If you read and follow what people say about quantum computing, you quickly realize that there’s a considerable gap between the almost unimaginable scale that quantum computing will enable and what today’s quantum computers can actually do. So when our friends and business partners at Dynex reported that their quantum simulation platform offers the equivalent of thousands of qubits of computing power—compared to the tens of qubits offered by other publicly available quantum simulation platforms—we were both excited and skeptical.
If Dynex's claims are true, then it's conceivable that their platform could be used outside the lab to solve real-world problems. But is it really possible for Dynex to deliver a level of capability far beyond what would otherwise be possible? In our opinion, the only way we could answer this question was to conduct a controlled test ourselves.
Dynex claims that their quantum simulation platform can execute quantum algorithms at a scale and performance far exceeding that of other publicly available quantum computing platforms. We have now completed this research, and I can confirm that we successfully replicated the results reported by Dynex in all six test cases.
That being said, I kindly ask you to be cautious with my words. Having worked with emerging technologies for over 30 years, I'm wary of technology "breakthroughs." Therefore, we're willing to provide our complete test suite (including the actual test scripts) to anyone who wishes to verify that they can reproduce similar results.
We also have an agreement with Dynex to allow them access to their platform for this type of verification.
As those who follow FSE know, our core principle is that emerging technologies create value only when they enable businesses to achieve previously unattainable goals. This presents a higher bar for Dynex; they must not only demonstrate superior performance to other quantum platforms, but also demonstrate that their platform's capabilities are comparable to or even superior to those of traditional computing solutions when solving the same real-world problems.
Even by this higher standard, Dynex has demonstrated promising and meaningful results across multiple high-value use cases that deserve further investigation.
Because the report is 26 pages long, only some of its contents are excerpted due to space limitations:
0. Action Plan
In early 2025, Dynex Corporation will make its quantum simulation platform generally available. This platform, running on a classical compilation infrastructure based on CPUs and GPUs, provides users with the ability to model and execute gate-based and annealing quantum algorithms. Due to its diverse modeling approach, Dynex asserts that its platform can run true quantum algorithms with capabilities and performance levels far exceeding those achievable using other publicly available quantum computing platforms.
Another key claim made by Dynex concerns the relationship between performance and complexity. For most quantum algorithms, every time you add a factor, your computational complexity increases exponentially. Dynex asserts that their platform's performance increases sub-exponentially. In other words, the increase in runtime from adding factors is less than the increase in complexity from adding those factors.
To provide evidence and independent verification of these assertions, Dynex asked its integration partner, Finserv Experts (FSE), to conduct an independent assessment of its platform’s capabilities. FSE agreed to this assessment on the condition that it make the testing tool (including the source code for the test itself) publicly available so that anyone interested could fully replicate the test results.
FSE and Dynex have jointly agreed upon six scenarios, each of which represents a common reference case for expressing quantum capabilities.
And each has potential real-world value as the basis for this assessment.
(1) n-bit adder
(2) Traveling Salesman Problem
(3) Reverse circuit execution
(4) Shor's algorithm
(5) Protein folding
(6) Reverse hashing using Grover's algorithm
For each test scenario, FSE attempts to answer two questions:
[1] Is it possible to independently verify Dynex's results?
[2] Are there potential real-world use cases where the capabilities of the Dynex Platform can be used to deliver solutions with measurable value?
After completing the evaluation, FSE can confirm that it can independently achieve the same performance levels claimed by Dynex in all six cases. FSE can also confirm that Dynex's assertion of sub-exponential scaling appears to be correct, at least within the range measured in this evaluation.
Definitive answers to the second question require more than simply evaluating the reproducibility of Dynex's performance metrics. These metrics may clearly outperform those of other publicly available quantum platforms, but to have real-world value, they must also demonstrate advantages over classical computing solutions that achieve similar goals. Full confirmation of such advantages will require further research, but there is sufficient evidence for several use cases to warrant further investigation.
1. Opportunity Cost and Motivation
Any investment of time has an opportunity cost; this is especially true for a service company like ours.
Because our time is our product. Finserv experts choose to invest their time for two reasons:
(1) Dynex has selected Finserv Experts as their integration partner. In order to deliver solutions based on the Dynex platform to our customers, we needed to come in with confidence that we could replicate the level of performance and functionality that Dynex claims.
