Can we increase capacity of hard drive?

Hard disk drives have become an essential component of modern computing systems, from personal computers to data centers. As demand for digital storage continues to grow, driven by big data, internet of things, digital media and other applications, there is a need to continually increase hard drive capacity. But how much further can capacities be pushed with current technologies? Let’s examine the methods available for increasing hard drive capacity and the challenges involved.

What limits hard drive capacity?

The storage capacity of a hard disk drive depends on two main factors – the areal density, which is the number of bits that can be stored per square inch of disk surface, and the number of platters in the drive. For decades, areal density has been steadily increasing through advances in magnetic recording technologies. However, we are now approaching the theoretical limits of perpendicular magnetic recording, the current approach used in hard drives.

Some of the challenges faced in continuing to scale up areal density include:

  • Magnetic grain size – Data is stored in tiny magnetic grains on the disk platter. But there is a physical limit to how small these grains can become before data stability is affected by magnetization switching errors.
  • Signal-to-noise ratio – As grains get smaller, the signal they generate also decreases, while noise remains constant. This lowers the SNR and makes data reads less reliable.
  • Superparamagnetism – When magnetic grains become too small, they can lose their magnetization at ambient temperatures through superparamagnetic effects. The data is then lost.

So there are physical constraints from magnetic recording principles that impose limits on how far we can push areal densities.

Perpendicular Magnetic Recording

Current hard drives use a recording technique called perpendicular magnetic recording (PMR). This was introduced in 2005 and allowed the recording density to be significantly increased by aligning magnetic orientations vertically rather than longitudinally along the surface.

Some of the technological innovations of PMR include:

  • High coercivity media – Materials like iron platinum allow magnetic bits to be packed more closely without interfering with each other.
  • Soft magnetic underlayers – Improve writeability and alignment of magnetic bits.
  • Double write heads – Allow both sides of a magnetic bit to be defined independently.

PMR enabled areal densities to be scaled up to around 1 Tb/in2. But we are now close to the limits of what can be achieved with this approach.

Shingled Magnetic Recording

One technology that allowed capacities to be further increased is shingled magnetic recording (SMR). This writes new tracks partially overlapping previously written tracks, ‘shingling’ them together to pack more data onto the disk.

The challenges with SMR include:

  • Overwrites require rewriting overlapping neighboring tracks.
  • Needs complex algorithms to arrange writes.
  • Performance impact from rewriting tracks.

So while SMR can increase capacities in principle, commercial adoption has been limited by technical difficulties.

Two-Dimensional Magnetic Recording

An advanced concept that has been researched to extend perpendicular recording even further is two-dimensional magnetic recording (TDMR). This uses nanowires arranged vertically on the disk surface to provide an extra dimension for data storage.

Benefits of TDMR include:

  • Allows for an ultra-high areal density in the range of 10 Tb/in2 theoretically.
  • Nanowires provide very high magnetic anisotropy for thermal stability of bits.
  • Large arrays of nanowires can be densely packed on disk platters.

However, there are difficulties in practical implementation:

  • Requires entirely new write heads optimized for vertical nanowires.
  • Nanofabrication of high density nanowire arrays is challenging.
  • New signal processing algorithms needed to operate in two dimensions.

So while promising, TDMR technology has hurdles to overcome before it is commercially viable.

Bit Patterned Magnetic Recording

Bit patterned media (BPM) is an alternative approach being researched to maximize magnetic recording densities. This involves etching an array of highly uniform magnetic pillars onto the disk where each pillar stores an individual bit.

Key advantages of bit patterned media:

  • Eliminates superparamagnetism issues faced by very small grains.
  • Allows for precise placement and size control of magnetic bits.
  • Enables ultra-high areal densities beyond 1 Tb/in2.

However, fabrication and integration challenges exist:

  • Requires nanoscale lithography to etch dense arrays of tiny magnetic pillars.
  • New servo tracking mechanisms needed to position heads over pillars.
  • Synchronization of reads/writes between pillars and heads.

So there are still technology hurdles to clear before bit patterned media can become economically viable.

Energy Assisted Magnetic Recording

To keep pushing magnetic recording densities higher, techniques like microwaves or lasers can be used to transiently heat the media and lower its coercivity. This allows stable writing of smaller bit sizes beyond traditional thermal stability limits. Some approaches include:

  • Microwave Assisted Magnetic Recording (MAMR) – Uses microwaves to heat media.
  • Thermally Assisted Magnetic Recording (TAMR) – Uses a laser diode for heating.
  • Helium Plasma Recording (HAMR) – Directs a beam of helium and electrons for heating.

Benefits of energy assisted recording:

  • Allows for stronger media with higher anisotropy.
  • Enables writing of very small stable magnetic bits.
  • Densities of up to ~2-4 Tb/in2 theoretically possible.

Challenges include:

  • Optical and thermal integration into write heads.
  • Media heating requires precise timing and temperature control.
  • Higher complexity of recording heads.

So energy assisted magnetic recording can potentially keep PMR going to higher densities. But integration for mass production will require significant engineering.

Hard Drive Actuator Density

In addition to areal density on the media itself, the number of bits per square inch can be increased by adding more platters and recording heads to the hard drive. However, the actuator that positions the heads has physical space constraints. Extremely close spacing of heads and platters risks collisions and interference.

Some innovations that help increase platter densities include:

  • Nitrogen-filled housings – Reduces turbulence and vibration for more precise head control.
  • Helium drives – Fill housing with low-density helium to allow closer platter spacing.
  • Shingled actuators – Overlapping but staggered read/write heads.

