Solid State Storage: How It Works
Posted: July 25, 2000
Written by: Tuan "Solace" Nguyen
Introduction
Welcome back to the next installment of the How It Works series. Recently, we’ve been covering storage technologies. We started out with hard drives then went into optical technology, and then holographic storage. What could we possibly come up with next? I bet some of you right now are thinking “solid-state” technology. Well, holographic storage is a type of solid-state technology. But the most common one would be a hard drive made entirely of RAM modules.
Let’s begin with solid-state hard drives.
Why Solid-state?
Why build extremely expensive equipment when 7,200RPM hard drives are dropping in price and offer large capacities? Let’s take a look at an analogy for a second. If you were to walk to the can to do stuff (heh), wouldn’t that take longer than if you could do stuff at the can while sitting where you are? If information can be accessed without having to move around to find it, then obviously data access time is greatly reduced.
The above diagram illustrates the massive speed advantage that a solid-state hard drive has over a conventional hard drive. You can expect performance gains that are 100 to 200 times that of today’s fastest Ultra2 SCSI hard drives, and those drives are already screaming fast.
Speed is only one of the advantages of using solid-state technology. Another advantage is reliability and durability. With solid-states drives you don’t have to worry about vibration and shock. Because there are no moving components, vibrations won’t affect the device.
Solid-state Hard Drives
The following picture illustrates the basic components of a typical solid-state hard drive.
This type of hard drive is made up entirely of RAM modules. They are stacked and connected in an array. It uses standard SDRAM modules to store information. But if you remember or know, SDRAM is volatile memory, meaning it loses its information when there is no power. That is why there is a rechargeable lithium-ion battery that sits in the back of the casing powering the drive when AC power is unavailable.
Speed
A high-end conventional hard drive has an I/O rate of about 60 to 90 IO’s/sec. A solid-state disk in comparison can range from 5000 to 9000 IO’s/sec.
A high-end conventional hard drive has an average access time of 6 to 8 milliseconds. A solid-state disk in comparison has an average access time of 50 to 100 microseconds.
As you can see, the benefits of having solid-state technology become more and more apparent as you look at the speed advantages it brings. Applications like video editing and mission critical issues can take advantage of the speed benefits.
Click for an actual size image.
As you can see, the drive is basically just made up of many SDRAM modules. The big black thing at the back would be the supporting battery.
Solid-state hard drives are incredible performers, but have one glaring problem -- price. The prices for SDRAM modules are still much more expensive than a conventional hard drive megabyte for megabyte.
This type of technology has been out for a long time. They exist in PCMCIA cards, flash cards, etc. The Diamond Rio and Creative Labs Nomad would be two common examples of commercially available solid-state products.
Reaching The Limits
Currently, hard drives have grown in capacity at a phenomenal rate. When SCSI was the king of capacity and IDE was left in the dust, hard drive prices were high. Now, you can find IDE drives ranging from 1GB to 75GB that spin at 7,200 RPM. If you have not read our How Hard Drives Works guide, hard drives increase in capacity by increasing areal density. Meaning they crease the number of bits that can be packed into a certain area on the platter. The more bits you can store in a smaller area, the more data you can pack.
But what happens when the physical limits of hard drives are reached? You can switch over to other technologies such as holographic technology, but it’s not available and won’t be for a few more years. Solid-state hard drives are still not a viable solution because RAM prices are still very high and optical drives are still not as fast as magnetic drives. Doesn’t that pretty much eliminate all of our options? It pretty much does. So what do we do to feed our hunger for more storage capacity? We go smaller. How small?
Molecular Memory Storage Technology
What if we could store bits of information at the molecular level? What would it be like if we could store massive amounts of information in chunks of molecules? If you haven’t read my article about holographic storage technology, you should zip by and read that; as it helps to understand the following information about molecular memory.
How It Works
How this technology works is relatively simple. This technology really is very similar to holographic storage techniques. What you have going is an LCD array. Which is basically a screen, like your laptop screen, which has cells that light up when electricity is passed through it. Now, all the LCD’s are turned on, which means that they do not let light pass through them. Think of it as the LCD screen on your watch. There is a low powered laser behind the array, which is lit. Now, the laser cannot pass through with the array on. When certain cells on the LCD array turn off, the laser will pass through those inactive cells. Once they pass through, they will strike a data page, which is sensitive to the laser.
One the laser hits the data page, the sections that were struck by the laser is in a state of Q. Q state denotes a binary value of 1. The hII state denotes a binary value of 0, which were parts that were not struck by the laser.
The data is now recorded on the data page. There could be many data pages stuck together. So how does the laser strike a specific data page? Well, there is a yellow low intensity laser that lights up a specific data page setting it to recording mode, or 0 state. While the other one’s remains in idle state, which the write laser doesn’t affect.
