The crypto door beeped once, waking Susan from her depressing reverie. The door had rotated past its fully open position and would be closed again in five seconds, having made a complete 360-degree rotation. Susan gathered her thoughts and stepped through the opening. A computer made note of her entry.

Although she had practically lived in Crypto since its completion three years ago, the sight of it still amazed her. The main room was an enormous circular chamber that rose five stories. Its transparent, domed ceiling towered 120 feet at its central peak. The Plexiglas cupola was embedded with a polycarbonate mesh-a protective web capable of withstanding a two-megaton blast. The screen filtered the sunlight into delicate lacework across the walls. Tiny particles of dust drifted upward in wide unsuspecting spirals-captives of the dome's powerful deionizing system.

The room's sloping sides arched broadly at the top and then became almost vertical as they approached eye level. Then they became subtly translucent and graduated to an opaque black as they reached the floor-a shimmering expanse of polished black tile that shone with an eerie luster, giving one the unsettling sensation that the floor was transparent. Black ice.

Pushing through the center of the floor like the tip of a colossal torpedo was the machine for which the dome had been built. Its sleek black contour arched twenty-three feet in the air before plunging back into the floor below. Curved and smooth, it was as if an enormous killer whale had been frozen mid breach in a frigid sea.

This was TRANSLTR, the single most expensive piece of computing equipment in the world-a machine the NSA swore did not exist.

Like an iceberg, the machine hid 90 percent of its mass and power deep beneath the surface. Its secret was locked in a ceramic silo that went six stories straight down-a rocketlike hull surrounded by a winding maze of catwalks, cables, and hissing exhaust from the freon cooling system. The power generators at the bottom droned in a perpetual low-frequency hum that gave the acoustics in Crypto a dead, ghostlike quality.

TRANSLTR, like all great technological advancements, had been a child of necessity. During the 1980s, the NSA witnessed a revolution in telecommunications that would change the world of intelligence reconnaissance forever-public access to the Internet. More specifically, the arrival of E-mail.

Criminals, terrorists, and spies had grown tired of having their phones tapped and immediately embraced this new means of global communication. E-mail had the security of conventional mail and the speed of the telephone. Since the transfers traveled through underground fiber-optic lines and were never transmitted into the airwaves, they were entirely intercept-proof-at least that was the perception.

In reality, intercepting E-mail as it zipped across the Internet was child's play for the NSA's techno-gurus. The Internet was not the new home computer revelation that most believed. It had been created by the Department of Defense three decades earlier-an enormous network of computers designed to provide secure government communication in the event of nuclear war. The eyes and ears of the NSA were old Internet pros. People conducting illegal business via E-mail quickly learned their secrets were not as private as they'd thought. The FBI, DEA, IRS, and other U.S. law enforcement agencies-aided by the NSA's staff of wily hackers-enjoyed a tidal wave of arrests and convictions.

Of course, when the computer users of the world found out the U.S. government had open access to their E-mail communications, a cry of outrage went up. Even pen pals, using E-mail for nothing more than recreational correspondence, found the lack of privacy unsettling. Across the globe, entrepreneurial programmers began working on a way to keep E-mail more secure. They quickly found one and public-key encryption was born.

Public-key encryption was a concept as simple as it was brilliant. It consisted of easy-to-use, home-computer software that scrambled personal E-mail messages in such a way that they were totally unreadable. A user could write a letter and run it through the encryption software, and the text would come out the other side looking like random nonsense-totally illegible-a code. Anyone intercepting the transmission found only an unreadable garble on the screen.

The only way to unscramble the message was to enter the sender's "pass-key"-a secret series of characters that functioned much like a PIN number at an automatic teller. The pass-keys were generally quite long and complex; they carried all the information necessary to instruct the encryption algorithm exactly what mathematical operations to follow tore-create the original message.

A user could now send E-mail in confidence. Even if the transmission was intercepted, only those who were given the key could ever decipher it.

The NSA felt the crunch immediately. The codes they were facing were no longer simple substitution ciphers crackable with pencil and graph paper-they were computer-generated hash functions that employed chaos theory and multiple symbolic alphabets to scramble messages into seemingly hopeless randomness.

