\
Computers / Buying Guide

Storage and Hard Drives 101

         

One of the most important features of computing is one of the least understood. A hard drive is at the core of any computer, and yet many consumers don’t realize the complexity of the different components that make up a hard drive. Let’s look inside the drive and explain some of the basics while we explore different types of drives and their uses.

External versus Internal Drives

First of all, if you want to store data, you need a hard drive. Internal hard drives are drives that fit inside a desktop computer, laptop, enclosure, or server array—basically, they are made to be enclosed in a computer or server. If you’re looking to replace a hard drive “inside” a desktop computer or laptop, you need to find an internal hard drive. Some people use servers to store multiple internal hard drives, usually for long-haul business applications. We’ll talk about internal hard drives used inside of enclosures or NAS arrays in a little bit.

External hard drives work a little differently. An external hard drive is usually designed for portability. Most slimmed-down Ultrabook computers these days feature limited hard-drive space. An portable external hard drive can give you the extra storage space you so sorely need, especially when accessing photo, video, or audio files. Portable external drives can easily fit into cases, bags, and backpacks. Some external drives (called Flash drives or thumb drives) can even fit in your pocket or on a keychain. These Flash drives have a much smaller capacity (128GB is the norm, although some makers are reportedly breaking the 1TB barrier) but are far more portable.

Type and Mechanics

So what makes a good hard drive? And what are the differences between a hard drive, a Flash drive, and a solid-state drive?

Traditional Hard Drive

A traditional hard drive is made up of very thin platters of material, coated with a magnetic medium, which stores data in magnetic patterns. Each platter in a hard drive can store billions of bits of data, and three or more platters are stacked on top of each other. A spindle, which runs through the center of each platter, spins the platters at incredible speeds (anywhere from 5400 to 15,000 revolutions per minute) while a read/write head on each side of the platter reads data. The head must completely avoid touching the platters or the disks will crash, causing a loss of data. What does this mean to you? A platter moving at 5400 rpm is obviously not accessing the platters (and data) as quickly as one traveling at 7200 rpm.

What are the drawbacks of a traditional hard drive? The 15,000 rpm drives are usually reserved for business-application servers where speed and stability are of paramount importance. You also should take great care when operating any unit with a spindle drive. Drops and bumps may upset the spinning drive, causing a malfunction. Heat generation is always a problem, as is the weight—a traditional hard drive can add to any laptop computer’s weight.

Solid-State Drives

Solid-state drives (SSDs) represent a relatively new and increasingly popular technology. Why? Well, an SSD uses Flash memory, which means that information is stored in cells on the drive's surface, and is accessed by a printed circuit board (PCB). Breaking it down even further, there are no moving parts in a solid-state drive—and that means fewer opportunities for crashes. The absence of moving parts means the heat signature is greatly reduced on an SSD, which makes your Ultrabook™ a lot more comfortable on your lap than a traditional notebook. An SSD is also much lighter than a traditional spindle hard drive, noticeably decreasing your computer’s weight. 

Internal SSDs are available in many different shapes and sizes. The first is a 2.5" SSD, typically found in laptops. Then there is the super-slim 2.5" SATA SSD, which is thinner than the previously mentioned 2.5" SSD. Since Ultrabooks are even thinner than notebooks, they typically use mSATA SSDs, which are smaller than both the previously mentioned form factors. If you want to go even smaller, there is the M.2 form factor that makes even the mSATA SSD look gigantic.

PCIe Solid-State Drives

Push your workstation for speed and performance with a PCIe solid-state drive (PCIe SSD). With a PCIe SSD, you'll skip the limitations of SATA and connect through a PCIe slot. Since PCIe offers point-to-point architecture, it allows for faster data-transfer rates. This makes PCIe SSDs useful for high-performance tasks, such as 4K video editing. PCIe SSDs can be found in Apple Mac Pros and HP Z Workstations

Hybrid Drives

The newest kids on the storage block are hybrid hard drives. These drives meld the capacity of spindle drives and the speed of SSDs. How? They include a small SSD, incorporated into the physical hard drive. This smaller SSD is usually used for caching, which allows you to access your most commonly used programs, or those that you designate as the programs you need to access the fastest. The cache stores information so that you don’t have to waste time spinning the spindle looking for it. It cuts down on boot times and program startup times.

