[comp.periphs.scsi] How big is a Gigabyte disk?

paul@stan.Solbourne.COM (Paul Orland) (05/30/90)

In article <11805@cbmvax.commodore.com> jesup@cbmvax (Randell Jesup) writes:
>In article <2814@medusa.informatik.uni-erlangen.de> csbrod@medusa.informatik.uni-erlangen.de (Claus Brod ) writes:
>>jesup@cbmvax.commodore.com (Randell Jesup) writes:
>>>	Try Quantum.  The Q40S is 40 MB, the Q80S is 80 MB, etc (formatted).
>>And the Q105S is 100.1 MB (formatted) X-(
>	Quantum MB are 1,000,000 bytes.  Oh well.  It has 105,033,728 bytes
>of usable storage (formatted, 512 bytes sectors).  Still a lot more useful
>than unformatted numbers.

Are there *any* disk vendors who don't consider a MB of disk storage to be
1,000,000 bytes?

It seems that most 5-1/4" SCSI vendors list the *unformatted* capacity of 
their drives, while most 3-1/2" SCSI disk vendors list the *formatted*
capacity of the drives. Even when the same vendor produces both form factors,
this discrepancy exists. When a formatted capactiy is listed, it is usually 
(but not always) determined using 512 Byte sectors and 1 spare sector/track.

Unformatted capacity comparisons can not be skewed by a vendor being 
more/less aggressive in their default defect handling schemes. As an 
example, giving 3 spare sectors/cylinder on a 9 data head 200 MB class 
drive instead of the more usual 1 spare sector/track (9 spare sector/
cylinder) will squeeze a few more MB out of the drive (3.9 MB in the 
case of a Seagate ST1239NS, 6 sectors saved per cylinder * 1272 cylinders 
* 512 bytes/sector). Usually, you take some of this back by allocating 
more spare cylinders per volume, but the benefits are still useful in 
gaining capacity, but can be misused by vendors when stating the capactiy 
of their drives.
--
--  --  --  --  --  --  --  --  --  --  --  --  --  --  --
Paul Orland   |   I/O Substystems   |   Solbourne Computer
paul@Solbourne.COM  -- --  --  ...!{boulder,sun}!stan!paul