Why no 5-inch hard drives?

In the race to create ever bigger HDs, why hasn’t anyone simply gone the route of increased surface area and built a 5-inch form factor model? Lots of people have spare 5-inch bays in their computers.

I remember back in the day when a 9-inch 20MB disk cost a few grand. Those things were cool.

There were. I had one serving as a doorstop some time ago.

Several reasons:

High cost. No demand. Perfectly viable alternate solution. People clamboring like mad for SMALLER SMALLER SMALLER.

Computers are heading towards becoming “appliances.” I blame Apple.

There’s a few reasons:

Basic physics. In order to maintain the same performance numbers of today’s common 3.5" hard drive, it would take a much larger and stronger motor to spin the larger and heavier platters of a 5.25" HD at the same speeds. This would greatly increase the power used and heat produced. Also, the actuator arm in a 5.25" HD would have to be able to move much faster than one in a 3.5" drive, simply because the distance between the outer and inner edge of the platter is greater.

It’s relatively simple to combine multiple drives into a single, larger logical drive using RAID.

Like almost all things in computers, the trend is towards a smaller physical footprint, not a bigger one. I wouldn’t be surprised if in a few years the 2.5" HD becomes common or even standard in desktop PCs.

Good answer.

(Note: 'At was a JOKE. Y’know… jo-o-o-oke? Sorry…)

There’s something wrong with your basic physics.

What really counts is data transfer rate, not rpm. The transfer rate is limited by the number of bits actually moving past the head each second. There are three ways that this rate can be increased:

  1. Increase the data density
  2. Increase the rpm
  3. Increase the platter diameter.

A 5.25" platter has a diameter 1.5 times greater than a 3.5" platter. To achieve the same transfer rate as a 3.5" platter at 7,200 rpm, the 5.25" platter would only need to spin at 4,800 rpm.

The Quantum Bigfoot drive used this principle back in about 1997. It only spun at 3,600 rpm, but was claimed to achieve the same transfer rate as a 5,400 rpm 3.5" disk.

Yeah, but that’d only be true near the outside edges of the platters. If it was reading something from the inside edge, the only two things that would matter would be the speed and data density. The total diameter of the disc would be completely irrelevant.

Of course, I suppose it could average out. A slower RPM 5" disc could read slower than a faster RPM 3" disc on the inside, but it could read faster on the outside.

This is knowleage from back in '95 so things might have changed a bit in the computer world in 9 yrs.


Hardrive ‘rings’ had the same amout of data, just more compact nead the center and spread out more on the outside. The reason for this that was given was the HD can maintain a steady speed, instead of a variable speed that a CD-ROM does. This allows the head to change ‘tracks’ without a variable speed motor. Also another reason that was given was that the OS or BIOS wasn’t written for this.

If that still holds then enlarging the platterwould not gain you all that much and you might have to slow down the drive in order to read the outside track.

The common practice for larger drives seems to be having multiple platters.

As a side question, all HDs I´ve seen have one actuator arm only, why not add a second one to increase the reading/writing speed?

Maybe I should have been more clear that, other than platter diameter and capacity, I was considering all basic specifications (RPM, seek time, areal density, etc.) to be the same.

Let’s play with the Bigfoot example of increasing the platter size and lowering the RPM. The linear speed of the outer track of a 5.25" platter at 3600 RPM is the same as that of a 3.25" platter at 5400 RPM (990 in/sec). So, given equivalent areal density, both would have identical internal throughput for that track (i.e. peak throughput). However, once you move inward the disadvantage of the lower RPM really starts to show. For example, reading a track 1" out from center on the Bigfoot’s platter will be slower than reading from the same position on a faster 3.5" drive. 188.5 vs 282.75 in/sec, 33% slower.

Also consider seek times with a larger platter. If the actuator arm has to swing from the inside to the outside of the platter to get where it needs to be, it has far more distance to cover if the platter is larger. If the actuator arm in the 5.25" drive is the same as that in a 3.25" unit, the 5.25" would be at a disadvantage. This is why I mentioned the need for a faster actuator in a 5.25" unit.

Then there’s rotational latency, which is taken together with the seek time to give a more realistic indication of how long it might take to get to the needed sector to read/write the needed data (click here for more). In this respect, a lower RPM drive is always at a distinct disadvantage compared to a faster drive.

The most important thing about HD’s is not transfer rate, it’s latency which is directly correlated with rpm. Most apps dont use anywhere near the maximum transfer rate but they certainly start thrashing hard if the latency is increased.

Just want to throw in that I pulled a 9gig 5.25" drive from an old Compaq a few months ago. It was at the end of it’s useful lifespan, so I opened it up and pulled the neato magnets and mirror surface platters.

Fun for the whole family.