Location: Editorial offices of Byte (San Francisco), PC Magazine (New York) and Science & Vie Micro (Paris) / International computer shows.
Date: Spring 1993.
Point of view: Omniscient.
For nearly three decades, the global semiconductor industry was not governed by simple commercial rules or stock market fluctuations. It submitted to an absolute dogma, a mathematical religion decreed in 1965 by Gordon Moore, the co-founder of Intel.
This postulate, resembling a self-fulfilling prophecy, asserted that the number of transistors on a silicon chip would double every eighteen months, leading to an exponential increase in computing power at a constant unit production cost.
This law wasn't just an empirical observation etched in the marble of Silicon Valley; it was the frenetic metronome dictating the heartbeat of technological capitalism. A perpetual, violent, and ruthless race forward.
It pushed battalions of engineers to nervous exhaustion, condemned foundries to constantly reinvent themselves, and forced consumers to discard their barely unboxed machines to avoid sinking into the abyss of obsolescence. A year earlier, in March 1992, the world had still vibrated to the rhythm of this cadenced frenzy when Andy Grove and Bill Gates jointly unveiled the 66 MHz Intel i486 DX2 and the enticing Windows 92, to the fanatical cheers of San Francisco's Moscone Center.
However, in the spring of 1993, the metronome stopped abruptly.
In the smoke-filled newsrooms of major specialized magazines, from the American West Coast to the banks of the Seine, astonishment gradually gave way to the cold anxiety of the blank page. The perennial editorial topics of spring, which usually punctuated the life of the technology press, demanded their annual tributes: benchmarks of new super-powerful processors, comparisons of revolutionary operating systems, analyses of the meteoric rise in clock frequencies, and grand, lyrical pronouncements on the bright future of microcomputing.
But the newsrooms' fax machines remained hopelessly silent. There was nothing to test. Nothing to compare. Nothing to prophesy.
The global IT industry had just entered a silent winter, an industrial freeze that defied all economic rationality. The market, usually so talkative, so quick to oversell the slightest minor innovation, was struck by a deafening silence from its leading figures.
In Santa Clara, the nerve center of Silicon Valley, tech giant Intel kept postponing, quarter after quarter, the official announcement of its highly anticipated fifth-generation processor, the famous »P5« project, which rumors were already dubbing the »Pentium.« Company spokespeople, their faces tense under the photographers' flashes, politely cited complex heat dissipation problems or yield issues related to the submicron manufacturing process.
The reality, carefully suppressed under the guise of state secrecy and national security, was utterly bleak. The press was unaware that on May 1, 1992, »Ogre Day,« the VoltaOS-M virus had encrypted Intel's Genesis server with a scalable, virtually unbreakable asymmetric RSA key.
Deprived of their original lithography masks and their most intimate source code, American engineers had been violently thrown back to the Stone Age. They spent the year trying to redesign the Pentium from old paper drafts, fragmented memories, and reverse-engineering their own components.
A daily humiliation that drained the company's coffers and paralyzed its development schedule.
Further north, in the perpetual grayness of Redmond, the Microsoft behemoth remained silent with the same morbid and shameful obstinacy. No major update had been deployed for Windows 92. No triumphant announcement for its successor had reassured the financial markets.
And for good reason: Lazarus Bonaparte's digital weapon had vaporized the file allocation tables (FATs) of their development servers. Bill Gates's teams were struggling to piece together millions of fragmented lines of code, painfully recreating a fundamentally unstable software ecosystem.
But the most agonizing silence, the most laden with meaning, came from the Old Continent.
In Ivry-sur-Seine, the prodigy who had humiliated the French state, robbed continental banking networks, and subjugated the Japanese arcade industry seemed to have fallen asleep on his colossal war chest. Volta SA announced no new chips for the civilian market.
The VESLA-II processor, although it still outperformed everything else on the planet, was technologically outdated from the previous year. The industrial world awaited the French response, the next iteration of terror.
But Lazare Bonaparte remained a categorical ghost. Locked away in the secrecy of the MegaFab under construction in the Alsace plain and focused on debugging his next superscalar architecture, he let nothing leak out.
Faced with this glaring void, journalists, unable to admit or even suspect that they were witnessing the consequences of a devastating geopolitical shadow war, invented narratives to justify the breakdown. They theorized, quite seriously, about a » technological glass ceiling.« Eminent academics wrote that Moore's Law had finally hit the implacable wall of the limits of quantum physics.
They asserted that the heat from the components could no longer be dissipated by conventional means, that miniaturization was generating uncontrollable electrical arcs, and that computing had simply reached its final plateau of maturity.
