CHAPTER 4

The following few days I got settled in and filled out the SF-86 security form. That thing is electronic and asked for details about my past, which had been wiped from existence by The Rain. I asked Larry (no longer "Mr. Waterford," he said) about how it would be determined if I really ever existed or not since records and witnesses to my life had all been destroyed.

"Just fill out the forms and let Defense Security Services handle the rest," he told me. "They have had these difficulties since the meteors and have found ways to get information. Also, since The Rain, as you call it, they increased the requirements for polygraphs. Expect to take a lie detector test sometime in the next few weeks."

So that was that. I was on my way to a new career, with a security clearance, even. I was put in for a Top Secret and was told to read all the documents on the so-called AFSPSEC website. I found out the acronym stands for Air Force Special Programs Security Education Community. I was also told I should look up the National Industrial Security Program Operating Manual, the Director of Central Intelligence Directives, and a few other security documents that read like stereo instructions.

Finally, on my second Monday, I was set up with a workbench in a lab space and actually given a technical task to do. Larry brought me what appeared to be a small motherboard in a plastic static bag (and by small I mean about twice the size of a sticky-note) and told me to figure out what it was.

"Do you have any information on it?" I asked him.

"Sure I do. But I want to see what you come up with first." He smirked.

"When do you need this figured out? I mean how long do I have to tinker with it?"

"Take as long as you need," he said smugly. "But, uh, don't take too long." And he left it at that. How damned vague could you be?

How long was "too long" and how long was "as long as I needed"? This was some sort of test for the new guy I figured. So, I sat there for twenty minutes or more just rolling the thing over in my hands and looking at it before I came up with a plan of action. Since it was pushing four-thirty p.m., my plan of action was to go home and take Lazarus for a walk, which is just what I did.

The next morning I had a doctor's appointment with my new headshrinker. After talking with her for some time, she decided that I had developed a tolerance to the Zoloft and wanted me to try a new drug, I couldn't pronounce or spell the name, which was just out on the market. Fortunately, she had plenty samples of the drug and gave me a handful of them, since they were apparently expensive as hell. I had decent insurance now, but the deductible on brand name drugs was two hundred and fifty bucks and I sure didn't want to have to pay that.

So, I took one of the pills with lunch and headed into the base for work. That little motherboard was still sitting on my desk in the plastic baggy.

"Well, first things first," I told myself and carefully pulled the board out of the baggy. Then I took out my lab notebook and began making a diagram of the circuit. It took about three hours to get what I thought was a complete diagram for the thing drawn and each part that I could identify labeled. Then I tried to develop a block diagram of the circuit with the nonlabeled chips marked A through D; the rest of the components on the board were standard parts.

It was my guess that the board was a "onesy" and had possibly come from a multi-card chassis. My reasoning for that conclusion was that all of the leads were covered and as short as possible, and there were outlines in each corner of the board for some sort of mounting hardware. The short leads would help prevent interference to and from other boards mounted above and below (or beside) it in the chassis. From this assumption I drew a dotted line around my evolving block diagram to represent a chassis.

The block diagram consisted of all five "main" chips, A through E, represented as blocks of size proportional to those on the board. From the top left of the page there was a chip about an inch square, which appeared to be some sort of standard input/output (I/O) conversion chip (analog to digital/digital to analog). To the right of the I/O chip was a chip much larger, about one and a half inches across and three inches tall. I had no clue what that chip would do, but it looked similar to a main processor chip like a Pentium VI. I labeled it A.

To the right of chip A was a slightly smaller chip, about one inch square, and I labeled it B. To the right of chip B was an identical twin to chip A, which I labeled C. To the right of chip C was another I/O chip. The chip sequence I/O, A, B, C, I/O made up the top row of my block diagram. Also, all of the chips on this row were connected by a copper stripe on the circuit board. The I/O chips had what appeared to be standard serial outputs connected to them.

