Exchange Ranking
Top Exchanges
24h Volume Ranking
Popularity Ranking
Exchange BTC Balance
Proof of Reserves
Decentralized Exchanges
Funding Rate
Funding Heatmap
Liquidation Data
Max Pain
Long/Short Ratio
Whale L/S Ratio
Binance/Okex/Huobi L/S
Bitfinex Margin L/S
ETF Tracker
Solana ETF
XRP ETF
Hong Kong ETF
Bitcoin Treasuries
Crypto Reversal
Ethereum Reserves
HyperLiquid Wallet Analysis
Hyperliquid Whale Watch
Large Transactions
On-chain Movement
Bitcoin ROI
Stablecoin Market Cap
Options Analysis
News
Articles
Economic Calendar
Features
Wallet
Contract Calculator
Security
Collections
Watchlist
FollowingGenerative models significantly shorten quantum circuit tuning time
IonQ shares fell slightly on Wednesday as new research demonstrated faster methods of quantum-optimized circuit design. IONQ shares closed at $36.88, down 0.45%, after falling below the reference price of $37.05. The research was jointly conducted by IonQ, Oak Ridge National Laboratory (ORNL), NVIDIA and the University of Tennessee.
Generative models reduce quantum circuit tuning time
This study tested a generative model that was able to create quantum-optimized circuits without repeated parameter tuning. Traditional methods require researchers to run circuits, measure results, adjust settings and repeat the process. As researchers increase the size of the quantum problem, the cost of this cycle increases significantly.
The team trained the Transformer model using circuits that produced near-optimal results in early testing. The model then directly generates candidate circuits for each optimization subproblem. The researchers simulated ten candidate circuits and selected the strongest result for the next global solution update.
Tests have shown that the time to find the circuit remains around 28 seconds on previous test problems of different sizes. The traditional method has increased from about 34 seconds for 4 qubits to more than 11 minutes. Therefore, as the scale of quantum circuits expands, generative methods reduce the tuning burden.
ORNL-led research tested 100-variable optimization problems
该项目由橡树岭国家实验室（ORNL）牵头，参与研究的还有来自IonQ、英伟达和田纳西大学的科研人员。该研究考察了一个包含100个决策变量的密集高阶基准测试。结果显示，随着研究人员增加量子子问题的规模，解决方案的质量大致翻了一番。
比较的重点在于两种量子电路生成方法，而非量子计算与经典计算的对比。在基准测试期间，这两种方法都使用了相同的加速计算基础设施。这种设置使研究人员能够隔离主要与电路生成过程相关的差异。
然而，研究人员是对每个电路进行模拟，而不是在物理量子处理器上运行它们。测试过程中通过CUDA-Q平台使用了英伟达的cuQuantum软件。这项工作是在橡树岭Defiant2计算系统中的一个英伟达H200 GPU上运行的。
IonQ研究为量子优化推进提供背景支持
该研究为将生成式模型与混合量子优化相结合提供了基准规模的证据。混合方法在组合结果之前，将大型优化问题划分为较小的部分。较大的部分可以提高答案质量，但传统的调优方法可能会急剧增加计算需求。
研究人员设计的新方法旨在消除大部分重复的优化过程。这种方法可以在不显著增加寻找电路时间的情况下，支持更大规模的量子子问题。然而，目前报告的工作仍集中在模拟基准测试上，而非商业量子工作负载。
IonQ在多伦多的IEEE Quantum Week 2026大会上展示了这项研究。该论文成为9月13日至18日举行的活动中被接受的九篇IonQ研究论文之一。它还获得了最佳论文奖，因为研究人员正在继续测试更大的科学和工程应用。