Nature, however, just like rampant capitalism, abhors a vacuum.
With the titans silent, holed up in their fortresses licking their wounds or stockpiling ammunition for the final assault, the media and commercial scene was suddenly empty. At the major IT trade shows in the spring, under the immense glass roofs of CeBIT in Hanover and in the sweltering aisles of Comdex in Las Vegas, the spotlights abandoned the surprisingly discreet booths of IBM and Intel to focus their beams on a veritable court of miracles.
It was the dance of amateurs. The era of illusion.
Dozens of startups born in San Jose garages or Hsinchu technology parks in Taiwan, ambitious small British foundries, and newly capitalized American university labs seized the opportunity to step into the limelight. Convinced that the era of monopolies was faltering and that the kings had died of old age, these minor players flooded the trade show aisles with their own creations.
They imagined themselves experiencing a new gold rush, strutting their stuff in magazines, giving pompous interviews about the »democratization of silicon« and the »liberation of architecture.« Among this diverse group, two companies stood out in particular by claiming to offer the spiritual successor to the French prodigy: Zenith Dynamics, a Californian startup financed by aggressive venture capital funds, and Aether Technologies, an Anglo-Taiwanese consortium.
Their engineers had studied from afar the prowess of the DGA's IMPERATOR servers and the triumph of Japanese arcade cards. They understood that the future no longer belonged to Intel's CISC (Complex Instruction Set Computer), cumbersome, energy-intensive, and burdened with obsolete instructions.
They had therefore thrown themselves headlong into the design of RISC (Reduced Instruction Set Computer) chips, hoping to emulate the architectural purity of Lazare Bonaparte's VESLA-II.
However, reproducing the surface of a masterpiece is not the same as capturing its soul.
Zenith Dynamics' flagship processor, pompously dubbed the »Vanguard-1,« drew crowds at CeBIT. On paper and on the glossy slides handed out by smiling hostesses, the chip promised parallel instruction execution and a staggering clock speed of 75 MHz.
Its designers had appropriated pipeline concepts that Lazare had secretly patented during his 1987 »fever.« They believed they had found patent-free loopholes in the publicly available mathematical literature.
But the physical reality of this piece of silicon was pathetically rustic compared to Volta's goldsmithing.
The original VESLA-II, forged in the mind of a man from 2026, was a marvel of balance. Its manufacturing process, mastered thanks to the ASML machines that Lazare had smuggled to Huawei, approached the micron level with perfect efficiency.
The French processor integrated a meticulously precise mathematical coprocessor (FPU) directly onto its die, and above all, it communicated with memory via address buses free of any latency. The VESLA-II was a monolith of black ceramic, radically elegant, that calculated at devastating speeds while remaining structurally cool.
The optimization of its electron paths was such that it required only a simple passive heat sink to dissipate its heat.
The »Vanguard-1,« on the other hand, was a thermal nightmare. Lacking access to deep ultraviolet lithography lasers (monopolized by Volta and ASML), Zenith Dynamics had to make do with old Nikon and Canon machines.
Their etching capabilities were limited to 1.5 microns. To compensate for the size of the transistors and try to match Volta's clock speed, the Americans had to overclock their chip. The result was a component that burned extremely hot, requiring enormous, noisy fans that transformed the computer cases into veritable industrial wind tunnels.
Even more tragic was their attempt to mimic Lazarus's smooth execution. The VESLA-II excelled in pipeline fluidity because Lazarus had integrated branch prediction algorithms of unparalleled simplicity and robustness (the very same ones he had used in VoltaOS-M to destroy the NSA).
Zenith's processor, on the other hand, ran into logical walls. Without effective prediction, their execution pipelines constantly emptied. Every time a conditional instruction wasn't correctly anticipated, the processor had to cancel its calculations, flush its cache (a »pipeline flush« ), and start again.
This incessant hiccup literally killed overall performance. On paper, the Vanguard-1 boasted 75 MHz; in reality, it floundered as soon as it was presented with a complex financial calculation.
On the other end of the copier spectrum, Aether Technologies presented the »Quantum-X.« Their engineers had adopted an even more hybrid and inevitably corrupt approach. Aware that the American civilian market remained terrified of abandoning old software developed under MS-DOS, Aether had attempted the impossible compromise: a RISC chip, but equipped with an internal software translation layer (microcode) intended to emulate Intel's CISC instructions to guarantee backward compatibility.
It was a silicon Frankenstein.