With the top line of the block diagram completed I then added chip D directly below chip B and then chip E directly below chip D. Chip D had copper leads to both chips A and C. There were also fiber optic cables connecting chips A to B, B to C, D to A, D to C, and E to D. Below and on each side of chip E there was a small power processing unit (PPU) board just like the type in a game console power supply unit (PSU). Each of these PPUs was in turn connected to a single transformer and fuse box that had a standard power cable input. There were also power connections between several pins on each of the chips and the PPUs and other standard components on the board. It looked to me as though there were two separate PPUs because the board was actually two systems in one. Perhaps the two identical processor-looking chips were parallel processors on a single board, or maybe one was a backup to the other. I didn't know.

By the time I had figured out the rough block diagram, it was six-thirty. I just knew Lazarus was going to kill me since I was late for his evening walk. I packed up my things, put the board in my bag, along with my notes, and headed out for the evening. There couldn't be any security risk or anything; hell, my clearance hadn't come through yet. So I knew this stuff was public stuff. No problem; if it had been classified they would never have let me see it.

Laz was damned happy to see me. No sooner than the door had opened did he jump up and lick my face. Of course, I dropped everything and cursed some. "Damnit Laz," I yelled. But then, it wasn't his fault I was late.

"I know buddy, sorry about that. You wanna go for a walk?" I tugged his left ear just above the white spot that met his neck. I grabbed his leash off the wall, a handful of smelly liver treats, and pocketed a couple of grocery store bags; then we were off, Lazarus pulling me the entire way. We strolled down the sidewalk to his favorite fire hydrant where he marked it as his. We walked for a good twenty minutes or so. It was my guess, in retrospect, that Lazarus was the only thing keeping me from having a heart attack. Since he was old enough, we have walked every day for about twenty or thirty minutes. I'll bet he has caused me to lose about ten pounds, not that I couldn't stand to lose another twenty-five or more.

We finished the walk and I took another one of my new pills. I realized that I hadn't cried all day. Was it the fact that I was preoccupied with the new circuit board or was it the new drugs? I didn't really give a shit as long as I quit crying all the time.

Lazarus and I played around a bit and then I threw together a sandwich. I was eager to get back to working on that circuit board, so I booted up my system and ate at my computer station. My block diagram was too primitive and paper is hard to manipulate, so I "CADed" it up in 3-D and loaded my circuit-modeling software. I had built enough game system modifications in the past to lead me to go out on the Framework and find a circuit-modeling shareware program. That way, I could draw up the circuit mods and run the software, which in turn would simulate the system's response and tell me if I needed to alter the circuit or something. This was nothing fancy; we used PSPICE back in my first Circuits class at University of Dayton, a standard approach. But PSPICE had become expensive, so I found the cheap version I still possessed. The coolest part was the interface program I had written a few years ago to translate my CAD drawings into circuit information, which would then interface with the circuit-modeling software. Cool.

So, I finished drawing up my circuit and I pulled a power source icon over to the PSU input on the diagram and connected them. I double-clicked the power source and typed in for it to be a standard one hundred twenty volts alternating current source with a ten-ampere breaker. Then I put oscilloscope icon leads at several places in the circuit. Since, I had no clue what was inside chips A through E, I connected the leads from each pin of each chip and ran them to the analog-to-digital/digital-to-analog board of my computer system. My plan was to put the thing in operation and record what happened to the outputs after placing certain small voltages all around the board. I would use a couple of volts and a few microamp signals. I finally got all that coded and wired up by about midnight. I turned it on and got some of the most random outputs I had ever seen. I tinkered with the thing for hours, trying a different voltage here and measuring a different output there. Nothing made any sense. Neither of the chips seemed to have any standard purpose that I could determine. About three forty-five I decided to go to bed.

I loaded up my data and hardware and lugged it back to work bright and early the next morning and set up the circuit on my workbench in the lab. I felt great and not tired at all. I wasn't sad either. I went right to work on the circuit board. This time I decided to back up and focus on one chip at a time. Chips A and C had way too many pins coming out of them but chips D and E only had a couple of connected fiber optics cables and a couple of what appeared to be I/O and power pins. So, I disconnected chip E's fiber optic cable and powered up the board. A bright beam of green light came out of the end of the fiber optic cable.

