IonQ bought Oxford Ionics for an undisclosed amount. Both make trapped ion quantum computers. Now there's one fewer company making quantum computers that need to be kept colder than space.
Why This Acquisition Might Not Be Complete Bullshit
Oxford Ionics has some interesting tech for controlling individual ions in traps. Instead of using lasers to manipulate qubits (which is expensive and error-prone), they use electric fields. This is like using a keyboard instead of a laser pointer to control your computer - more precise, less likely to fuck up.
IonQ claims their trapped ion systems maintain quantum states for around 10,000 microseconds before decoherence kills everything - that's assuming perfect lab conditions and a team of PhDs babysitting the system. In the real world? Who knows.
Oxford Ionics might extend this, but I've seen too many acquisition promises turn into integration nightmares. Quantum algorithms need hundreds or thousands of operations before decoherence makes your results garbage, and nobody's really solved that yet.
The Reality of Quantum vs Classical Computing
Current quantum computers are not faster than classical computers for anything useful. IonQ's 64-qubit system can theoretically explore 2^64 states simultaneously, but every time you measure the result, quantum mechanics randomly picks one state and throws away the rest.
The quantum advantage only exists for very specific problems:
- Molecular simulation: Quantum computers naturally model quantum systems like molecules
- Optimization: Some routing problems where you need to explore many possibilities simultaneously
- Factoring large numbers: Which breaks current encryption but has no other practical use
Everything else? Classical computers win by orders of magnitude.
Production Reality: 64 Qubits vs 40 Billion Parameters
IonQ's latest system has 64 trapped ion qubits that actually work on good days. Each qubit can theoretically be in superposition of 0 and 1 states, giving you 2^64 possible states simultaneously. That sounds impressive until you realize GPT-4 has somewhere between 1.8-220 trillion parameters depending on who you ask.
Quantum limitations that make you want to quit:
- Circuit depth: Can only run ~100 quantum gates before decoherence kills everything
- Connectivity: Not all qubits can talk to each other directly (like having USB ports that only work with specific devices)
- Measurement: Reading quantum states destroys them, so you get probabilistic outputs that may or may not be right
- Error rates: IBM publishes ~0.1% gate errors but that's best case scenario - by operation 100 your results are complete garbage
The Oxford Ionics acquisition might fix some connectivity issues. Their ion trap design lets you move qubits around physically, like having dynamic GPU memory allocation but for quantum states.
Where Quantum Computing Doesn't Completely Suck
Drug discovery is the one area where quantum computers might eventually be useful. Classical computers struggle with molecular simulation because molecules are quantum systems. Simulating even a 30-atom molecule means tracking something like 10^30 possible quantum states, which will melt your supercomputer.
Quantum computers naturally represent molecular energy states, so they might be able to simulate drugs before they're synthesized. Big emphasis on "might" - current quantum hardware has too many errors to simulate anything more complex than like 3-4 atoms without the results being complete bullshit.
What's Actually Broken With Quantum Computers
Quantum computers are finicky as fuck. Here's why they barely work:
Temperature: Need to be kept at 0.01 Kelvin (colder than space) or thermal noise destroys quantum states
Calibration: Quantum gates drift out of alignment every few hours like trying to keep a server farm tuned with a screwdriver
Error rates: IBM publishes 0.1% error rates for single operations, but quantum algorithms need hundreds of operations, so errors compound exponentially
Measurement: Every time you read the result, you destroy the quantum state, so you can only get probabilistic answers
Money Reality: Still Burning Cash
IonQ is publicly traded and loses money every quarter like most quantum companies. Revenue comes mostly from government contracts and research partnerships while they burn tens of millions on R&D.
Every quantum company burns investor cash building tech that might work someday. Could be like early internet companies that eventually paid off, or like most startups that disappear.