Where Lazare Bonaparte had made a clean and definitive break with the past by imposing its own native language, Aether tried to please everyone. The emulation layer generated monstrous latency. Requests piled up in the Level 1 (L1) cache, which consistently overflowed.
Moreover, unlike Volta SA, which was secretly developing a revolutionary new generation of dynamic memory (CELLA-64M) with the Soviets, incorporating deep-trench capacitors, these startups had to make do with cheap, slow, and asynchronous Asian DRAM modules. The super-powerful processor spent most of its time waiting for the congested system memory to provide it with data.
A pathetic bottleneck.
Finally, and this was the ultimate blind spot of all these new contenders, they did not possess the soul of the machine.
The empire of the Ivry prodigy rested on perfect symbiosis, the definitive coupling between hardware and software. The VESLA-II was invincible only because it ran on VoltaOS, a masterful UNIX kernel, encrypted and purged of all unnecessary code, designed specifically for its instruction set.
And above all, it was equipped with the SONG graphics coprocessor, a second brain dedicated to the image that relieved the main CPU of all visual burdens.
The CeBIT startups, for their part, delivered bare processors. They threw them to the wolves of a fragmented software market, forcing them to run on cobbled-together versions of DOS, sluggish iterations of OS/2, or fledgling Unix kernels.
They lacked hardware cryptographic shields like the famous CENTURION module that protected the DGA's IMPERATOR servers. They didn't possess the vision of Lazare Bonaparte; they had only the arrogance of opportunists.
Yet, in the halls of Comdex, the tech press was intoxicated by this profusion of new products. Journalists applauded the end of the original Silicon Valley's monopoly and celebrated the rise of these young upstarts.
They published glowing articles on the user-friendliness of new alternative operating systems, such as the recently launched BeOS, which attempted to replicate, with garish colors, the fluid windowing that Karim Belkacem had imposed on the professional world. They spoke of a new golden age, a springtime of free software, completely unaware that these companies had ventured onto territory heavily mined by the intellectual property of the most dangerous man in Europe.
All this excitement, all this media and commercial commotion, was methodically making its way back to the glass-fronted offices in Ivry-sur-Seine.
In the inner sanctum of Volta SA, reports poured in, scrutinized with obsessive precision by the company's legal department. Armand Delacroix, Lazare's ruthless corporate lawyer, chafed at the bit. He saw these startups appropriating his client's patented concepts.
He saw Volta's register architecture, parallel data buses, and primitive graphics rendering algorithms being shamelessly copied and commercialized. For the lawyer, this was a clear case of intellectual property infringement that demanded a swift legal response, customs seizures, and lawsuits seeking tens of millions of dollars in damages.
But Lazare Bonaparte, motionless behind the windows of his office, gazing at the Parisian suburbs in the spring rain, had expressly forbidden his lawyers from attacking. The titan refused to swat flies.
He watched these startups parade by with the cynical pity of a predator watching scavengers clear the area before his own meal. He knew that the illusion of an open market was the best cover to mask the construction of his total monopoly, a silicon fortress that, when the time came, would sweep away these ridiculous copper prostheses to impose the true law of the Builder.
Location: Volta SA executive office (The « Bunker »), Ivry-sur-Seine. Date: Spring 1993. Point of view: Omniscient (Sliding focus on Armand Delacroix, Édouard Renault-Tessier and Lazare Bonaparte).
The sky over the Parisian suburbs unleashed a slanting, cold rain, typical of this indecisive spring, which lashed against the immense bay windows of Volta SA's world headquarters in Ivry-sur-Seine. The atmosphere inside the »Bunker« contrasted sharply with the grayness outside.
The factory was a living, feverish organism, its metabolism running at full speed to maintain the massive production rates demanded by government contracts and Asian partners.
Yet, on the third floor, in the hushed corridor leading to the general management office, the tumult of the assembly lines was now only an imperceptible murmur, muffled by military-grade soundproofing.
Armand Delacroix strode forward across the thick, dark carpet, his face impassive, his jaw clenched. The business lawyer, well into his sixties, with his silver mane and predatory gaze, was not a man easily intimidated.
He had managed the interests of the upper class and the largest European conglomerates for decades. But what he had discovered in the last forty-eight hours had shattered his legendary composure.
On his heels, almost struggling to keep up with his pace, Édouard Renault-Tessier was sweating profusely. The Volta financial director, usually so meticulous about the symmetry of his silk tie, clutched a heavy leather briefcase embossed with his initials to his chest.