"Aha!" I exclaimed. "It's a laser on a chip." I played around with it a bit until I figured out how to control the laser output by adjusting the power on the proper pin. Then I connected the cable back to Chip D. "So obviously, whatever is happening in chip D is optical!"

I removed the cables from chip D and powered up the laser chip. A faint beam of green light came out of the fiber optic cable fitting on each side of chip D.

"Chip D has, at the least, a beamsplitter," I muttered to myself. I connected the cables back to chips A and C and disconnected the cables between chips A to B and C to B. Or was it B to C? No way to know. When I powered the system up again, I didn't see light exiting the chips as expected. I turned off the lights and held a piece of paper in front of them and could barely see a faint beam coming out of A, nothing from B, and nothing from C.

I had no idea what this board was for and I wasn't sure that I would ever figure it out. I worked like mad on the thing for the entire semester I was supposed to work full-time. By the end of the semester, all I had figured out was that you could put an input signal of however many bits into one I/O chip and you would get the exact same thing out the other I/O chip. I guessed that this meant the chips A, B, and C were shorted out. I wrote up a report and handed it to Larry for my cooperative education credit. Then I signed up for my last four classes.

I had taken two classes while I was working and managed a B and a C, good enough to graduate. Once I finished the four classes this semester I would start as a civil servant for the Air Force full time. I did have to tailor my transcript a little for the job though. I was previously looking only at software coding, but the fun I had had at the ICG was making me think more about hardware.

My last four classes were senior level electives, so I got to pick basically whatever senior to graduate level computer, physics, math, or electrical engineering class I wanted to. I registered for ECT 466: Microcomputer Architectures, ECT 460: Advanced Microprocessor Systems, and I took a useless music class and a mate-selection class. I needed two liberal arts or humanities electives.

Near the end of the semester Larry called me up and told me that he had kept my desk just as I had left it and that my Top Secret security clearance had been granted! I was thrilled since I hadn't expected to be cleared, my parents and everyone I ever knew being dead and all.

I couldn't wait to get back to work on that circuit, but I had to focus on finals first. I hadn't cried since that first day on the job months before and I had only felt sad a little the night I got the call about the security clearance. The reminder of my family must have done that. My new headshrinker was doing a pretty good job with me and I was down to half the dosage that I had started with of the new drugs. The only bad side effect was the insomnia. The pills kept me awake until about three every morning and then I would be up again by seven. The good side effect was that I had lost another ten pounds. Things had not looked up for me so much in the three and a half years since The Rain.

I hadn't been worried at all about my final exams since I only had two "real" classes. The other two liberal arts/humanities classes were a waste of time and a joke all in one. The final exam for the Advanced Microprocessors class was simple, to the point, and interesting as hell. The exam read simply,

Design an I/O system to input a handwritten page via a scanner, conduct a character recognition algorithm on the page, convert it to data of any format you choose, broadcast that data to another remote computer, convert the data back into handwritten form, and output it to a printer. Show a block diagram of the system, show all switching hubs and routers, and explain where all of the data latencies and bus bottlenecks will be. Also, bonus points will be given to innovative approaches to remove bottlenecks. Then give a short essay on how this system is similar to a motherboard, and how the motherboard might be replaced by a single chip. Good luck!

I know it seems complex at first and why am I telling you about this now? I'm getting to the point, I assure you. I started out by drawing how a simple fax machine works. On the left side of the paper I drew a rectangle to represent a written page and showed it via an arrow going into another box labeled FAX. I also showed the page out of the other side of the FAX. This part of the drawing went from bottom to top. Page in at bottom of paper, FAX in the middle, and page out at top. Then I drew a horizontal arrow from the FAX on the left side of the page to a box in the center of the page labeled Router/Hub and then on to an identical FAX on the other side of the page. I drew the page coming out above the FAX on the right.