USB Flash Drives

Another storage medium is the USB flash drive. These small, inconspicuous drives work along the same principles as SSD drives.

They also use cell technology to erase and write data with something called a "floating gate" transistor. This transistor has two “gates” or on/off switches that transform electrical signals into binary 00s and 01s. Large-size (64GB and greater) thumb drives have become extremely affordable. The downside is that flash drives, although infinitely useful, don’t provide the large-capacity storage of hard drives or even solid-state drives. And their small size makes them very easy to lose. Also, if you’re using anything less than USB 3.0, the transfer speeds could be dismal.

Storage Capacities

So why the need for so much storage? As we move toward lives totally dominated by digital photos, high-resolution video, and high-end lossless audio files, it’s imperative to have room for all that information.

If you’re getting by, storing your whole digital life on the 16GB on your smartphone, then you don’t need this chart. But if you find yourself losing precious photos or cherished music because you have no storage space, this may interest you: (Click to enlarge)

Of course, this is just an approximation. File sizes—uncompressed video, RAW photo files, and lossless high-end audio files, compounded—will definitely take a bite out of your storage space. But you get the idea.

Storage Solutions

Note that not all internal hard drives have to be physically inside of your host computer. Internal hard drives can be sold separately and stored in an enclosure, which usually houses multiple hard drives. Unlike a server, which has the added functionality of hardware and software to serve your data to you, an enclosure is simply a shell that holds hard drives and can be attached to your host computer. You can buy internal hard drives separately from an enclosure or in kits that include the enclosure and hard drives. Some enclosures feature hard drives already built in; for flexibility, you want an enclosure that features hot-swappable hard drives, which means you can insert or remove hard drives at your leisure, while the computer system using them remains in operation.

There are also Network Attached Storage (NAS) arrays. You store hard drives in the mini-server, and that information is served back to you via your home or office network. The advantage of a NAS server is the amount of storage possible. By swapping out hard drives or adding to the NAS server (some can hold as many as 12 hard drives), your storage possibilities are endless. Even with a small home NAS server, you could easily store millions of photos, huge amounts of audio and video, and any other part of your digital life that is hanging around the house. NAS servers are great for people who share their digital files with others, as you can access your NAS server through an Internet connection. NAS systems can include room for both traditional hard drives and SSDs. When purchasing a NAS server, ask yourself these two important questions: what do I need it for and how much storage space do I need?

RAID

Another thing to consider when purchasing hard drives or NAS arrays is security. You want your information to be safe, you want the drives to be reliable, and you don’t want to come home and find that a critical drive has failed and you’ve lost all the information on it.

Let’s talk Redundant Array of Independent Disks, or RAID, for short. Put simply, RAID configurations allow you to protect data in several ways, across multiple hard drives.

RAID Level 0: Striping

Striping can make the data throughput of a drive faster, because it doesn’t duplicate any data, just pushes it through the pipe. All the disks perform at the same level, but your data is never backed up.

RAID Level 1: Mirroring

Mirroring does the same thing as striping, pushing the data through quickly, but it duplicates the data on another disk, so in the event of a drive failure you can always recover the data from the unaffected drive.

RAID Level 3

RAID Level 3 stripes the information across several drives, but uses one drive to store your main data. This allows you to rebuild after a data failure, but only if one drive fails (it rebuilds using “pieces” of data from the other drives. If more than one drive fails, however, you may lose all the data. Hardware acceleration is needed at the host machine.

RAID Level 5: Distributed Parity

Although similar to level 3, this level is best used for data distributed in small chunks. Large data flows may hamper performance and safety.

RAID Level 0+1 (also known as RAID 10)

A combination of RAID levels, this is the most costly, as well. One drive is mirrored across several alternate drives, and the information is safe, even in the case of drive failure—as long as the failed drive is not the mirrored drive. For example, you have a remote that controls several products. If a product fails, you can use the other products, but if the remote fails, nothing works.

Data Transfer Speeds

So you have all this data and a great big hard drive on which to store it. Now, how do you get that data from Point A to Point B?

