Quantum computers promise significant leaps forward, but today they struggle to solve practical problems because they’re noisy and unreliable. One major challenge is calculating the “ground state” of molecules, essentially the most stable arrangement of electrons. Think of it as finding the lowest valley in a complicated mountain range.

The Problem with Current Methods

Today, it's challenging for quantum algorithms to descend this mountain as quantum computers struggle with noise. Typically, researchers use a Variational Quantum Eigensolver (VQE), which mixes everyday computers with quantum computers, to test different settings on a quantum computer while using classical computers to decide how best to tweak those settings.

However, a VQE encounters two major issues:

• Flat Spots: Like wandering into a flat area on the mountain, the algorithm doesn’t know which way to go.
• Too Many Measurements: It needs to check energy levels so frequently that it becomes impractical and noisy quantum computers can’t handle it.
  
  

A Smart Shortcut: The Greedy Algorithm

Researchers have come up with a simpler version, called the Greedy Gradient-Free Adaptive VQE (GGA-VQE). Here’s how it works, using our mountain analogy:

• Quick Checks: Instead of looking everywhere, it quickly checks a few paths down the mountain.
• Pick the Best Path: It immediately chooses the path with the steepest descent and, therefore, the fastest route.
• No way Back: Once it decides a route, it's committed. There's no looking back, preventing further measurements and more complex calculations.
  
  

The simpler and faster steps help the algorithm to avoid mistakes and solve the problem quicker.

Why Does This Matter?

The greedy gradient approach reduces the work load of a quantum computer as there are fewer measurements and, therefore, fewer mistakes. When compared with previous methods, it was much improved, especially when simulating simple molecules. For example, it was twice as accurate when looking at a water molecule, despite encountering noise.

The researchers didn’t just simulate this method on normal computers, they tested it on a real quantum computer with 25 quantum bits (or qubits) to solve complex puzzles. Despite the noise, the algorithm was more than 98% accurate.

This breakthrough isn’t just theory, the benefits are real:

• Medicines: The ability to find precise ground states can help researchers find new therapeutics, materials and even chemicals even faster.
• Efficiencies: The noise from quantum computers can be reduced, which means we can use them sooner and not in decades' time.
  
  
  

What are the future applications:

This breakthrough means researchers can significantly accelerate the development of quantum computers as it shows we don’t need perfect hardware, just smarter algorithms, to predict reactions and create new materials that are beyond the reach of today's classical computers.

Qubit Pharmaceuticals' Greedy algorithm overcomes many of the hurdles facing quantum computers. The new approach reduces the need to wait for error-free quantum computers and allows us to use quantum tech today to solve real-life problems.

This breakthrough is a big step towards accessing the immense potential of quantum computing and unlocking new realities.

Access the full paper here