Hard drive companies continue to innovate on actuator design to pack more platters into drives. For example, new types of piezoelectric materials are being used to allow fine head positioning control. But there are still physical limits to how closely platters and actuators can be packed while preserving reliable operation.

Non-Volatile Memory Integration

Non-volatile memory (NVM) technologies like phase change memory, spin-torque transfer RAM and 3D XPoint can be integrated with hard disk drives to create hybrid storage devices. These exploit the low cost per bit of HDDs and the fast read/write speeds of NVM to optimize storage system performance.

Some options for NVM/HDD hybridization include:

  • SSD caching for hard drives – Flash solid state drive as cache.
  • Non-volatile cache on HDD – Small NVM cache on drive.
  • Separate NVM chips – Flash or 3D XPoint chips alongside but separate from HDD.

Advantages of hybrid drives include:

  • Improves performance and lifetime of HDDs.
  • Provides faster access to frequently used data.
  • Enables new applications like computational storage.

There are still challenges around intelligently managing and tiering data between the NVM and HDD components for maximum benefit. Hybrid drives also add to complexity and cost. But overall, integration of NVM technologies can significantly boost hard disk drive capabilities.

Shingled Magnetic Recording

One technology that allowed capacities to be further increased is shingled magnetic recording (SMR). This writes new tracks partially overlapping previously written tracks, ‘shingling’ them together to pack more data onto the disk.

The challenges with SMR include:

  • Overwrites require rewriting overlapping neighboring tracks.
  • Needs complex algorithms to arrange writes.
  • Performance impact from rewriting tracks.

So while SMR can increase capacities in principle, commercial adoption has been limited by technical difficulties.

Hard Drive Form Factors

Increasing the physical size and dimensions of hard drives expands the area available to add more platters and heads. For example, newer drive form factors like M.2 allow components to be packed more densely compared to traditional 2.5″ or 3.5″ HDDs.

Larger drives also tend to incorporate more advanced features like:

  • More powerful processors for caching algorithms.
  • Dual actuator arms to reduce head seek times.
  • Independent head control for finer servo positioning.

But bigger is not always better. Large multi-terabyte HDDs face technical hurdles like:

  • Increased vibration and shock sensitivity.
  • Difficulty maintaining alignment over more platters.
  • Higher power requirements and heat build-up.

So drive form factors require balancing size against practical engineering limits and use case requirements.

Helium vs Air in Hard Drives

Using helium instead of air in hard drive enclosures can boost storage densities. Benefits of helium drives include:

  • Thinner platters can be stacked closer together.
  • Low turbulence allows more precise head control.
  • Convective cooling is more efficient in helium.
  • Less drag allows lower platter rotation speeds.

Challenges with helium drives:

  • Require sealed enclosures to contain the helium.
  • More complex manufacturing and assembly processes.
  • Susceptible to leaks that can render data unrecoverable.

Overall, helium allows around 40% higher storage density than air for a given form factor. But leakage risks and higher costs limit adoption to high-end enterprise drives. Lower density air drives remain preferred for consumer applications.

Hard Drive Reliability Factors

To maximize capacity, storage density has to be balanced against reliability. Some factors affecting HDD reliability include:

  • Electromechanical components wearing out.
  • Degraded magnetics on platter surfaces.
  • Vibration interference during seeks.
  • Coping with dust particles and debris.
  • Maintaining head and platter clearances.

To improve reliability, technologies like the following help:

  • Fault tolerant signals and error correction coding.
  • High precision actuators and head controllers.
  • Sensors to detect impending failures.
  • Algorithms to remap bad sectors.

There are always tradeoffs between push density higher and preserving data integrity and drive lifespan. Manufacturers rigorously test and validate drives to hit specific reliability targets.

Perpendicular vs Shingled Magnetic Recording

Perpendicular recording is generally superior to shingled magnetic recording for high capacity hard drives thanks to advantages like:

  • Simpler, more reliable writing of overlapping tracks.
  • Maintains performance by avoiding track rewrites.
  • Not limited by write serialization issues.
  • More consistent readback signal-to-noise ratio.

However, SMR can reach slightly higher densities in principle by overlapping tracks.

SMR may be preferable in archival data storage applications where rewriting is very infrequent. But for general-purpose drives, PMR avoids performance penalties from shingling complexities.

So PMR strikes a better overall balance between density gains, performance and reliability for most usage cases.

Summary

There are a variety of techniques available for continuing to scale up the storage densities and capacities of hard disk drives. However, we are approaching the fundamental limits of what can be achieved with current perpendicular magnetic recording approaches.

To maximize capacities further, technologies like bit patterned media, energy assisted recording and two dimensional magnetic recording will be needed. These bring their own engineering challenges around fabrication, costs and reliability.

Integration of higher-density platters, head actuators and emerging non-volatile memories can also expand hard drive capacities. Larger form factors allow greater storage potentials assuming issues like vibration and heat buildup can be controlled.

There are inevitable tradeoffs between pushing areal densities versus operational reliability and lifespan. Capacity increases cannot compromise data integrity and drive usability.

Hard drive evolution will likely focus on hybridizing with solid state memories and specialized form factors for different usage models rather than just maximizing magnetic recording bit densities.

By combining incremental advances across multiple facets like heads, media, integration and form factors – there is still potential to sustain the historical growth in capacities and remain competitive with flash, tape and optical storage alternatives. But fundamental changes will be needed beyond just improving existing perpendicular recording approaches.