Molecular Memory Storage Technology (cont.)
To read from the data page, the yellow laser sets the data page to a state of Q, which permits a low intensity red laser to penetrate it. The laser then fires at the data page and passes through everything but the parts that were marked. The cells, which have no laser light passing through them, show up on a CID detector which reads the pattern of on and off cells of the data page, thus, representing ones and zeros.
Image courtesy of Scientific American
So what’s keeping this great technology from coming out to the public? Cost and product. Currently, the product costs for this technology is as expensive as holographic technology. And an economical production unit that’s usable in size is still a long way off.
What now you say? What is left for the power hungry and storage greedy? More unavailable technology! That’s what! I know, I’m really saturating you readers with all these high-tech and cool technologies and yet, they are unavailable to you. Why exactly am I doing this? No, I’m not evil, but I would like to inform all you readers of what’s ahead of us (hopefully) and where computers will go in the future. So, noting that, let’s continue.
Atomic Resolution Storage Technology
What do you think of storing a few gigabytes in the space that’s about the size of your thumb? This is the promise of atomic resolution technology. Why is it called that? Because each bit of data is recorded in a very tiny space. So tiny that each bit is the size of a single atom, or a few atoms. But then again, if it’s too many atoms, then it’ll be a molecule, and would be molecular memory technology. Which, this is not.
Image courtesy of Scientific American
What It Will Offer
Atomic resolution technology will enable us to store vast amounts of information in a very tight space. It will offer storage capacities that are unheard of and are limited by imagination, and there will be some physical limitations. Electron beams from an array of probes with write heads that are the size of atoms write data onto the storage medium (phase change medium) by heating tiny data spots on the medium and changing its physical state or phase. Under the array, the medium is moved with nanometer precision.
Currently, Hewlett Packard is developing the technology (ARS or Atomic Resolution Storage) and is looking at ways to commercialize the devices. Currently HP has achieved densities in the terabits per thumbnail size mediums.
Atomic Resolution Storage Technology (cont.)
Challenges
So now that we can achieve insane storage capacity, what’s stopping us from using the technology?
The first problem is finding a suitable medium to record on. It must stay stable at room temperature and its state must not change when the temperature varies by a few degrees or more. And the medium must be able to phase change fast enough.
The second problem is the write head that must emit an accurate electron beam when power is induced. The write head must be able to accurately heat up a spot on the recording medium to either write or erase a bit. Using a low powered beam will let the head read the bit data.
The write head reads the bit by sensing the bit's electrical resistance. And if it is high, then a 1 is denoted while a low resistance says it is a 0. HP is also looking into using optical pickup devices to read the bits instead of switching between low and high current to sense the medium.
Image courtesy of Scientific American
The third problem is developing the reading mechanism to be extremely accurate when moving. Obviously, this makes it less “solid-state”. However, its movement is so small that you can call it solid-state. HP needs to develop the mechanism to be able to move with nano-accuracy. A slight nano-meter off and the whole read/write process is destroyed.
The final challenge to overcome is packaging. The device chassis needs to be extremely rugged and must be able to withstand shocks and vibrations of varying degree. The device must also be sealed in a vacuum package because anything can disturb the flow of electrons. The reason for this is the same as for a monitor being in a vacuum.
Conclusion
All these technologies are great for the future. For now however, conventional hard drives will still lead the way. And more promising technologies such as serial ATA and faster SCSI technologies will still offer us more and more storage space and/or faster access speeds. And besides, what will you do with a drive that can store a few terabytes? I doubt you can really fill it (well, some day...). And besides, defragging will take eons. ;)
Well, I hope this guide has given you a glimpse to the future of storage technology. No longer will we be tied down by a few gigabytes. But that is still a few years off. However, the technology is very promising. Right now, a few other technologies are in development. DVD-RAM and DVD-R drives will become mainstream just like CD-R and RW has. And the speed will grow, and writing DVD’s will become a norm. That too, will take some time. I wonder sometimes if anyone is still using the 1.44MB floppy disk. CD-R disks are so cheap these days it makes no sense not to get a writer. I know, not all have the money, but there are many writers out there with varying speed and it’s really worth the purchase. Besides, if you want to tweak your system, you’ll need space for all those tweaking utilities you’ll be downloading. And once you get broadband (if you don’t already have it) you’ll definitely need a burner.
Well, I’ll be writing more on storage technology (such as double layered CDs) and other guides, and more are sure to come. Until then.
Want to return to the normal guide? Click here!
All Content Copyright ©Dan Kennedy; 1998-2000
