What was the cost for a 1GB hard drive in 1986?

Robert Hankey

Commendable
May 6, 2016
2
0
1,510
The first Gigabyte hard drive was created by IBM in 1980. It was the size of a refrigerator and cost $40,000.

What is the history from 1980 to present, (2016) for

[1] The maximum capacity of a commercially available hard drive

and

[2] What was the cost per Gigabyte?
 
Solution



Robert Hankey,

Sorry, I don't know the cost of a GB in 1986, but I do remember a couple of cost points on both sides of those dates.

My father, who was a writer for a government consultancy, about 1970 had a computer terminal at home that communicated with a mainframe at his office, an IBM 360 I think. He was excited with the news that the system was going to be getting a "hard drive". The systems then ran programs from punch cards and the storage...

Mark_1970

Reputable
Nov 14, 2015
1,391
3
5,960


Sounds like someone wants us to do their homework :lol:

Best go look at Wikipedia
 

Robert Hankey

Commendable
May 6, 2016
2
0
1,510


"Sounds like someone wants us to do their homework"

>> Then what is the purpose of this website?

"Best go look at Wikipedia"

>> If the answer was in Wikipedia, I would not have posted this question. If you don't know the answer, what is your purpose for responding with a non-answer?
 

Mark_1970

Reputable
Nov 14, 2015
1,391
3
5,960


Here you go............https://en.wikipedia.org/wiki/Hard_disk_drive

Mostly the purpose is to help with pc issues
 



Robert Hankey,

Sorry, I don't know the cost of a GB in 1986, but I do remember a couple of cost points on both sides of those dates.

My father, who was a writer for a government consultancy, about 1970 had a computer terminal at home that communicated with a mainframe at his office, an IBM 360 I think. He was excited with the news that the system was going to be getting a "hard drive". The systems then ran programs from punch cards and the storage drives were tape. I never saw this hard drive- but it was described as being like a small barrel about 2' in diameter. I think those drives had four or more platters. The capacity was 5MB and the cost was about $50,000 or $10,000 /MB, which equals $10,000,000 / GB. I think files then were 2-10KB.

In 1993, I bought my first modern system, an IBM 486. This was a 50MHz with a DX2 math coprocessor that allowed the system to run AutoCad and graphics programs.

The system cost $2,300 and had an 85MB IBM HD. This system DOS 6 /Windows 3.1, WordPerferfect 6, AutoCad 10 for DOS, and Corel Graphic Suite 3. WordPerfect 6 was the first version with GUI where the fonts displayed

The file size was tiny in those days- a big AutoCad drawing was 20-80KB. It took about 6 months to fill the 85MB drive. At the time, DOS could only see 528MB, so I bought the largest HD available, a 540MB of which 528MB was available. The cost of the 540 MB was $570 or about $1.05 /MB. Therefore, in 1993, that particular HD cost $1,050 /GB. It's amazingly to think that a 1TB drive at that rate would cost $950,000.

The price per MB was so high that I was obsessively frugal with drive space- drive compression was common in early Windows. I never kept old files on the system, but wrote them to 1.44MB floppy disks and later to CD. It wasn't until about 2000 that the cost of storage dropped to the point where I didn't take the cost seriously.

I often wonder why people have such huge drives. I have a quite high percentage- perhaps 90% of everything I've made or kept on a computer since 1993- including many WordPerfect files still with DOS extensions and that total is about 75GB. That doesn't include reference images/ media files and some program bits- all together I keep about 400GB.

But, as far as hard drives are concerned- the astounding speed of M.2 NVMe SSDs and recently 2TB SSD's- the golden age is now- I have no nostalgia for $10,000,000 / GB!

Cheers,

BambiBoom

Modeling:

1. HP z420 (2015) > Xeon E5-1660 v2 (6-core @ 3.7 / 4.0GHz) > 32GB DDR3 1866 ECC RAM > Quadro K4200 (4GB) > Intel 730 480GB (9SSDSC2BP480G4R5) > Western Digital Black WD1003FZEX 1TB> M-Audio 192 sound card > 600W PSU> > Windows 7 Professional 64-bit > Logitech z2300 speakers > 2X Dell Ultrasharp U2715H (2560 X 1440)>
[ Passmark Rating = 5064 > CPU= 13989 / 2D= 819 / 3D= 4596 / Mem= 2772 / Disk= 4555]

Rendering:

2. Dell Precision T5500 (2011) (Revised) > 2X Xeon X5680 (6-core @ 3.33 / 3.6GHz), 48GB DDR3 1333 ECC Reg. > Quadro K2200 (4GB ) > PERC H310 / Samsung 840 250GB / WD RE4 Enterprise 1TB > M-Audio 192 sound card > Logitech z313 > 875W PSU > Windows 7 Professional 64> HP 2711x (27", 1920 X 1080)
[ Passmark system rating = 3844 / CPU = 15047 / 2D= 662 / 3D= 3550 / Mem= 1785 / Disk= 2649] (12.30.15)

Other:

3. Dell Precision T3500 (2011) (Rev 2) Xeon X5677 4-core @ 3.46 / 3.73GHz > 12GB (6X 2GB) DDR3-1333 ECC > Quadro 4000 (2GB) > PERC 6/i + Seagate 300GB 15K SAS ST3300657SS + WD Black 500GB > 525W PSU> Windows 7 Professional 64-bit > 2X Dell 19" LCD
[Passmark system rating = 2751, CPU = 7236 / 2D= 658 / 3D=2020 / Mem= 1875 / Disk=1221]
[PT9 BETA > Passmark system rating = 2696, CPU = 6595 / 2D= 636 / 3D=2391 / Mem= 1811 / Disk=1203]
 
Solution

Paperdoc

Polypheme
Ambassador
"Commercially Available" puts a big restriction on what you want. In 1986 the newest desktop PC equipment was hitting 16 to 20 MHz CPU speed and IDE hard drives were typically about 30 MB. The GB hard drive you cited by IBM certainly was NOT aimed at the desktop market - it was too huge and MUCH too expensive for that. I suspect, in fact, that it was a developmental device aimed mostly at research and bragging rights, and eventually for mainframe houses.

Bambiboom's comments triggered memories. At that time I was a graduate student in Physical Chemistry dong research in Far-Infrared absorption spectroscopy. Hardware limitations meant that such work was possible then only by using Fourier Transform Interferometry, and that only had become possible in the mid-1960's when MAINFRAME computers got powerful enough. The university's computer centre at that time ran an IBM 360/60 machine, I believe, and we were all happy when it expanded its core memory from 256 to 512 KB. Large data storage was by reel-to-reel tapes (1" wide). Fast random access data storage was by units called "Disk Packs". These were an earlier form of hard disk unit, but did NOT use the Winchester design of having the heads "float" on a thin boundary layer of air close to the disk surface. They used heads supported by rigid moving arms to maintain the proper clearance of head to disk surface. A "Disk Pack" consisted of about half a dozen 12" diameter magnetic platters mounted on a common vertical spindle about 6" high. It was kept in a protective case with a handling mechanism to make direct contact with the disk by hand unnecessary, The entire Disk Pack assembly could be transferred from its case into the drive unit that opened at the top and then closed for running. As it operated the disk pack spun on its central shaft while the drive unit moved arms carrying the heads over the individual disk surfaces. There were separate arms with heads for each disk in the pack all on a common head arm system. I don't know but I suspect they were using only ONE side of each disk. I don't remember the capacity of a single Disk Pack, but it must have been in the several MB region. I do remember that one service offered for scheduled times each day was the use of an interactive language and OS called AP/L. To run that, the mainframe operators shut down some other batch processing services and mounted the AP/L Disk Pack(s) in the drive(s). Users used a typewriter terminal (IBM's "golf ball" Selectric type) for interactive access, and were able to use a workspace of 32 KB at a time, although they had assigned to them each several possible 32 KB workspaces. Each workspace contained both programs and data and resided on the Disk Packs. So each user might have had up to 200KB of storage resources, and the system could support several dozen users simultaneously, probably using 2 to 3 Disk Pack drive units in the mainframe operating room.

In my research the data from the interferometer was recorded on 1" punched paper tape, then taken with job control punched cards and a bit of data to the computer centre, where the Fourier transform work was run as one of the batch jobs. The AP/L system I described above did not do this job. (The Fourier Transform job required significant storage, processing speed and file access not available in AP/L.) But I did use the AP/L system extensively for processing data from many other instruments measuring dielectric properties of samples covering the broad frequency range from 50 Hz up to 150 GHz. Most often these data analyses required fitting experimental data to linear regression lines and deriving dielectric propertied from them, and we had several AP/L programs written for these.