Most internal drives, whether spindle hard or solid state, use a SATA interface to connect to the controller in your hardware. These interfaces determine the speed and age of your hard drive. SATA I drives have a bandwidth of 150 MBps, Bus Speed of 1500 MHz, and Signal Rate of 1.5 Gbps. This means that a slower SATA I interface in your computer can only transmit at those speeds. Newer SATA II is twice that, at 300 MBps, 3000 MHz bus speed, and 3.0 Gbps signal rate, while SATA III is four times as fast, with a 600 MBps, 6000 MHz bus speed, and 6.0 Gbps signal rate. This does not necessarily mean that a SATA III (sometimes called SATA 6 Gbps) transfer speed is 6 Gbps; after all, your personal configuration and the type and amount of data you try to move may bottleneck those speeds, but it does mean that it is capable of transferring at much faster speeds than its predecessors.

External drives transfer a little differently, using an external protocol like USB or Thunderbolt. USB 3.0 boasts transfer speeds of 5 Gbps and Thunderbolt doubles that, at 10 Gbps. 10 Gbps, theoretically, means you can transfer a full-length HD movie in less than 30 seconds.

We say, theoretically, because achieving the maximum output of any transfer protocol is always hit and miss. Anything can change the throughput of data, from files' sizes, software configurations, or even your computer’s processing speed or RAM. If you have a computer with Thunderbolt as an option, you should use that. Regardless of the speeds, it would be far better than using USB 2.0, which tops out at 480 Mbps. The chart below shows you the comparison between the different speed throughputs of various transfer protocols.

So now you know all about hard drives, but do you know what to look for when buying one? Are you buying one as a gift or for your own home use? Do you need lots of storage, or just a supplemental storage unit for an already crowded laptop or desktop computer? Do you want access across your network, or just a plug-and-play device? When you can answer those questions, you’re on your way to making a great hardware purchase.

For more information on internal, external, and solid-state drives, stop by the B&H SuperStore in New York, speak with a sales professional on the telephone at 1-800-606-6969 or contact us online via Live Chat.
 

Discussion 16

Add new comment

Add comment Cancel

Thanks for what seems to me a very cursary overview of this complex topic, at leat for those of us who do not spend every day and night on it.

Pehaps it would have not been a proper subject, or the proper venue, but I would have liked to have seen some recommended manufactueres products that are reliable and a good buy, as well as those that best to stay away from.  I have a Western Digital BLUE 500gb that had a phsyical crash, and was only in service for just over a year.  The recovery service that has looked at it was not surprised, and said these drives have a LOT of failures.  I guess good for their business, but bad for me.  I guess they will be able to rebuild the file allocation table, but only about 90% worth.  I have an automatic backup and recovery application provided by the computer manufacture (HP), but it was a second letter drive on the same physical drive.  I am HOPING, I'll be able to put my whole drive back together with that, but I need to get the money together for them to restart the recovery process.  I would have gladly spent more money on another drive that would have proven more fuctional and more importantly, reliable!!  I'd welcome any recommendations B&H would offer, as I know that they would be fair ones. 

I've heard the opinion in the past that flash drives were not good for long term storage, yet here, flah drives are the same as SSDs, only smaller. 

A 64G flash drive can hold a LOT of information, so are there any facts to explain why it wouldn't be a good choice for long term backup and archiving?

In the past it may have been true that Solid State Drives were not good for long term use. This is no longer the case, and most reports show that they have a longer life than traditional drives. Flash drives can be used for part of your backup, but always try to follow the three two one rule. Have three copies of any critical data, on two different types of storage, and keep one copy offsite.

Where does the eSATA come in here? I have a 4x eSATA3 PCI card connecting 4 separate external SATA3 discs including two as RAID 0. The RAID 0 array scores highest among all the mechanical-external drives as one would expect.

Another issue is the USB 3.0 interface. First, I've always had problems getting computer 3.0 ports to run as 3.0. Second, putting the USB 3.0 driver-computer port issue aside, I've never been able to achieve transfer rates even close to eSATA3 or SATA3 speeds. Now i cannot find any new computers with eSATA ports; as fi USB 3.0 makes them redundant. Hence the necessity of the PCI card.