(2) Our company’s core mission is to help businesses derive measurable value from emerging technologies. Quantum computing is one of the emerging technologies in which we have chosen to develop capabilities, but for the most part, quantum technology has not yet matured to the point where it can begin to deliver real-world business value. If Dynex’s claims are true, then it is possible to deliver solutions with real-world value now, even though Dynex currently offers a quantum simulation platform.
2. Introduction to Dynex
The Dynex platform introduces a new approach to quantum computing by using ordinary differential equations (ODEs) to simulate the dynamics of physical systems. This approach allows Dynex to create what it calls "algorithmic qubits." Its goal is to provide a scalable and efficient way to simulate quantum behavior on classical hardware.
Most existing simulators tend to rely on solving partial differential equations (PDEs) such as the Schrödinger equation; they quickly hit a wall due to exponentially increasing complexity. Dynex designed its ODE-based approach to scale sub-exponentially with complexity, unlocking performance that is difficult to achieve with existing simulators.
For comparison: most leading quantum simulators, such as IBM’s Aer Simulator or Google’s Quantum Tensor Network Simulator, struggle to handle more than 30-40 qubits before exhausting practical computing resources. In terms of hardware, today’s largest physical quantum computers, such as IBM’s Osprey, currently operate at around 433 qubits.
Dynex claims—and this white paper attempts to verify—that their platform can process the equivalent of thousands of qubits of data on classical GPU infrastructure, potentially delivering solutions that are orders of magnitude more complex.
3. Test Case 1: n-bit adder circuit
We examined public benchmarks provided by Dynex, where the team computed the exact same quantum circuit on both the IBM AerSimulator and IBM’s quantum platform (127-qubit Eagle r3 QPU, ~30K CLOPS). These results, publicly accessible in the Dynex SDK repository (https://github.com/Dynexcoin/DynexSDK/tree/main/circuit_scaling_benchmark), show the practical limitations of both approaches.
On the IBM AerSimulator, the limit is the number of qubits that can be simulated before resource requirements grow exponentially, preventing scaling beyond a few dozen qubits.
On the IBM Quantum platform, the bottleneck is error correction: despite the physical qubit count of 127 on the Eagle r3 device, noise and decoherence limit the circuit depth and fidelity, making it impractical to scale the same computation beyond small problem sizes.
To further clarify, Dynex also released a side-by-side comparison video illustrating these findings.
(https://github.com/dynexcoin/DynexSDK/tree/main/circuit_
scaling_benchmark).
Together, these control studies highlight the current limitations of leading simulators and available quantum hardware and provide a transparent benchmark against which alternative approaches, such as Dynex's ODE-based architecture, can be evaluated.
The results are shown in the figure below. The Dynex quantum simulation platform can successfully process numbers up to 100 digits in less than 13 seconds.
analyze
Can Dynex assertions be independently verified?
Yes. We were able to successfully and independently replicate the results reported by Dynex.
Can the current level of business capabilities be used to deliver real-world business value?
For some use cases, yes.
What is being measured?
Quantum adders on public platforms can only process numbers up to 26 or 64. The Dynex platform has demonstrated that it can process numbers up to 2100: 1,267, 650, 600, 228, 229, 401, 496, 703, 205, 375.
Observe the running time.
Execution time remained practical across the entire range tested, increasing only from approximately 6 seconds for 6 bits to approximately 13 seconds for 100 bits. As discussed in Section 3 above, one of Dynex's key claims is that its approach supports sub-exponential performance scaling relative to complexity. At least within the range tested, the evidence appears to bear this out.
Background and Applicability
The 13-second processing time for a 100-bit number far exceeds results reported for other quantum platforms, but it is still non-trivial, meaning that real-world solutions must specifically exploit this power, rather than relying on brute-force computational power.
On its own, a quantum adder’s 100-bit threshold might be large enough for targeted use in portfolio analysis, fraud detection, or supply chain planning, but still slightly below the requirements for drug evaluation, cybersecurity, or climate modeling.
4. Test Case 2: Traveling Salesman Problem
Conclusion: The Dynex platform can handle TSP problems that are much larger than those handled by the IBM and IonQ platforms.