This was not enough to get the question right, of course. Much more detail would be required. So, I drew another level of detail showing the I/O input to the leftmost FAX renamed A. Box A was subdivided into three boxes: one box labeled RAM (for random access memory), a box labeled Algorithm, and one labeled Arithmetic Logic Unit (ALU). Then I penned in below the A the letters CPU for central processing unit. The Router box I renamed Box B stayed the same and then I gave another level of detail for the right FAX, now labeled C, which had a corresponding I/O box to its right. I then wrote a page each about the I/O, A, B, and C from the diagram.

On another page I drew an even deeper level of detail about each box and box within each box. To keep from boring you here I will just cut to the chase, since my actual response to the question contained ten more pages of circuit and chip and motherboard diagrams. I also drew some logic timing diagrams and bus and interconnect bandwidths per each pin. So I won't bother you with all that. But the point is that the data flow is slowed down every place there must be a wired connection. If somehow the data could be transferred from the RAM locations at each CPU directly to the RAM locations in the remote CPU, then a lot of time and therefore bandwidth could be saved. It's like taking a five-gallon bucket full of water, using a one-inch hose to transfer water from it into a one-gallon bucket, and then going from the one-gallon bucket to another five-gallon bucket again with a one-inch hose. It would sure be a lot faster if you could dump the water from one five-gallon bucket directly into the other five-gallon bucket, skipping all the in-between hoses and jugs.

I made a B in that class, which stands for "Better than needed to graduate." Larry showed up to shake my hand at graduation. He was the only person other than my instructors and a few students I had studied with that I knew. He was the only person there besides me that I had even had a meal with. I had wanted to bring Lazarus with me badly, but there was a no pets policy. That made me cry, but wouldn't it have made you cry? Laz was my only family. I told Larry as much, and I could sense he felt sympathetic for me.

"Larry, it's two fax machines on one board," I told him while he shook my hand and patted my back.

"What's a fax machine?" He looked puzzled.

"It's a machine that you put paper in and send it over phone lines to another one that prints out the paper, but that's not important right now." I chuckled my response.

"Steven, nobody likes a smartass," he laughed. "What are you talking about that is a fax machine?"

"The little circuit board is, well it's two of them actually. The I/O could be anything, a page of text for example. The data goes into the big chip on one side of the board, which is some sort of optical CPU. Then it is routed over via the little optical chip in the middle to the other identical optical CPU, and then out the other I/O device." I smiled triumphantly.

"Hmm . . . never thought of it quite that way." He tugged at his tie clip. "Then what are the other unknown chips on there? There are at least two others, right?" He smiled.

"I just figured . . ." I caught my tassel as it fell off my cap. "Uh, damn thing." I straightened it out and put my hat back on. "Uh, I just figured they supply the optical power. Am I right?"

"Why would I have two fax machines on one board, Steven?"

"Beats me. Maybe you're just playing around with an idea or something." I shrugged and that damned cap fell off again. Larry chuckled, so this time I just held it in my hand.

"When do you start back, Steven? It's about two weeks from now, right?" he asked me.

"Yeah, I was going to ask about that. Can I start earlier? I mean, uh, I'm not going . . . to . . . visit anybody . . . or anything." I looked down at my shoes for a second since I wasn't sure if I would tear up or not. "So, couldn't I just get started and get on with my life?" I asked.

"Sure, Steven." He paused. "That would be fine." Larry patted me on the shoulder, nodded, and left it at that.

"Thanks," is all I could manage to say.

"I will tell you this though—"

"Yeah, what's that?"

"You are about eighty percent right and I will tell you the rest when you get in the office. Helluva job! Let me buy you a beer, what d'you say?"

I took him up on it. Then I went home and Laz and I curled up on the couch and Sequenced for the first time in months. That was the extent of my graduation party. It was a good gesture for Larry to come to my graduation like that. Nobody else there knew me as more than some guy that was in one of his or her classes. Nobody, but Larry, knew me enough to shake my hand. That's the way it had been every birthday, Thanksgiving, and Christmas since . . . The Rain. I almost cried again, so I took another pill for it.


Загрузка...