Home
About
Help
Feedback
|
The Harrow Technology Report
Insight,
analysis, and commentary on the
Copyright © 2001-2005, Jeffrey R. Harrow. All rights
reserved.
YOU Speak Out.
Listen to this Issue.
Copper Is So,
Well, 20th Century!
There's MUCH More I Can Do For You!
Light Tweezers,
Redux.
Backup Insights,
Revisited.
Our Final Words on
'Digital vs. Analog.'
New Cheap Labor! Schedule Note.
The next issue of The Harrow Technology Report will publish on May 26, 2003. Enjoy Spring!
Listen to this Issue.Do you prefer to let your ears do the work of keeping you in-touch with, and thinking about where technology is taking us? If so, "The Harrow Technology Report" is also available in an audio-on-demand, Web-based, MP3 version. If you have an MP3 player on your system (and most do, such as Window's Media Player, RealPlayer, etc.), clicking on the link below will either stream the file to you, or, depending on how your system is configured, it might download the file before playing it. Alternatively, if you specifically want to download the file, simply right-click on the link, and choose "Save Target As..." Also, to learn how you can listen at whatever speed is most comfortable to you, check out the FAQ at http://www.theharrowgroup.com/help.htm . So, if you wish, just click on the following link to listen to this issue! http://www.theharrowgroup.com/articles/20030505/20030505.mp3 .
Copper Is So, Well, 20th Century!
Even though they might seem so,
today's integrated circuits are not single
monolithic devices that magically provide memory,
calculations, interconnect functions, and more.
Instead, each chip can contain tens of millions of
different devices (transistors and more) that work
together to perform the intended tasks. (As
examples, a contemporary 3.06 GHz Pentium 4 chip
contains 77 million transistors, while News.com
suggests that this will hit the half-billion mark by
the end of this year -
http://news.com.com/2100-1001-961210.html).
And it's necessary to wire all those devices
together! For example, the outputs of some
transistors have to be routed to the inputs of
others to create logic gates. Logic gates in turn
have to be wired together to provide higher level
functions. And the multiple "wedding cake" layers
of circuits that form a chip have to be wired
together as well. On today's chips, that takes a
lot of "interconnect wiring."
"The new process includes 'growing' microscopic, whisker-like carbon nanotubes on the surface of a silicon wafer by means of a chemical process. Researchers deposit a layer of silica over the nanotubes grown on the chip to fill the spaces between the tubes. Then the surface is polished flat." According to Meyya Meyyappan, director of NASA's Center for Nanotechnology, "Roadblocks exist in several common technologies such as interconnects, lithography and others currently used to make the chips. However, I think our new process could be in use by industry for the next generation of ICs, removing some of these roadblocks." This is just one innovation
that address but one of the roadblocks that will
continue to appear as we unfold the onion-like
wrappings of the billionths-of-a-meter world.
(Consider that, following a different route, IBM
demonstrated back in 2001 that they could build an
experimental transistor entirely out of
semiconducting nanotubes -
http://www.research.ibm.com/resources/news/
And especially considering that carbon nanotubes were discovered by accident (scientist Sumio Iijima decided to explore the soot-like residue left by an experiment - http://www.personal.rdg.ac.uk/~scsharip/tubes.htm#history), and that we're just beginning to understand the unique ways that carbon nanotubes can be put to use, this is another great example of the power of innovative curiosity. Do question the world around you. There's a tremendous amount for us to yet learn!
Begin Self-Serving Advertisement There's MUCH More I Can Do For You!
You may not realize it, but there's much more to The Harrow Group than just "The Harrow Technology Report." For almost twenty years, as I've been sharing my research on the ever-faster-moving and converging technologies that are changing how we work, live, and play, I've also been working directly with businesses and organizations, large and small, to help them understand and address how these changes may affect them, their customers, and their customers' businesses, through a series of:
·
Presentations -
Highly engaging, interactive, multimedia,
constantly-updated presentations and keynote
speeches to individual businesses, internal groups,
and trade organizations, helping participants to
viscerally understand and appreciate how technology has
brought us to where we are today, and where it's
likely to lead us tomorrow.
