get 510GB out of 200gb or 120 out of 80.. etc
- Thread starter Cha0s
- Start date
Just put a sticker on the drive that says "Capacity: Eleventy-billion GB" it's just as effective!
You can add racing stripes and a spoiler to improve the sustained transfer rate too, who needs RAID0?
You can add racing stripes and a spoiler to improve the sustained transfer rate too, who needs RAID0?
I agree with the follow up letters. This is NOT a good idea, there's a good chance that the space you're "creating" is actually a terrible idea, you're probably corrupting the partition tables; I'd be surprised if you could actually use this "extra" capacity. As was stated in one of the letters, there is some extra space on hard disks but it's there for things like error correction or to use if bad block turns up.
if anyone has a harddrive they dont need, please try it. Or send it to me and i'll be happy to see if it works
It's a year old article and people are still not doing it. It should be blindingly obvious that it does not work reliably.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,570
- 10,202
- 126
Nah, it's complete BS. TheInq is being irresponsible for even running that article, before having a tech-guru editor review it. Although IDE HDs do have a small amount (1-2%) of "spare sectors" on them, for repair/replacement of ones that go bad via a host-transparent remapping mechanism, they most certainly DO NOT have 200% of the listed capacity or more in them. Especially when you look at the maximum theoretical per-platter storage, and the number and generational age of the platters used in the drives that they tested. That should disprove it right there.
All I can figure is that they are creating erroneous and overlapping partition tables, who's total overall size is collectively larger than the drive itself, but that's purely an illusion, and if you attempt to fill/write-to all of the partitions to their maximum capacity, you'll simply overwrite and corrupt the filesystems.
The *only* time that such "space reclaimation procedures" might be possible, is when you have a HD with an underlying physical capacity, that is much greater than the current capacity reported to the host system, because the drive has been programmed with a "soft LBA clip" setting, smaller than the "physical max LBA" setting. This could happen, sometimes, if you recieve an RMA replacement drive, and the drive that they send you is larger than the one that you sent them. Sometimes, the larger drive will be "soft-clipped" to the size of the smaller, original drive. You can simply "unclip" it to the max size again, and return the larger drive to its natural state. But for drives already sold as such, you can't make them "magically bigger".
All I can figure is that they are creating erroneous and overlapping partition tables, who's total overall size is collectively larger than the drive itself, but that's purely an illusion, and if you attempt to fill/write-to all of the partitions to their maximum capacity, you'll simply overwrite and corrupt the filesystems.
The *only* time that such "space reclaimation procedures" might be possible, is when you have a HD with an underlying physical capacity, that is much greater than the current capacity reported to the host system, because the drive has been programmed with a "soft LBA clip" setting, smaller than the "physical max LBA" setting. This could happen, sometimes, if you recieve an RMA replacement drive, and the drive that they send you is larger than the one that you sent them. Sometimes, the larger drive will be "soft-clipped" to the size of the smaller, original drive. You can simply "unclip" it to the max size again, and return the larger drive to its natural state. But for drives already sold as such, you can't make them "magically bigger".
No I've actually heard of things like this and manufacturers do this intentionally! Though this software thing is simply screwing up the partitions and there is no real benifit, there are such things as unused space in that way (not using a particular side of a platter). Manufacturers do this as a marketing thing where they have 3 drives, 1 250GB, 1 200GB and 1 160GB drive. The 250GB drive has 5 platters which uses both sides and the 200GB drive can theoretically use 4 platters. But the 160GB drive has a problem, since the density in this case is 25GB per platter, they can't get an even number for the 160GB drive. So in order to achieve a size of 160GB, they just use a 4 platter drive and only use 7 sides of the 8 sides available.
In some cases, HDD manufacturers can get SO lazy that they will leave all 5 platters in a drive but only sell a particular capacity because it's simply easier. In theory the HDD manufacturers will also use platters where one side is damaged in drives that only need an odd number of sides.
In some cases, HDD manufacturers can get SO lazy that they will leave all 5 platters in a drive but only sell a particular capacity because it's simply easier. In theory the HDD manufacturers will also use platters where one side is damaged in drives that only need an odd number of sides.
harddrives should be super cheap. I dont think it costs them much to make one if they ignore 5 platters or so
They are super cheap, n00b.
Prices are about 1/100 (one one hundredth) of what they were just 7 years ago, and that's ignoring 7 years of inflation. I paid $116 for a 2.5 GB drive on 9-11-98.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,570
- 10,202
- 126
Actually, you don't mean "unused", you mean "unusable". Small difference there.
Whether one side of a platter is either: 1) too defective to make effective use of, or 2) remains "unfinished", in order to save on costs, in either case you cannot somehow convince the HD to use that "extra space", since the platter isn't finished, and even if it was, there is no factory-recorded low-level servo information. Remember, you can't low-level format IDE drives in the field.You're more-or-less right about some drives not making use of both sides of a platter, but the numbers are slightly off. Current 160GB HDs, use two platters, each 80GB in size, which implies 40GB/side. So newer 120GB HDs, are generally 160GB drives that didn't make the grade, and have only three finished platter sides to them. Older 120GB HDs used 60GB platters, so there were two platters, each with two 30GB sides. Generally, higher density per platter is better, because it implies a higher linear media-transfer rate, and thus a higher continuous data-transfer rate from the drive to the host. (So the newer 3-sided, 80GB/platter 120GB HD, would likely be faster than the older 4-sided, 60GB/platter 120GB HD.)
Few consumer-level IDE HDs ever used more than three platters per spindle; one of the few that did was the ill-fated IBM 75GXP family, and recent IBM, Maxtor and Seagate drives have reached back up to four platters/spindle. (At 100GB/platter, that gives a max size of 400GB per drive.) Prior generation drives generally were one, two or three platters/spindle, like Maxtor's DM+9 and WD's "JB" Caviar Special-Edition drives.
That's about the only thing that's not true - unlike CPUs, that come off of the same assembly-line, and cost the same amount to make, regardless of final speed grade (well, testing costs may vary a tiny bit) - HDs have a mfg cost that differs depending on the number of platters in the drive, and HD mfg's operate on such slim profit margins, it would be financial suicide to intentionally leave good platters in a HD, unused.
Yes, that's true.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,570
- 10,202
- 126
Oh, and as far as pricing, I remember paying $300 for a WD 1GB HD, around 96 or 97. I went from 1GB to 2GB, and it seemed like heaven at the time!
Well back when HDs were only 1gb I remember actually needing every bit of that space. But now, a $80 HD has more room than I'll ever need, this is certainly a nice technological advancement.
Originally posted by: VirtualLarry
Oh, and as far as pricing, I remember paying $300 for a WD 1GB HD, around 96 or 97. I went from 1GB to 2GB, and it seemed like heaven at the time!
I remember paying almost $4K for a Dell Dimension D300 in 1997. It had a 8.4GB IBM HDD. I wonder how much this drive would have cost, alone?
You young guys! I remember getting a great deal. A 200mb drive for $200. A dollar a meg! Of course ram was still $50 a meg then.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 19K
-
-
-
Facebook
Twitter