What is measured
This evaluation focused on the Traveling Salesman Problem (TSP) as an annealing problem. Current quantum annealers, such as DWave's Advantage system, are limited to around 5,000 physical qubits. In practice, this limits the size of solvable TSP instances to relatively modest city counts due to embedding overhead and noise. In contrast, Dynex successfully executed an 80-city TSP mapped to 6,400 qubits, significantly exceeding the effective hardware limit of the leading commercial annealer.
Background and Applicability
The study compared Dynex to a quantum annealing platform, rather than to state-of-the-art classical TSP solvers. Classical systems have achieved impressive feats—for example, the Concorde solver successfully solved an 89,000-city TSP, but only after running it for over a year at peak CPU load.
Therefore, an open question is whether there is a useful range of TSP complexity in the real world where Dynex can match or outperform classical computers without requiring the lab conditions, specialized hardware, or runtimes that would make them unusable for everyday practical purposes. Given the potential impact of successful TSP optimization across a wide range of high-impact industries, we believe this warrants further exploration.
- Reverse circuit execution
Conclusion: Dynex was able to successfully perform reverse circuit execution for numbers up to 1,028,171, significantly exceeding results obtained with gate-based or annealing platforms.
Can the current level of business capabilities be used to deliver real-world business value?
For some use cases it may be, but further research is needed to determine this.
While performing reverse circuit execution on a classical computer is theoretically possible, it's practically impossible to implement on a large scale. This is because only a subset of classical logic gates are reversible (e.g., NOT and XOR), while most gates that are fundamental to modern computing (AND, OR, NAND) are inherently irreversible. Building a classical machine using only reversible gates is impractical and inefficient.
What is measured
Benchmark tests confirmed that Dynex can factor numbers up to 20 digits (≈1,028,171), scaling up to 400 qubits. By comparison, currently available quantum systems are limited to very small factors: IBM and IonQ demonstrated results at N=21, and D-Wave at N=35.
Background and Applicability
In this case, the Dynex platform demonstrated the practical ability to execute quantum circuits in reverse, using integer factorization as a reference workload. Compared to other tests, the increase in execution time across the measured range was closer to proportional scaling than exponential scaling. While factoring a 20-bit number is still small relative to encryption standards like RSA-2048 (617 bits), it is orders of magnitude greater than anything currently achieved on publicly available quantum hardware.
In terms of real-world applicability, it is difficult to comment on the usability of reverse circuit execution on the Dynex platform for many of the above-mentioned real-world use cases because there is no reference point to compare it to, as reverse circuit execution is rarely implemented on classical computing hardware.
Further research and prototyping are necessary to determine whether using the Dynex platform will have a meaningful impact on software development, hardware design, or improving fault tolerance. On the other hand, it may be worth exploring whether reverse circuit engineering on Dynex can be an effective debugging and error checking tool for quantum software development, as it may facilitate more rigorous error checking than is otherwise currently possible.
6. Shor's algorithm
Conclusion: Dynex was able to successfully factorize a 40-digit number (a 13-digit number, specifically N = 1,099,510,308,317) in less than a second. Interestingly, a shorter 20-digit number took over 2 seconds to factor.
Test case description
Shor's algorithm, first proposed by MIT professor Peter Shor in 1996, is a concise and elegant algorithm for finding the prime factors of any integer, even very large ones.
Shor's algorithm is simple to understand and implement. It has only six steps; someone with a bachelor's degree in mathematics and a basic understanding of Python can code it in less than an hour.
The problem with integer factorization isn't writing such a script, but running it. If you could combine all the computing power on Earth into a supercluster of galaxies and have that supercluster factor a 617-bit number (the size of an RSA2048 private key), it would take longer than the current age of the universe to run.
Because of its ability to process multiple possibilities simultaneously, quantum computing has the potential to make Shor's algorithm trivial.
That’s why it’s the single most cited example of how quantum computing can offer capabilities that classical computing could never accomplish.
Real-world use cases
While Shor's algorithm has many potential use cases, the most interesting and concerning aspect is its impact on network security. If Shor's algorithm could make breaking RSA-2048 trivial, it would render almost all modern encryption useless, meaning that nearly all existing software applications would be open to misuse, theft, and malicious attacks.
The industry takes this threat very seriously and is investing billions of dollars in developing and adopting new encryption methods (often called post-quantum encryption, or PQE) that will remain secure once quantum computing becomes feasible.