·
Workshops
- Beginning with the presentation described above
(to give all participants a common understanding and
insight), the workshop further engages attendees to
explore how this march of technology might affect
their individual businesses and organizations, and
their specific needs.
·
One-On-One Consulting -
Individualized
consulting services, available via phone or
in-person, to help you explore the topics and trends
discussed in The Harrow Technology Report,
and related issues. Please continue at http://www.theharrowgroup.com/consulting2.htm for additional information. Then, contact me at Jeff@TheHarrowGroup.com with any additional questions, to discuss fees, and to schedule a consulting event. I look forward to working with you! End Self-Serving Advertisement Light Tweezers, Redux.
In the last issue, we explored
a novel use of light that had been processed in
unusual ways -- as physical tweezers that could move
nano-sized things around and even spin them like a
top (http://www.theharrowgroup.com/articles/
"This is actually nothing like a revolutionary rules change---the fact that photons exert force is not only taught at school, but can be demonstrated fairly easily with something I have seen in toy stores: a light bulb-shape evacuated glass bulb, containing a pivot-mounted 'propeller' with blades painted silver on one side and black on the other side. When you look at them from the side (perpendicular to the axis of rotation), the blade on the right of the axis is silver, whereas the incident light bounces off of it; the blade to the left of the axis is black, absorbing light. The explanation of this phenomenon is fairly simple, if you get down to it. Photons are radiating electromagnetic fields, and such fields can exert action at distance, obviously: vide electrostatic and magnetic force, as seen by owners of polyester pants and refrigerator magnets.
The way to describe that is: photons have an energy,
hv, and momentum, p=hv/c. If they strike the black
blade, they are absorbed, and therefore give their
momentum "p" to the blade. If they strike silver,
they bounce back, changing the momentum by 2p, and,
due to momentum conservation, imparting 2p to the
silver blade. So, the black blade is pushed one way,
and the silver blade is pushed twice as hard the
other way; it turns out that this is enough to cause
the propeller to rotate when light is shone onto
it." Alas, there's still contention in paradise, at least on the subject of those "toys." Geoff Canyon offers a different explanation in this paraphrase: "HEAT is involved in the toy with the four black-on-one-side, white-on-the-other-side vanes; it depends on the vacuum in the bulb not being very good. In bright light the black side heats up more than the white side, thus the few air molecules that are left bounce off more energetically, pushing the vane. This is demonstrated by putting one of the four-vane toys, minus its bulb, in a high-vacuum chamber. Shine light on it, and there's no motion. The other thing to note about the four-vane toy is that if it were light pressure causing the movement, then the vanes would spin AWAY from the white side, since white reflects and black absorbs. But they spin away from the black side..." On the
other hand, Adrian Jackson points us to the
"Physics FAQ" explanation of the "light mill" (their
name for this "toy," which is actually a type of
radiometer), hosted by the University of California,
Riverside, at
http://math.ucr.edu/home/baez/ As we close the door on this discussion, these responses (and the MANY more I've received but can't fit into this issue) show that our collective knowledge is quite something; we'll certainly tap into it again in future issues. Everyone may not quite agree, but we've all learned a lot! My thanks to ALL who
participated and increased our overall knowledge! Backup Insights, Revisited.
Last issue, we continued
exploring the world of Backup strategies, and in the
process of explaining why I still used tape backup
as one prong of my Backup fork
(http://www.theharrowgroup.com/ "One thing that's missing from this image backup strategy (and one reason I continue to use a file-based incremental tape backup as well) is that, because of the size of the image files vs. the speed of my tape drive, I can't (reasonably) write the image files off to tape for off-site storage. I could write the image backup files directly to an external disk drive (connected via USB 2.0 or FireWire), but they're rather expensive for an off-site backup rotation schedule." One reason that I considered external hard drives too expensive was that the least expensive 100 - 120 gigabyte external USB or FireWire drive costs about $260 (and goes up considerably from there). Additionally, although I didn't mention it, there's also the issue that external hard drives, nestled in their protective cases, are fairly large. If your off-site storage space is limited as in a fire safe, this could be another issue if your backup data exceeds a few drives. But reader Dave LeVine
reminds us that "case-less" (no external case)
removable hard drives, which slip into and out of a
tray in a removable drive bay, are another, less
expensive (and sometimes physically smaller!) backup
media alternative to tape (see
http://www.pcconnection.com/scripts/ "Actually, Jeff, you hit the nail on the head but had it upside down. You indicated that backup to disk is too expensive to store off site. I disagree.
