This month Hitachi announced and demonstrated at the CES the world’s first 3.5” terabyte disk drive. Note, a terabyte is 1,000 times 1 billion bytes or characters. The drive is shown below.
It looks like any other 3.5” disk drive but inside it embodies the very latest in recording technology known as Perpendicular Recording. It is the latest in the long string of ‘pulling the rabbit out of the hat’ innovations that have characterized disk drive technology. On many occasions when it seemed that the limit of capacity and density had been reached someone would come up with a way to push the barrier further out and create a way to maintain the incredible rate of growth in capacity and performance that is the legacy of these devices.
Last year we celebrated the fiftieth anniversary of the disk drive. The first drive, developed by IBM, comprised 50 24” disks each one of which had the capacity of 100,000 alphanumeric characters of 6 bits each (the byte had not been invented yet) for a total of 5,000,000 characters. The box housing the spindle, disks and access mechanism weighed 1.5 tons! The new gigabyte drive has 26,000 times the capacity on 1/10th the number of disks in 1/11,000th of the volume and about 1/3,000th of the weight. The new disk drive can access more data in a single operation than the entire contents of the original drive and do it in 1/60 of the time it took to find a single record. All of this for 1/300,000th of the price! We have come a long way; the question is how much further can we go. Below is the original IBM 305 disk unit.
Before we try to answer that question let us look at perpendicular recording to understand how this latest improvement works. The recording basics really have not changed since the beginning. Magnetic spots are created on a rotating disk surface that can be detected by a magnetic pick-up known as a read head. The spots are actually oblong on the surface of the disk along the direction of travel. The read head senses the magnetic field of the spot and identifies it as either a digital 0 or 1 depending upon the direction of the magnetic field. Over time ways were found to make the spots smaller and smaller and tracks closer and closer until it appeared that no further improvement could be made.
Looking at the illustration above, the dark arrows represent the magnetic spot son the disk looking at a cross section of the disk, with the pick-up sitting above the disk detecting the magnetic fields as the disk rotates. The practical density limit of this approach was reached about 2 years ago. What to do? Well it turns out that a Japanese professor; Professor Sun-ichi Iwasaki in 1975 came up with the idea of creating vertical magnetic spots instead of horizontal spots. Sounds easy but it required a lot of hard work with materials, design and manufacturing techniques to make it work. But finally all the problems were solved. Below is the equivalent illustration for perpendicular recording. As you can see,several spots can be squeezed into the same horizontal space as was achieved with horizontal recording.
So here we are with another great leap forward. We should be able to squeeze another 2maybe 3 times the density of the new devices before we run into the wall again. What then?
Unfortunately it appears that there are no more rabbits in the hat. We are staring at the molecular limit of what can be done with the current technology. We are now back to research looking at several different approaches to storing data, none of which bear any resemblance to our current devices. Since these efforts are in research it is going to be several years before we know what the best solutions will be. Several appear promising. For sure, there will be something that comes along but today our vision is clouded.
So Ben Higginbotham, with respect to your pool, guessing when there will be a laptop with a 1 terabyte disk drive in it, I’m afraid I’m going to have to say “it’s not in the cards, at least not today”.
What does a 1 TB hard drive look like? Just like any other hard drive from my perspective, which goes to show why people shouldn’t judge hard drives by their understated looks.
On a related note, we have a pool going at work over which month the first notebook with a 1TB hard drive will ship. Care to wager a guess?
Alan F. Shugart passed away on December 12th of complications following open heart surgery six weeks earlier. He was 76 years old.
While he has not been active in the disk drive business for several years his footprint will last for decades. His most visible legacy today is Seagate Technology which he co-founded with Finis Conner in 1979 as Shugart Technology, later to be named Seagate Technology. We know Seagate Technology today as the largest producer of disk drives in the world.
A few months ago we celebrated the 50th anniversary of the disk drive. Alan was an engineering member of that first development team in San Jose, California. We were colleagues in San Jose during the 50‚Äôs and 60‚Äôs, Alan was on the disk drive side and I was responsible for systems development. Over this period the disk drive moved from an interesting concept to a fundamental, integral part of every new computer system. We worked closely together to insure the proper trade-offs as we integrated this new medium into an on-line facility. When the IBM S/360 was announced in 1964, after only 5 years in the marketplace, the disk drive had transformed the concept of a modern computer system forever. The success of the early disk drives was due, in no small part, to Alan‚Äôs skill, stubbornness and determination. He was known for speaking his mind and not being afraid of whomever he expressed his opinion to. In the end, it was this characteristic that ended his IBM career in 1969. Before he left, however, one of his pet projects was the development of the 8‚Ä? floppy disk drive. It was designed for loading microcode into the S/360 mainframes. He knew that there was a much larger potential for this device so he took the idea with him and as the PC business started to emerge he re-incarnated it as the 5.25‚Ä? floppy that became the backbone of the early PC products.
Alan Shugart was a true pioneer. Unafraid of the unknown, full of self confidence and a tough task master. His name does not appear in the lights of Broadway but his legacy lies deep within every single disk drive that is produce today.
I shall miss him.
As I write this my laptop is telling me that my 60 Giga Byte hard drive is almost full and I must do something about it. This is serious because my laptop and my smart phone are my inseparable companions. They are my constant companions and connect me with the rest of the world through Wi-Fi and a Sprint data card. My laptop is my desktop, my library, my contacts, my e-mail, my presentations, my digital photo albums, my personal information, my professional work, even my personal history. Without the disk drive my life would be a mess. I would have to resort to tons of pieces of paper and my professional life with IBM would never have materialized.