Control baseline
At the time of writing, the largest number that a quantum computer has successfully factored using Shor's algorithm is N = 21. In this representation, N represents the actual number being factored, so the solutions for N = 21 are 3 and 7.
Photonic Systems (2011):
Politi, A. et al. (2011). "Shor's quantum factorization algorithm on a photonic chip." Science
Science, 325(5945), 1221–1224. DOI: 10.1126/science.1173731.
o Details: Demonstrates factorization using a photonic quantum chip with 2-4 qubits. The implementation uses
A compiled version of Shor's algorithm where the modular exponentiation operation is pre-simplified, thus reducing circuit depth.
The success probability is about 90%, slightly lower than the experimental results.
Now step.
o Relevance: Confirmed as a milestone for photonic systems, but scalability is limited by photon losses and gate fidelity.
Ion Trap System (2012):
Martín-Lopez et al. (2012). "An experimental implementation of Shor's quantum factorization algorithm in nuclear magnetic resonance." Nature, 486, 195-199. DOI: 10.1038/nature11111.
o Details: Decomposition is performed using a 7-qubit nuclear magnetic resonance quantum system excited by an ion trap.
The overmodular exponentiation operation is heavily optimized and simplified for specific cases (, Issue 2). Classical preconditioning reduces the quantum operation to a few controlled gates.
o Correlation: verified as the largest factor for ion trap systems, but due to its dependence on specific
The method cannot be generalized due to the numerical attributes of .
Superconducting qubit systems (2012–2016):
Lucero et al. (2012). "Computing prime factors with a Josephson phase qubit processor." Nature Physics, 8719–723. DOI: 10.1038/nphys2385.
o Details: Factoring was performed using a 4-qubit superconducting processor (Google/UCSB team). A later experiment
The experience (not fully published, but mentioned in the talks) was extended to a similar platform with
Has 5-6 qubits, uses compiled Shor's algorithm and precomputed modular operations.
o Relevance: Shows superconducting qubit implementation, but with simplifications that limit generality.
Comprehensive review (2020-2023):
Amico, M. et al. (2020). "The status of quantum computation of integer factorization." Quantum Science
and Technology, 5(3), 033001. DOI: 10.1088/2058-9565/ab8b6d.
o Details: Review the experimental progress and point out that cross-platform (photon, ion trap, superconductor) physical hardware is the most
It is emphasized that all demonstrations use custom circuits and are hard-coded for specific situations.
Larger numbers (for example) are considered in analog or classical quantum hybrids, while
Not pure quantum hardware.
o Correlation: Confirmed as a practical limit due to NISQ constraints (qubit count ~5–10, gate error ~1%).
Recent Tutorials and Community Insights (2024–2025):
Qiskit Tutorial: Shor's Algorithm.
ibmq_nairobi etc. 7 qubit system). Note that due to the circuit depth (approx.
20–30 gates) and noise, is the maximum demonstrated on the hardware.
Quantum Computing Stack Exchange (2023): "The largest number that Shor's algorithm can factor on real hardware is
The consensus points out that there is no other way to quote the photon and superconductor experiments due to coherence limitations.
There are solid demos.
Can the current level of business capabilities be used to deliver real-world business value?
While the Dynex quantum simulation platform clearly outperforms published results from other publicly available quantum platforms, it is still a far cry from what can be achieved by running other factorization methods on classical computers.
What is measured
The evaluation used Shor's algorithm, implemented as a quantum gate circuit for integer factorization. Current quantum hardware is still limited to very small numbers: IBM and IonQ have reported results for N=21, while DWave has demonstrated results for N=35. In contrast, Dynex successfully executed short circuits for N=1,099, 510, 308, and 317 in less than 1 second.
Background and Applicability
Dynex has shown that it can factor a 13-digit number in less than a second, which is very impressive when you consider that other quantum platforms can only factor numbers of 2 digits at most. This result is not yet achievable on specialized classical hardware; algorithms such as Pollard's Rho can already factor numbers of 100 to 130 digits in a similar time frame on a single CPU.
Even so, the results obtained by Dynex still warrant further exploration and experimentation. The scaling curve along the measurement range is quite different from what we have seen on other quantum platforms, low enough to make us wonder what the current Dynex platform might be capable of.
7. Protein Folding
Conclusion: Dynex was able to successfully use a 2D lattice-based optimization approach to predict the final shape of a 77-amino acid chain protein in 26 seconds.
