discs/partitions goes along way in performance of a system, seperate disc for swap/temp space, ew yeah.
Now spinwrite, thats a HD tool you may wich to play with, check the disc controller buffers and the disc details and tune to perfection and then adjust your filesystem acordingly to use a blocksize that gets the best, thats were you can tune your HD. Also alot of HD’s have NVRAM settings which need extranal programs to set and can adjust noise or indeed speed it up, spin down and the like. Speed/power is a trade of between noise and cost, mostly running costs.
Now go tweak.
You should really explain at least why you think this is happening.
Now for me i think it’s simply because you are forcing the drive to use the inner most of the disk which is obviously faster. Since the 1.5TB is one of the highest capacity drives at the moment it will use as many platters as seagate can fit while the velociraptor in its 2.5″ enclosure will have less.
Therefore I don’t really see what you proved. Limit the velociraptor to use 80GB of its inner platters and you’ll see a super super fast drive. It’s not magic. All you proved is that seagate’s utility is quite clever.
Funnily enough the only thing that wouldnt change at least much is the seek speed, which is probably whats nicest about having a 10k drive in the first place no? I guess the fashion atm is to go back to 5400rpm drive and claim you are green… I’m gonna ditch my 15krpm SCSI array and go SSD….
There was a similar article a long time ago in THG (with diff. sized hdd). However the same question that bugged me there applies here – why exactly the 1.5TB model? Can we do the same with 1TB or 500GB HDD? After all, you benefit most if you transfer your Windows boot partition to a faster hdd and for that 300gb is an overkill…
“or the data which you read or write is sequential (most of the time it is) ”
Most of the time it isn’t. If you’re running the OS from the drive, small random writes keep happening in background as long as you run the OS. How many times per day do you copy/paste multi-gig files? Although I agree that if this really translates into real world performance, you’ll be getting really good speeds for the price. Maybe you could put 2 Barracuda’s in RAID instead of getting a single velociraptor for the same money.
Great Deal your brain had ! ….
Why not havin an Idea and check it out ?!
Yes you did !
WTF didnt you just create a 300 GB “velociraptor – patrtition” ?
And another one for some unspectular data.
So, that you have a superfast velo’ (velo is a bycycle in Switzerland) plus a additional supebig data hd ?
dont mind me joking.
kalkzone