A 40 GB (compressed) DLT-IV tape costs about $40.00, and its tape drive costs at least $550.00.
Exabyte VXA-1 tapes (66 GB compressed) cost about $40.00, and its tape drives cost more than $500.00. Assuming one backup per disk (and that is not the most efficient approach), each 100 GB disk holds 200 GB of data (compressed at typical 50%) and costs about the same as tape on a $/GB basis. Add in a "mobile rack" (for under $50.00, which allows "tray-mounted" hard disk drives to slide in and out of a removable drive bay "rack"), and put each drive in a tray for the "rack" (for about $15.00 each) and the results are enlightening -- with 8 hard drives and related equipment, the cost is LESS than for tape storage's 20 tapes plus their tape drive! Cost Comparison - 1.6 Terabytes Of Off-Site Tape & Tray-Mounted Hard Disk Storage
I admit that the controllers were not included (IDE is built in, SCSI is seldom free) and speed was not addressed (even 5400 RPM drives are much faster than tape) and I didn't use a really high end "mobile rack" to maximize its cost, but I didn't use list on the drives either. Can you afford $120 for an off site backup? Actually, after you consider the cost of your time to restore, I think disks are cheaper than many tapes." On the "space" issue of offsite storage, Dave offers:
"Consider a 240 GB (uncompressed) [today's
higher-capacity but more expensive] disk drive
versus the equivalent number of tapes (in different
common formats) that would be required to store that
same 240 GB of uncompressed data: DLT tape format
would require about 6 tapes; Exabyte "Mammoth-2"
tape format would require 4 tapes; "DDS-4" would use
up 12 tapes; and that same 240 GB could be stored on
50 DVD-Rs (or on 342 CD-Rs). Want to try that size
equation again? The Mammoth tapes MIGHT be smaller
than a hard disk in a tray, but a box of DDS tapes
is bigger, DLT tapes are huge by comparison, and we
all know the size of a spindle of 50 DVD-Rs. Removable hard disks are less rugged than DVD and CD discs, which will survive 500 G shocks easily, but that is the ONLY drawback to removable hard drives that I have found. And who cares if the disk "only" runs at 5400 RPM and PIO mode 4, it's still LOTS faster than any tape I have seen." Dave certainly makes some persuasive arguments. There's no question that hard disk backup is far faster than writing to tape. And as Dave so clearly describes, the dramatic and continuous advances in hard drive capacity vs. cost per gigabyte have indeed tilted the cost equation. Similarly, depending on your storage needs, the space needed for off-site storage of removable tray-mounted hard disk drives may also cross-over from tape. (But note that most tray-mounted removable hard disk drives are NOT "hot-swappable," as are USB and FireWire external hard drives, as brought to our attention by reader Perry Bebbington - you usually have to shut down the system prior to removing or installing an inexpensive tray-mounted hard drive.) As you consider which technology might be best for you, do consider your backup needs: On the "small" end of the scale, one small business backup scenario I'm familiar with fits its daily backup nicely onto one 30-gigabyte (compressed) OnStream DI-30 tape cartridge, which is significantly smaller than a hard disk. Yet if their storage needs grow to spill over onto several tapes, the hard disk would then win on size. Moving up the business size scale, to where up to 240 gigabytes of backup storage is needed, the removable hard drive backup solution would seem ideal. But when their storage needs grow beyond the capacity of a single hard drive, in order to keep the process automated without the need for an attendant, they will either need multiple removable hard drive bays (so that the backup software can extend onto other waiting volumes), or a "tape library" system that can automatically feed a sufficient number of blank tape cartridges or other storage media into the drive as needed. (Of course with this amount of backup data, the typically slower speed of tape could make this infeasible -- without special procedures, it would be hard to make daily backups during the night if the backup process took longer than one night! There are specialized backup solutions for such needs, but they're outside the scope of this discussion.) To take this from the realm of the abstract to the real, reader Carl Taylor explains how he actually uses such a tray-mounted configuration to safeguard his work: "I have both my C and D drives in the "drawers" and normally run off of "C:". At backup time, I put a blank drive in the D drawer and clone C to D. There are dozens of programs to do this -- I use a DOS batch file with a whole bunch of switches to do the job... At crash time, no recovery procedure! Just change the jumper on the cloned drive to make it the master, slide out the crashed C drive, slide in the cloned drive (now configured to be "C:"), and you are back in business. The backup drive is ALWAYS pulled out of the "drawer" an inch or so -- that way nothing can harm it..., can harm it..., can harm it... ( I am sure you remember that joke! ) I
