BlueQubit

🔬 Breakthrough: Google DeepMind's AlphaQubit is making quantum computers more reliable! Their new AI system can spot errors in quantum processing with record-breaking accuracy - 6% better than previous methods. ✨ This could accelerate the development of quantum computers that solve problems classical computers would take billions of years to crack. 🧮 A major milestone for quantum error correction! 🚀 https://lnkd.in/djbAEzkX

Making progress in Quantum Computing and Quantum Error Correction with 28 logical qubits! Let's dive in! 🧠💫 Check out the latest paper 📄 "Logical computation demonstrated with a neutral atom quantum processor" by researchers at Microsoft Azure Quantum & Atom Computing! This architecture was demonstrated with 256 Ytterbium atoms, showing potential for larger-scale quantum error correction 💾 What are the results? 🔹 The team created a 24-qubit logical cat state using [4,2,2] code with error rates significantly better than unencoded versions 📊 🔹 Implemented Bernstein-Vazirani algorithm with up to 28 logical qubits, demonstrating better-than-physical error rates 📈 🔹 Demonstrated fault-tolerant quantum computation with repeated loss correction across multiple rounds 🎯 🔹 Achieved repeated loss and error correction using the distance-three [9,1,3] Bacon-Shor code 🔄 🔹 Single-qubit Clifford gates achieved 99.85(2)% fidelity, with two-qubit gate infidelity as low as 0.4(1)% ⚡ 🔹 All-to-all connectivity achieved through atom movement with loss probability below 0.1% per move 🎯 The work successfully detects and corrects both qubit loss and computational errors, marking a crucial step toward practical quantum computing! 🎉⚛️ Read the full paper here: https://lnkd.in/eMu435DT 📚

🎉 A big Bluequbit welcome to our new software engineer, Arman Gevorgyan! 🚀 Arman brings incredible talent, passion and experience to our team. We’re thrilled to have him on board as we continue building cutting-edge software for quantum computing 🚀 Welcome to the team, Arman! We can’t wait to see the amazing things we’ll achieve together. 💻✨

What can neural networks tell us about quantum states? 🧠💫 Important new paper 📄 "When can classical neural networks represent quantum states?" by researchers (including John Preskill) at Caltech, UC Berkeley, UIUC & AWS Quantum Computing! Key Findings: 🔹 Neural networks with depth O(log log(n)) and width n^O(log^(D-1)(n)) can efficiently represent quantum states with exponentially decaying conditional correlations 🔹 Recurrent neural networks need memory size O(log^D(n)) for D-dimensional lattices 🔹 Revealed how rotation angles in cluster states control the range of correlations, critically affecting neural network training 🔹 Identified phase transitions in random 2D circuits and tensor networks at small constant depth/bond dimension 🔹 Proved connection between conditional independence and entanglement area laws for sign-free quantum states This work establishes when neural networks can efficiently represent quantum many-body states through conditional correlations! 🎯⚛️ Read more here: https://lnkd.in/g3QEXyjx

We’re excited to welcome Jake Malliaros to the BlueQubit team 🚀 Jake previously served as Quantum Stream Lead at Canada’s Creative Destruction Lab (CDL), which has helped launch some of the most successful quantum companies. Now, he’s joining us as Vice President of Business Development, bringing years of experience in the quantum industry and a deep understanding of where the field is heading. As a founder, it’s incredibly rewarding when someone who truly understands and believes in your vision chooses to join the mission. #quantum #business #growth #cdl #bluequbit

📄 "Unravelling Quantum Dynamics Using Flow Equations" shows why GPU acceleration with platforms like BlueQubit is important for quantum computing research! Researchers from Free University of Berlin demonstrate a powerful new GPU-accelerated method for simulating quantum dynamics, offering a significant breakthrough in tackling previously intractable quantum systems! Quantum dynamics simulations are essential for developing new materials, solar cells, batteries, and sensors - but they're incredibly challenging due to the exponential scaling of quantum states. 🔋 While traditional methods struggle with highly entangled systems, this new approach offers a way forward that's limited only by desired numerical accuracy rather than entanglement barriers. 🎯 Key Results: 🔹 Achieved over 8x speedup using a single GTX 1660Ti GPU vs CPU for 24-particle systems 🔹 Reduced computation time from 2+ hours (CPU) to under 15 minutes (GPU) 🔹 Successfully demonstrated simulation of 2D systems by "unfolding" into 1D representations 🔹 Developed novel "scrambling transforms" to improve convergence in challenging cases P.S. BlueQubit quantum computing platform is powered by NVIDIA GPUs for efficient quantum circuit simulations. Try it out! https://lnkd.in/dMm783_x

Welcome to BlueQubit, Jake Malliaros!

The BlueQubit team was honored to speak at Inside Quantum Technology last week! We showcased our recent achievements in applying quantum machine learning methods to image processing, discussed experiments conducted on IBM and Quantinuum QC quantum computers, and dove into the latest trends and opportunities in QML. Try the BlueQubit platform, the easiest way to run quantum algorithms on CPU, GPUs, and QPUs! https://lnkd.in/gckr-kfy

📄 New Paper Alert! "Efficient charge-preserving excited state preparation with variational quantum algorithms" 🧑💻 Researchers from NVIDIA, Stony Brook University, and Brookhaven National Laboratory introduce CPVQD (Charge-Preserving Variational Quantum Deflation), a breakthrough algorithm that makes excited state calculations more efficient on quantum computers! Key Results: 🔹 Successfully tested systems up to 24 qubits using GPU-accelerated quantum simulations on NERSC's Perlmutter system 🔹 Achieved 99.7% quantum computation ratio for largest systems (24 qubits) 🔹 Reduced 4-qubit HeH+ ion system from 16 possible states to just 4 states through charge preservation 🔹 Demonstrated successful convergence for Schwinger model calculations at large mass Why It Matters: Understanding excited states of quantum systems is crucial for advancing everything from drug development to new materials design. While classical computers struggle with these calculations due to the exponential scaling of quantum states, quantum computers offer a natural way to simulate these systems - if we can make the algorithms efficient enough for today's noisy quantum hardware. 🎯 Read the full paper here: https://lnkd.in/dKVC_M7q

Why do we use NVIDIA GPUs to accelerate quantum algorithms on the BlueQubit platform? Quantum simulations require massive parallel computations for matrix and tensor operations. While traditional CPU approaches struggle, GPUs can handle thousands of calculations simultaneously, enabling: 🔷 Universal state-vector simulations up to 36 qubits 🔷 Multi-node GPU simulations pushing beyond traditional limits 🔷 Tensor network simulations handling 100+ qubits in some cases While CPUs have to process these sequentially, GPUs can handle thousands of calculations simultaneously - making them perfectly suited for quantum computing problems as a stepping stone to run. Additionally? There's positive environmental impact! 🌱 As researcher Steven Thomson notes in the NVIDIA Developer blog on "Simulating Quantum Dynamics Systems with NVIDIA GPUs": "Without GPUs, simulations would have taken tens or hundreds of times longer to run... [coming] with a huge environmental cost due to the energy required." GPU acceleration enables entirely new areas of quantum research while reducing computational energy costs. ⚡ Sources: 🔷 NVIDIA Developer Blog (October 16, 2024) - https://lnkd.in/dMm783_x 🔷 BlueQubit BlueQubit Offers Managed Quantum Simulators, Powered by NVIDIA Technology Blog (March 18, 2024) - https://lnkd.in/dAFZPEE4