The two men didn't bother waiting for the secretary to announce their arrival. Delacroix pushed open the heavy solid oak door of the presidential office with such force that the brass handle slammed against the wall stop.
Lazare Bonaparte stood with his hands clasped behind his back, gazing at the rain that was drenching the Seine Valley. The sixty-year-old engineer, trapped in the twenty-six-year-old body that still bore the scars of the Eindhoven ambush, didn't flinch.
He didn't even turn around immediately. The giant from Ivry pivoted with exasperating slowness, his obsidian gaze gliding over his two lieutenants. »Good manners are dying out, Armand,« the Builder said softly in a glacial voice, whose low resonance instantly chilled the atmosphere of the room. »Good manners are a luxury reserved for peacetime, Lazarus!« thundered the lawyer, striding over to the immense mahogany desk and slamming his briefcase down. »We're being robbed in broad daylight!« Édouard hurried around the table to open the briefcase, his hands trembling.
He pulled out bundles of documents printed on glossy paper, technical datasheets retrieved from the stands at CeBIT in Hanover, and offprints from the American technology press. »Have you read the strategic intelligence reports I sent you last night?« asked the CFO, his voice choked with anxiety. »These Californian and Taiwanese startups, Zenith Dynamics, Aether Technologies, and all these small groups launching into graphical operating systems...« « I've read them, » Lazare replied with stony indifference, approaching his leather armchair to sit down, his back perfectly straight.
« The Vanguard-1, the Quantum-X. Touching attempts, but desperately poorly executed, to appropriate RISC architecture. » »Touching attempts?!« exclaimed Maître Delacroix, his temples throbbing. »Lazare, you're in denial!
I've just spent three sleepless nights with the top intellectual property experts in my firm and two of your Level 4 engineers. We've dissected the technical specifications of their processors. This isn't inspiration.
It's industrial-scale counterfeiting!« The lawyer planted a vengeful finger on one of the architectural diagrams spread out on the leather.
« Look at the layout of their branch prediction unit. Look at the topology of their ghost registers for out-of-order execution! They've even dared to replicate the principle of the matrix-integrated cache controller.
These idiots think they've found royalty-free flaws in the public mathematical literature, but they're directly and brazenly violating the dozens of patents you had me file in 1987! »
Delacroix opened a scarlet cardboard folder, revealing dozens of pages of legal letterhead, initialed and ready to be sent. The predatory excitement of the courtroom gleamed in his eyes.
« We know how to make our enemies bleed, Lazarus! Look at the effectiveness of our current legal strategy against Intel! The patent infringement lawsuit we filed against them through our Delaware front consortium is a marvel of attrition warfare!
This colossal lawsuit for the unauthorized use of our patents on their Pentium project is crippling them. It's currently costing them millions of dollars a month in legal fees, audits, and provisional penalties!
We're suffocating them under mountains of technical requests! »
The lawyer slammed his hand down on the CeBIT startup files.
« Give me permission to apply the same guillotine to these parasites! I have the preliminary injunctions. The cease and desist letters.
I can have all their booths at the next Comdex seized by the U.S. federal marshals. I can drag them before the World Intellectual Property Organization and destroy them in six months. Let's make a dazzling example! »
Édouard Renault-Tessier nodded vigorously, his eyes shining with accounting greed. »Armand is absolutely right, Lazare. We can't let them establish themselves. The Intel lawsuit proves that our intellectual property is a weapon of mass destruction.
If we let these startups flood the market with chips that mimic our architecture, they'll get developers used to inferior products! They'll cannibalize our market even before we launch our consumer PC with Compaq!« Lazare Bonaparte glanced down at the colorful brochures.
He ran his pale index finger over the technical diagrams. He recognized the imprint of his own mind, his equations born in the fever of isolation, crudely copied by engineers striving to mimic the incomprehensible.
It was a wretched plagiarism, a patchwork that, technically speaking, deserved to be destroyed.
But the young Titan did not react with the inflated ego of a Silicon Valley CEO.
Lazarus raised his head very slowly. »Gather your documents, Armand,« he ordered, his voice plummeting to depths of authority the lawyer had never heard before. »Take these orders, and put them through the shredder before the end of the morning.« The lawyer froze, stunned. His arm remained suspended above the documents. »Excuse me?« Delacroix stammered, unable to process the information. »You... you refuse to defend your patents against these thieves?« »You have the vision of a notary, Armand,« observed the titan of Ivry, leaning back in his armchair, clasping his hands over his abdomen. »You're reacting with the primal instinct of a shopkeeper whose apple has just been stolen from his stall, without considering the topography of the battlefield.« Lazare leaned forward, the aura of the old shadow strategist instantly regaining control.
