Welcome to another weekly recap, where we look at key updates in quantum computing, as the industry continues to move forward at a pace. Today, we follow fresh steps in scale and error control. First, Rigetti Computing (RGTI) has advanced its chiplet design, then Terra Quantum shows a new way to cut errors, and next, Nvidia (NVDA) adds tools for faster codes. We also note new cloud features from Amazon (AMZN) and fresh results from Quantum Computing Inc. (QUBT).
Elevate Your Investing Strategy:
- Take advantage of TipRanks Premium at 50% off! Unlock powerful investing tools, advanced data, and expert analyst insights to help you invest with confidence.
Rigetti Computing: Scaling with Chiplets
We start this week with Rigetti Computing (RGTI), which has reported progress in its chiplet-based design. The company introduced Cepheus-1-36Q, a 36-qubit system built on a modular layout. The design doubles the performance of its earlier Ankaa-3 platform, cutting error rates by half and reaching a median gate fidelity of 99%. Rigetti says it plans to scale further, with a 100-plus qubit system targeted before the end of 2025. The company also strengthened its finances with a $350 million equity raise, bringing cash to $571.6 million and leaving it with no debt. Rigetti aims to use the funds for internal work, new deals, and future growth.
Terra Quantum: New Error Correction Method
Next, we move to Terra Quantum, which announced a method called QMM-Enhanced Error Correction. The new layer improves error handling on today’s hardware without changes to the design. Tests on IBM (IBM) processors showed up to 35% lower error, with logical fidelity raised to 94%. The method works with fewer qubits than standard codes, making it more efficient for early systems. Terra Quantum says the design can also boost quantum machine learning by cutting training loss and improving run stability.
Quantum Computing Inc: Wider Loss, Stronger Balance
Quantum Computing Inc (QUBT) posted a wider quarterly loss of $36.5 million, with revenue falling to $61,000. The drop came from higher operating costs and a non-cash warrant charge. However, the company’s cash position rose to $348.8 million after a private placement. Additionally, QCi opened a photonic chip foundry in Arizona, reported new sales in sensing and AI, and gained index inclusion in the Russell 3000 and 2000 (IWM). The company also secured deals with NASA and shipped devices for research use in Europe and Asia.
Nvidia: CUDA-QX Update
Moving on to the mega-cap companies, first is Nvidia, which has released CUDA-QX 0.4, adding tools for error correction and new solvers. The update includes a generative AI model that designs circuits adaptively, plus a GPU-powered tensor network decoder for faster and more accurate work. The company says the framework now allows researchers to design, simulate, and test codes all in one environment, backed by its cuQuantum libraries. The release is open-source and tuned to work across both simulation and hardware.
Amazon: Faster Braket Runs
We finish with Amazon expanding its Braket service with “program sets.” The feature lets users bundle up to 100 circuits into one task, cutting run times by up to 24 times. It also reduces costs since the per-task fee is paid once per batch instead of per circuit. Program sets are now live on Rigetti and IQM devices, as well as on Braket’s simulator, with more platforms expected to follow.
We used the TipRanks’ Comparison Tool to line up some of the top quantum stocks with a few names mentioned in this piece. It’s an easy way to see how they compare and get a feel for where the space might be going.