The eSATA interface has a maximum speed of 3Gbps. (Not to be confused with 3GB/s) When connecting a dive to any interface you will never see transfer speeds equal to that of the connection. For example USB 3.0 has a maximum speed of 5,000Mbps (625MB/s) The average spinning USB 3.0 drive will most likely be in the 120-170MB/s range. This isn’t a limitation of the USB 3.0 port, but of the drive.

You should also add that solid state drives are great for speed ,  however have a limited read / write life.  Most affordable solid state drives are MLC  , where as SLC is one of the most expensive.  Your typical Samsung Evo 850 is MLC based. A 64gb SLC server drive will run you $300 to $500 which isn't cost effective. People burn out MLC drives thinking they can be used for heavy read and write storage.  Ex. Video editing.  When dealing with solid states , it's always good to have some form of health management software so you can keep an eye on the condition of your drives. 

Would Obamacare work for this issue??

Excelente articulo. Muy buen trabajo.

Your discussion of RAID 10 is off a little. There are two ways do implement what is known as RAID 10. The Dell way is to have multiple disks (an array) and RAID 0 them, then create another RAID 0 array and Mirror the two arrays into a single RAID 0+1 structure. The bad thing about this arrangement is that if you lose a hard drive the RAID 0 part is broken and your risk of a failure goes up.

The HP way is to RAID 1 two drives for each segment of the overall array. Once you have a set of Mirrored disks you concatinate them together resulting in a RAID 1+0 configuration. You can lose up to half of the drives in the array and keep running as long as the original mirrored pairs do not suffer a simultainious failure.

On enterprise systems RAID is used for fault tollerance, disk hardware failures, capacity, ie very large volumes and performance. RAID 0 uses a minimum of 2 drives (we called them spindles). RAID 0 uses a minimum of 2 drives. RAID 5 uses a minimum of 3 drives. RAID 6 uses a minimum of 5 drives.

RAID configurations are not for backup. If you hit the delete key on a RAIDed array it will dissapear long before you can hit ctrl^C or escape, halt whatever. Remember too, there is no recycle bin on a NAS, SAN or other networked storage solution. My main system uses two RAID 1 volumes of 1TB and 5.5TB respectively.

Background: System Administrator for 20+ years manageing 7-300+ servers by myself over the years (member of a IT staff that managed 1000's of machines) Worked for two Fortune 50 Global Corporations with my last position supporting 28 clusters with just under 5PB of SAN storage.

These "graphs" (really illustrations) are very beautiful, but extremely confusing, unintuitive and wacky.

For example, in the "Storage Capacity" image, the color of the hexagons follows absolutely no progression (ie cooler colors for smaller drives, warmer colors as capacity increases), in fact the largest and smallest are nearly the same color when they should be opposites. The labels for the hexagons in the center of the page are in their respective colors, making the darker ones extremely difficult to read against a black background. In the "Data Transfer Speed" image, the metaphor is that of a speedometer ..except that for no explicable reason, this speedometer operates *counter-clockwise*.

It's great to make nice-looking images, but when the point of the material is to instruct and clarify, it's important that they function properly before dressing them up.

Excellent, concise explanations, bravo B&H.....

Great article! That's for providing it!

I don't know who designed the two charts displayed in this article, but I must take exception to them, as they are highly misleading.
In the first chart, concerning storage capacity, the most obvious issue is with the 1000 GB element compared to the 2000 GB element: the 2000GB part should be twice the area of the 1000 BG part, not four or five times the area. In this case, the chart graphically exaggerates the differences among the elements.
In the second case, concerning data transfer rates, the opposite is true: the chart significantly understates the differences among the elements. The scale is not linear, so you have 1 Gbps appearing to be half as fast as 10 Gbps instead of being one-tenth as fast, as the case truly is.
It's probably naive of me to suggest this, but I would like to see the charts revised so they graphically depict the relationships among the graph elements accurately.

Why don't you make the new chart and post it in the comment sections for all to see.

Thanks for the heads up, and I'll be looking forward to that new chart.

links within the discussion of RAID to tutorials on implementing the different levels of RAID would make this discussion more helpful.

seriously? how lazy are you?

Close

Close

Close