« The monster lawsuit we filed against Intel is a surgical, targeted maneuver. Bleeding Intel millions a month is a vital tactical necessity. It's a legal tourniquet designed to stifle their R&D, prevent them from investing in new foundries, and slow down the Pentium rollout.
We're fighting a king. But attacking these second-rate startups with the same fury? That's sheer recklessness. »How is it unwise to crush thieves?!« Edward protested. »
« Because if we unleash a global legal storm against every little smelter in California and Taiwan, we'll have to lay our cards on the table! » Lazare snapped, his voice razor-sharp. « You'll have to prove in federal courts that the myriad of empty shell companies in Delaware and Luxembourg foundations that hold our patents all actually belong to Volta SA! »
A sticky, revealing silence began to settle in the office.
« The moment you make this canvas public, » the founder continued, « the lawyers for Intel, IBM, and Microsoft will stop fighting amongst themselves. If they realize today that we hold a monopoly on the RISC architecture, as well as patents on future cellular networks, serial communication buses, and mass cryptography, they will panic.
They will unite. They will form a survival cartel. They will weep in the Oval Office, arguing for the national security of the United States.
The US Congress will pass exceptional laws to invalidate our patents with the stroke of a pen, under the pretext of judicial review. »
Lazare dismissed the formal notices with a contemptuous wave of his hand.
« The Intel lawsuit is working because it has the appearance of an isolated private dispute. I refuse to reveal the full order of battle of my artillery for crushing insects. »
Delacroix wiped his brow. The logic was irrefutable, but the potential financial loss still gnawed at him. »But Lazarus... While we remain silent, these amateurs are filling the void we created by paralyzing the giants!
They're selling chips! They're winning over the developers!« A spark of unfathomable cruelty, the glimmer of an evil god manipulating destinies, lit up in Lazarus's dark eyes.
« And that's exactly what I want them to do, Edward. »
Lazare stood up. His body, aching from the spring dampness, protested, but he ignored it with haughty demeanor. He walked around his immense desk to stand behind his two directors.
« Look at them bustling about, these little substitute geniuses, » the architect murmured, his voice dropping an octave. « For ten years, the industry has lived under the mental and software dictatorship of Intel's x86 architecture and MS-DOS.
It's an intellectual monopoly. A religion. And breaking a religion costs billions in marketing and evangelization. »
He turned around abruptly, his eyes fierce.
By flooding the trade shows in Hanover and Las Vegas with these alternative RISC chips and fledgling graphics systems, these audacious startups are doing our work for us. And the best part? They're doing it for free.
With their own American venture capital! They're shattering the foundations of Intel's hegemony. They're taking control of the situation by proving to developers worldwide that the American standard isn't inevitable.
They're forcing them to rewrite their programs, to adapt to a new way of thinking about hardware.
The Titan advanced to the center of the room, imposing his vision with a gravitational force of attraction that pinned his men to the spot.
« Those Zenith Dynamics enthusiasts aren't our competitors. They're our laborers. They're our useful idiots.
They're preparing the ground. They're uprooting the weeds of the old order, and they're sowing the seeds of change in the minds of consumers. Let them dance!
Let them exhaust themselves evangelizing the market at their own expense, reassuring the press, and dealing with the initial hardware glitches! »
Édouard Renault-Tessier was fascinated. The audacity of letting the competitor build the psychological infrastructure of his own defeat was breathtakingly cynical. »And when will the earth be ready, Lazarus?« asked the former banker, his throat dry.
« When the time is right, » replied the pioneer, touching the dark aluminum prototype of the future Volta personal computer, « when the market is hungry for a real solution after being disappointed by the crashes of these thermal clones, we will emerge from the Blind Spot. The VESLA 2100 series, manufactured with sovereign expertise, powered by the true VoltaOS 3.0, and assembled with Compaq's logistical firepower, will hit the market. »
Lazarus's gaze drifted for a moment, probing the depths of an inevitable future.
« As for those charming startups… The very second we unveil our product line, the technical comparison will be so overwhelming that their stock will plummet before lunchtime. That's precisely when, Armand, you'll pull out your patents to finish them off.
We'll buy out the most compliant ones for a pittance to absorb their best engineers, and we'll let the others suffocate under the weight of legal fees. Gentlemen, war isn't fought against foot soldiers when you're aiming for the heads of kings.
Put those cease and desist letters through the shredder. The predator's pity is merely waiting for the perfect moment to strike. »