It is a truth universally acknowledged, or at least it should be, that a technological breakthrough in possession of a grand name must be in want of a historical parallel. Behold, IBM’s latest offering: a microchip the size of a block of flats. British-designed, no less, which should immediately trigger either patriotic pride or deep suspicion, depending on how many Victorian-era industrial metaphors you can stomach.
This chip is not merely large; it is a vertical embarrassment of riches. Whereas the rest of the computing world has been obsessed with shrinking transistors to the point of quantum absurdity, IBM has gone the other way. They have built a chip that, metaphorically speaking, stands up rather than lies down. A chip that could double as a doorstop in a modernist housing estate. The press release speaks of ‘unprecedented performance’ and ‘reimagining compute density’, but one cannot help but wonder: have they simply built a very large microprocessor, stuck it on its side, and called it progress?
The historical precedent is instructive. When the Romans built the Basilica of Maxentius, they were not merely constructing a large building; they were making a statement about imperial might. Similarly, when Victorian engineers built the Great Eastern steamship, they were not just solving a transport problem; they were asserting British dominance over the seas. Today, IBM’s block-of-flats chip is being presented as a solution to the tyranny of Moore’s Law, but it smells rather more of hubris dressed up as innovation.
Let us examine the technical claims with the scepticism they deserve. The chip uses 3D stacking, which is hardly new. Sony has been stacking image sensors for years. But IBM has taken it to the extreme: 14 overlapping layers of circuitry, connected by what they call ‘through-silicon vias’. It is vertical computing as Brutalist architecture. The energy density must be terrifying. One imagines the cooling solution requiring a small river diverted through the server room. But no matter: the press release emphasises performance per watt, a metric so malleable it might as well be poetry.
What this really represents is the intellectual decadence of an industry that has run out of room on the horizontal plane. For decades, we shrank transistors. Now we cannot shrink them any further without quantum mechanics playing havoc, so we stack them. It is the same impulse that drove the Victorians to build taller and taller warehouses: not because they needed the height, but because the land was too expensive. Except here, the ‘land’ is the die surface, and the ‘landlord’ is the foundry. We have reached the physical limits of reduction, so we are now building upwards.
The national identity angle is particularly rich. British-designed, they say. But manufactured? In Taiwan, no doubt. The design is British, the fabrication is East Asian, and the profits are global. This is the post-industrial reality: a country that once built locomotives now sells blueprints. It is the intellectual property equivalent of a gentleman architect in Regency London: grand designs that others get their hands dirty implementing. One might even call it a return to the Victorian division of labour, where the British brain and the colonial brawn worked in tandem. Except this time, the colony is Taiwan, and the ‘brawn’ is a fabrication plant worth ten billion dollars.
Is this chip a game-changer? Possibly, but only in the way that the Crystal Palace was a game-changer for greenhouse design. It is a marvel of engineering, but it solves a problem of its own making. If we had not run out of space on the chip, we would not need to stack. The real breakthrough will be when someone finds a way to compute without silicon altogether. Until then, we will be stacking blocks of flats, each floor a little warmer, a little more absurd, and a little more British in its aspiration if not in its substance.
Do not mistake me: I celebrate the design and the ambition. But let us not confuse size with significance. The Empire State Building is not a better building than a medieval cathedral because it is taller. It is merely taller. And so, IBM’s chip is not necessarily better because it is thicker. It is merely thicker. And perhaps that is a metaphor for our entire age: we have run out of width, so we build height. Whether that makes for a stable future remains to be seen.