love the system... Cheap, easy and I think
foolproof... (Hummm... do you hear Murphy laughing
in the background?") Additionally, Jeremy Winston among others talked about the advantages of RAID (Redundant Array of Inexpensive Disks - see explanations of the eleven different Levels of RAID at http://www.acnc.com/04_01_00.html), as a way to assure that one hard disk mechanical failure doesn't mean "the end of (data) life as we know it:" Note that while RAID does indeed "back up," it does so in a very different manner than traditional backups, with both advantages and disadvantages, which we'll discuss below:
"Given the ever decreasing cost of ever increasing
disk capacity, another backup strategy to consider
is disk mirroring (RAID-1). I have a pair of 80GB
drives from Western Digital plugged into a Promise
FastTrak RAID controller built in to my Soyo Dragon
Plus motherboard. They are configured to mirror
each other, so that if one breaks down, the system
automatically shifts to the other one, and once a
replacement drive is installed, automatically
restores the mirror. RAID does have great value in certain circumstances, in that it makes it highly unlikely that a physical failure of one disk drive will cause you to lose data. But even with RAID, traditional backups DO remain necessary since RAID does not protect against "program errors" (such as a database that decides to corrupt itself through a bug in its code, since that will be replicated to both RAID disks), or the "Oh, DARN! (or similar)" type of error where you know, instantly but too late, that you'd REALLY like to have been able to undo your last pressing of the "Enter" key. But do consider the combination of RAID -PLUS- traditional backups -- that can be a Very Good Thing! Finally, for this subject, one last paraphrase -- an excellent reminder from reader James Ronholm that I should have mentioned earlier in these backup discussions: "Backup procedures by themselves are useless. It's the RESTORE procedures that are essential. Make sure that you practice the restore procedure at least once to find the bugs, BEFORE it is critical." He's quite right -- test the ENTIRE end-to-end process - both the backup AND the restore. You'll rest easier. MUCH easier...
Our Final Words on 'Digital vs. Analog.'
Finishing up our recent
discussions around the differences between digital
and analog music recording and processing
(http://www.theharrowgroup.com/ "Hello Jeff, I've been subscribing to your newsletter for 2 or 3 years now, and I still enjoy it immensely. Though the moment may have passed, I'd like to add further insight on the issue of analog vs. digital quality in sound recordings. I make my living as a musician, and I can tell you that despite the proliferation of advanced digital technology in today's professional recording studios (and "basement" studios), hard drives still give way to magnetic tape when recording certain instruments. Drums, in particular, are commonly recorded on tape before being transferred to digital format for editing purposes. The analog recordings are just more true, something sound engineers have always known. Straight-to-digital drum recordings sound very harsh and brittle, and are also very unforgiving on errant dB levels (an extra loud cymbal crash or bass drum kick). Digital starts to distort very quickly when dB's are pushed past nominal, whereas analog can start to sound even better when the recording gets a little "hot". Under certain circumstances, some engineers will even transfer digitally recorded drums to tape and then back to digital to warm them up. Or they may patch a vacuum tube processor into the drum signal if going straight to disk. The same is sometimes true for other instruments as well. So, many of the latest recordings made at today's most advanced studios still originate on reel to reel tape to capture the "true" sound of the instrument long before the general public plays it back "digitally"." On the other hand, reader Bill Hallahan takes umbrage that from a practical standpoint, our digital CDs sound worse than good vinyl: "I don't believe that records sound better than CD's. Now if some people actually like the clicks, pops, and hiss, and the poorer waveform reproduction that records produce, then this is a matter of personal taste. But I am referring to "true reproduction of the original audio" in my discussion below.