What is measured
The mathematical complexity of this problem ranges from a small number of operations for 3 amino acids to about 1.5×10 for 77 amino acids.
Despite the exponential increase in theoretical complexity, the runtime of Dynex increases only slightly: from 0.02 seconds for very small systems to 25 seconds for Swiss-Prot. This demonstrates a sub-exponential scaling curve within the measured range.
Background and Applicability
While there are reference studies focusing on using this mathematical technique for protein final state prediction on classical computing platforms, no two of these experiments are set up in the same way; further controlled experiments are needed to evaluate this.
The Dynex approach does appear to outperform or at least match the performance of alpha folds (depending on the choice of hardware), but asserting that Dynex is a potential real-world option also requires comparing it under consistent experimental conditions with other optimized 2D lattice-based implementations implemented purely on classical computing platforms (i.e., without quantum simulation).
- Reverse hashing using Grover's algorithm
Results: The Dynex platform has been able to successfully achieve reverse hashing up to 128 bits.
Test case description
Grover's Algorithm Explained
Grover's algorithm is a search method that exploits the ability of a quantum computer to perform multiple versions of the same operation simultaneously.
Imagine having to choose a specific possibility from a large number of possibilities, which we will call N. These possibilities do not have any order, so we have to test them one by one, which means that on average, we have to evaluate about half of the possibilities, or N/2, before we find the one we are looking for.
Grover's algorithm is a way of increasing the probability that each time we choose one of N possibilities to evaluate, we choose the correct one. This is called a quadratic speedup; the average number of possibilities we have to examine goes from N/2 to N (the square root of N).
So, for example, if we have to use regular search to pick the right choice out of a million possibilities, we have to look at, on average, 500,000 possibilities before we find the right one; if we can use Grover's search, we only have to look at 1,000 possibilities on average before we find the right one.
Reverse hashing explained
Hashing is the operation of converting one set of numeric or character data into another set of data. What makes hashes unique and useful is that they are irreversible; if you only have the converted data, there is no way to calculate what the original data was; you must go through each possible value of the input data one by one and apply the hash function to it to see if it is the correct value.
Reverse hashing is exactly that process: trying to figure out the original value before hashing it. Reverse hashing is practically impossible for any large dataset because there are so many possibilities, but Grover's algorithm can (at least in theory) make it easier by increasing the probability that you'll choose the right option and drastically reducing the number of options you have to check.
Real-world use cases
Like Shor's algorithm, the most important use case of Grover's algorithm is in the field of network security.
Hashing is the primary method used to store passwords. When a user creates an account on a website or software product, the password they create is hashed before being stored. Later, when the user attempts to log in using that password, the user's input is hashed again, and the hashed input is compared to the stored hash to see if the user entered the correct password.
A hash cannot reconstruct the original password; you have to keep guessing until you get it right. For very long passwords, the chances of guessing right are almost zero, but Grover's algorithm can drastically reduce the number of guesses you have to make to get the right answer.
Hashing is also the method used to similarly secure digital signatures and entries in a blockchain or distributed ledger.
Control baseline
Testing a reverse hash function on existing publicly available quantum platforms has not yet been demonstrated. Grover's algorithm has so far only been implemented as very small proof-of-concept experiments—typically in the range of 2 to 8 qubits—which is insufficient to test even the simplest hash functions in practice.
To date, no quantum computing company has released a useful Grover implementation that operates at a scale relevant to real-world applications. Public demonstrations by IBM, IonQ, Rigetti, and others have been limited to toy circuits designed to prove theoretical principles, rather than solving meaningful reverse hashing problems.
Therefore, there is no established baseline from existing quantum hardware to which the Dynex results can be directly compared.
The Dynex benchmark thus represents the first attempt to exercise Grover-style circuits at a large enough scale (up to 128 bits) to test their potential relevance to real-world problems.
Can the current level of business capabilities be used to deliver real-world business value?
Possibly; further research is needed.
What is measured
The benchmark evaluated Grover's reverse hashing algorithm. While current quantum hardware is limited to very small test cases, Dynex executed Grover circuits up to 128 qubits (128 bits), significantly exceeding the capabilities of publicly available quantum platforms.