At first, the pass-keys being used were short enough for the NSA's computers to "guess." If a desired pass-key had ten digits, a computer was programmed to try every possibility between 0000000000 and 9999999999. Sooner or later the computer hit the correct sequence. This method of trial-and-error guessing was known as "brute force attack." It was time-consuming but mathematically guaranteed to work.

As the world got wise to the power of brute-force code-breaking, the pass-keys started getting longer and longer. The computer time needed to "guess" the correct key grew from weeks to months and finally to years.

By the 1990s, pass-keys were over fifty characters long and employed the full 256-character ASCII alphabet of letters, numbers, and symbols. The number of different possibilities was in the neighborhood of 10120-ten with 120 zeros after it. Correctly guessing a pass-key was as mathematically unlikely as choosing the correct grain of sand from a three-mile beach. It was estimated that a successful brute-force attack on a standard sixty-four-bit key would take the NSA's fastest computer-the top-secret Cray/Josephson II-over nineteen years to break. By the time the computer guessed the key and broke the code, the contents of the message would be irrelevant.

Caught in a virtual intelligence blackout, the NSA passed a top-secret directive that was endorsed by the President of the United States. Buoyed by federal funds and a carte blanche to do whatever was necessary to solve the problem, the NSA set out to build the impossible: the world's first universal code-breaking machine.

Despite the opinion of many engineers that the newly proposed code-breaking computer was impossible to build, the NSA lived by its motto: Everything is possible. The impossible just takes longer.

Five years, half a million man-hours, and $1.9 billion later, the NSA proved it once again. The last of the three million, stamp-size processors was hand-soldered in place, the final internal programming was finished, and the ceramic shell was welded shut. TRANSLTR had been born.

Although the secret internal workings of TRANSLTR were the product of many minds and were not fully understood by any one individual, its basic principle was simple: Many hands make light work.

Its three million processors would all work in parallel-counting upward at blinding speed, trying every new permutation as they went. The hope was that even codes with unthinkably colossal pass-keys would not be safe from TRANSLTR's tenacity. This multibillion-dollar masterpiece would use the power of parallel processing as well as some highly classified advances in clear text assessment to guess pass-keys and break codes. It would derive its power not only from its staggering number of processors but also from new advances in quantum computing-an emerging technology that allowed information to be stored as quantum-mechanical states rather than solely as binary data.

The moment of truth came on a blustery Thursday morning in October. The first live test. Despite uncertainty about how fast the machine would be, there was one thing on which the engineers agreed-if the processors all functioned in parallel, TRANSLTR would be powerful. The question was how powerful.

The answer came twelve minutes later. There was a stunned silence from the handful in attendance when the printout sprang to life and delivered the cleartext-the broken code. TRANSLTR had just located a sixty-four-character key in a little over ten minutes, almost a million times faster than the two decades it would have taken the NSA's second-fastest computer.

Led by the deputy director of operations, Commander Trevor J. Strathmore, the NSA's Office of Production had triumphed. TRANSLTR was a success. In the interest of keeping their success a secret, Commander Strathmore immediately leaked information that the project had been a complete failure. All the activity in the Crypto wing was supposedly an attempt to salvage their $2 billion fiasco. Only the NSA elite knew the truth-TRANSLTR was cracking hundreds of codes every day.

With word on the street that computer-encrypted codes were entirely unbreakable-even by the all-powerful NSA-the secrets poured in. Drug lords, terrorists, and embezzlers alike-weary of having their cellular phone transmissions intercepted-were turning to the exciting new medium of encrypted E-mail for instantaneous global communications. Never again would they have to face a grand jury and hear their own voice rolling off tape, proof of some long-forgotten cellular phone conversation plucked from the air by an NSA satellite.

Intelligence gathering had never been easier. Codes intercepted by the NSA entered TRANSLTR as totally illegible ciphers and were spit out minutes later as perfectly readable cleartext. No more secrets.

To make their charade of incompetence complete, the NSA lobbied fiercely against all new computer encryption software, insisting it crippled them and made it impossible for lawmakers to catch and prosecute the criminals. Civil rights groups rejoiced, insisting the NSA shouldn't be reading their mail anyway. Encryption software kept rolling off the presses. The NSA had lost the battle-exactly as it had planned. The entire electronic global community had been fooled... or so it seemed.