Well, I have some choices, I can:
- Upgrade to a new laptop with 100 giga bytes of disk space. It solves the problem but what an ordeal to go through.
- Dump some of my least used files to CD?¢Ç«®Ç—¢s or a backup disk drive. It buys me some time but it is only a temporary solution and I will miss the data I dumped.
- Compress the data and gain some space but then I lose some performance in dealing with those files.
- Offload some of my files to a web based storage service such as IBackup for Windows but then I can?¢Ç«®Ç—¢t guarantee that I will always be able to reach them.
I don?¢Ç«®Ç—¢t think I am alone in facing this dilemma. Our daily lives have become dependent upon the digital data that these little devices soak up like pouring water onto sand. Now they have even jumped the computer boundary and have penetrated our entertainment domain. What young person would be seen without an iPod or similar musical device, or what would we do without our Tivo that captures TV programs for us and is soon to be embedded into the latest HDTV?¢Ç«®Ç—¢s. We can throw away our driving maps because all the information is captured onto a hard drive embedded in a GPS navigation device. One inch devices have even appeared as Compact Flash cards for digital cameras. Could ATM machines exist without these devices? Of course not. Who knows where the next appearance is going to be?
All of this says nothing about running businesses, airlines and the like. What businesses today could operate without immediate access to huge data bases managing transactions, schedules and a million other things that need to be attended to?
OK, so hard disk drives are a big deal, how did they happen? Innocently enough: to solve a problem that would limit how useful a computer could be. Let?¢Ç«®Ç—¢s follow the trail.
Before computers, records were kept on punched cards. There were thousands and thousands of these cards kept in trays of about 3,000 cards each. A file was one or more trays with cards that were kept in sequence. The sequence was important because that was the only way they could be processed. Additions to the files were made by merging the additions into the master file using special equipment. The first primitive commercial computers simply performed the arithmetic calculations for such things as payroll or inventory or accounts payable etc. and entered the results into the punched card
When the big mainframes appeared (that was the only way to build a computer in the beginning) the only digital data was on punched cards so card readers were attached to the computers. But, that was a terribly slow way to process because the computer would sit idly waiting for the next card to be read etc. The solution was to convert analogue tape drives into digital tape drives because tape could be read and written to much faster and could hold very large files in a more compact media. This was much better but still not ideal. To find a particular record the computer would have to search through a tape, record by record until it found the one it was looking for. Again, a lot of wasted time for a very expensive machine. Besides, doing work in batches was a function of the way records were assembled into files, not necessarily the way a business wanted to run. For instance, it was not possible to find out what the inventory of a particular item was without running the inventory tape and wading through lots of records before the item record appeared. A totally impractical use of a mainframe.
IBM in 1952, recognizing that in order for computers to realize their potential this problem must be solved, asked an engineer by the name of Rey Johnson to form a research group in San Jose, California to study possible solutions to this problem. He quickly formed a small group for this purpose. Univac was also studying this problem and had decided to use large magnetic drums rotating at high speed as a their solution. Rey Johnson felt that the outer surface of a drum would not hold enough data to solve the problem. He was intrigued by the idea of using disks somewhat like the long play record disks of that time but stacked in multiples to provide a much larger available surface for recording. After four years of development the basic engineering issues had been put to bed and IBM, on September 13th 1956 announced the 305 RAMAC ?¢Ç«®Çƒ˙ a Random-Access Method of Accounting and Control System. It comprised a disk unit coupled with a small vacuum tube, relay and drum computer, a console, card reader and stick printer.
The disk unit contained 50 - 24 inch diameter platters on a spindle with an arm with two read/write heads that were inserted between the disks. The arm assembly was driven up and down a track by compressed air and that was also fed through the read/write heads, preventing the heads from coming into contact with the magnetic surface of the disks. The capacity of the unit was 100 million characters of 6 bits plus parity each. The disk unit weighed one ton!
Just 50 years ago, shortly after the announcement, a truck pulled up outside one of my customers ?¢Ç«®Çƒ˙ Zellerbach Paper Co. on Battery Street, San Francisco. In it was the #1 RAMAC to be delivered to an IBM customer for application development. Needless to say, it was a thrill to be associated with such a break through product but it was beyond my wildest dreams that I would become deeply involved in the development of many of IBM?¢Ç«®Ç—¢s products that depended upon these disks.
Thus San Jose became the home of The IBM Disk Drive even though in later years disk development activity also took place in Hursley, England, Yamato, Japan and Rochester, Minnesota.
The disk drive was critical to IBM?¢Ç«®Ç—¢s success but it was the Personal Computer that really put the disk drive on the map. The PC forced the miniaturization of the device and large scale manufacturing that achieved economies of scale undreamed of until then.
My humble laptop has 18,000 times the capacity of the original RAMAC, is more than 3,000 times smaller, weighs a few ounces instead of a ton yet we know that there is more to come. The current technology curve will provide several more generations of improvements and, true to form, I expect the demand will be there waiting for the next step in the revolutionary process.
So, I guess my next step will be the newest, latest and greatest because my needs will inexorably continue to grow - this is the age we live in.
Recent Comments