Some people reason that because CD's use digital
sampling, the reproduction is less accurate because
it results in staircase waveforms. However, this is
not true at all. Because the sampling is more than
twice the highest frequency in the band-limited
source audio, i.e. the sampling is done above the
Nyquist rate (see
http://whatis.techtarget.com/definition/ Recording errors on a brand new perfect record are 65 decibels down at most. On CD's they are 90 decibels down. Note that the output low-pass filter on CD player can have some phase distortion at the higher frequency cutoff range. But to eliminate this problem, virtually all modern CD players upsample to 88 KHz (from 44 KHz.) before filtering out the information above 20 KHz. I have compared the output waveforms of digital systems versus analog systems. The digital systems reproduce a waveform with much less error than any analog system, including tape. I recommend that people try the following test. Record a record on a CD, so that the CD has the same clicks, pops, and hiss. Then, have a friend switch between the record and the CD while they are both playing, and then try to identify which is playing (make sure the volume is set the same for both.) I bet that few people will be able to tell which is the CD and which is the record. I know of no person who has been able to pass that test when done with good digital equipment. I assure you that no human exists who can detect the difference between the record and the CD, in that case."
We won't go further into what
is both a technical and a personal preference issue
here, although I suggest that you read some
additional interesting commentary and insights on
this issue from Chris Rakoczy at
http://www.theharrowgroup.com/articles/ From all of these discussions, it seems reasonable that at the CD level of quality, the perceptible differences between high-quality analog vs. digital recording may be only on the edge of [most] humans' perception, and may be more an issue of "coloring" than technical fidelity. I'll be interested in tracking the results of similar comparisons and technical analysis made on the newer Super CD and DVD Audio formats as they become common, as well as on the "next" generation, whatever that turns out to be. Analog may seem "more true" to some (Nyquist aside), but digital may well soon become "good enough" for everyone. We shall see...
New Cheap Labor!
Finally, wouldn't it be nice for manufacturers to be able to co-opt millions of workers who have the ability to build things at the nano-scale? Would this be helpful, even if these workers took so many coffee (and other) breaks that they only built these things at the rate of "4,000ths of a millimeter per minute?" The folks at the Forestry and Forest Products Research Institute say "Yes!" They see this as a first step towards these slow workers becoming very helpful indeed as, according to the Oct. 8 Nature (http://www.nature.com/nsu/021007/021007-1.html), these tiny workers lay down nano-sized organic paths on a substrate in nano-precise patterns (which you can watch in a movie at the link above). And that could be most useful for nano-building and for tiny circuits. Oh -- and by the way -- these workers are individual trained bacteria (Acetobacter xylinum)! The neat and orderly paths these bacteria lay down are currently made of cellulose, which isn't too good a material for either chips or tiny MEMS machines, but the scientists are working to get the bugs out (so to speak) by altering the bacteria's genetic makeup so they lay down different, more useful molecules. Of course "trained bacteria" hold the potential for both good and ill, and it's far too early to see how this might play out. But it is an excellent example of how people thinking very much outside the box can produce some startling "inventions." I guess it IS possible to teach old bugs new tricks...
About "The Harrow Technology Report."
"The Harrow Technology Report" explores the innovations and trends of many contemporary and emerging technologies, and then draws some less than obvious connections between them, to help us each survive and prosper in the Knowledge Age. "The Harrow Technology Report" is brought to you by Jeffrey R. Harrow, Principal of The Harrow Group. http://www.TheHarrowGroup.com . Where To Find "The Harrow Technology Report:"
Copyright (c) 2001-2005, Jeffrey R. Harrow. All
rights reserved.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