The complexity of the underlying problem increases exponentially across the measured range, from about 10 operations at 4 bits to about 10 operations at 128 bits.⁸ Despite this, Dynex’s runtime remained relatively flat across the entire range tested: about 15 seconds for 4-64 bits and only about 18 seconds at 128 bits, further supporting Dynex’s assertion that their platform offers more proportional than exponential performance/complexity scaling.
Background and Applicability
Modern hashing algorithms like SHA-256, combined with passwords of 8 or more characters, remain secure even in a fully capable quantum computer; even with Grover's quadratic speedup, there are still too many possibilities to test. However, there are many examples of legacy software and outdated hashing methods where, in principle, the capabilities demonstrated by Dynex could be used to expose vulnerabilities.
The purpose of this research is not to endorse attacks, but to identify when such capabilities may be accessed so that enterprises can proactively protect systems before they are exposed.
That said, there are many potential use cases beyond cybersecurity attacks where a Grover-style approach could add real-world value beyond the capabilities currently demonstrated by the Dynex platform. A few use cases worth further exploration might include:
(1) Data recovery: Locating information on damaged or degraded storage devices using earlier, smaller hashing methods.
interest.
(2) Protein design: Encode the desired protein properties as hash values and use reverse hashing to
The protein space can be efficiently searched to identify candidates that match these properties.
(3) Post-quantum cryptography research: Stress testing new cryptographic methods against Grover-style quantum searches to verify their resilience before large-scale deployment.
(4) Database search and optimization: Apply search technology similar to Grover to speed up the query of structured and unstructured data in large databases.
(5) Artificial Intelligence and Machine Learning: Explore Grover-based primitives for accelerating model search, hyperparameter tuning, or pattern matching.
(6) Error correction and verification: In large-scale distributed systems where classical checking methods become inefficient, reverse hashing is used as a tool for error detection.
9. Summary
Dynex has made some noteworthy claims about the capabilities of its quantum simulation platform, claiming it can significantly outperform most publicly available quantum computing platforms. If true, the Dynex platform will at least enable broader testing and prototyping of quantum computing software development techniques than is possible on publicly available quantum computing platforms. Beyond that, however, some of their claims, if true, raise the possibility that the Dynex platform can be used to deliver real-world solutions, even as a quantum simulator.
The purpose of our study was to answer two questions:
[1] Is it possible to independently verify Dynex's results?
The answer is absolutely yes.
We reviewed the test case implementation in detail, executed the scripts independently, and verified the results against known references. In each case, we received results similar to those reported by Dynex.
[2] Are there potential real-world use cases where the capabilities of the Dynex Platform can be used to deliver solutions with measurable value?
The answer is yes, but it is a more qualified yes.
In our study, Dynex appears to outperform publicly available quantum platforms in all six text cases, but these test cases necessarily refer to quantum algorithms; in some of these cases, traditional classical techniques still produce significantly better results (e.g., TSP).
However, there are some targeted use cases where the current Dynex platform has the potential to deliver results that exceed or compete with traditional computing techniques. These use cases may include, but are not limited to, portfolio analysis, supply chain optimization, or prediction of the final shape of proteins.
10. Next Steps
- Next steps for this research
The ultimate validation of any study of this kind is the reproducibility of its results. For this reason, we have agreed with Dynex to make the source code for this study available to anyone who wishes to run the scripts themselves and verify that they can reproduce the same results. The only condition we attach is that anyone who performs these tests share their findings with us afterwards. If you would like to verify these results yourself, please email us at
dynex_study@finservexperts.com .
As mentioned above, these initial results across many use cases demonstrate that the current capabilities of the Dynex quantum simulation platform are robust enough to add demonstrable value in real-world scenarios outside the laboratory. Because we are discussing a class of problems where quantum computing could potentially break new ground, it's no surprise that all of these scenarios represent highly complex problems that shouldn't be taken lightly. The next step will be to more robustly validate the Dynex platform's utility in these scenarios, ideally in partnership with public, private, or academic enterprises that already possess deep expertise in the relevant solution areas.
- What's next for Dynex?
The current version of the Dynex platform is software-based; it runs on CPU-based hardware. But even though this research has demonstrated significant improvements over published results from other publicly available quantum platforms, Dynex believes a hardware-based solution will offer even greater advancements.
To this end, Dynex is working to implement this approach on a dedicated silicon chip called Apollo, which Dynex says will run 1,000 physical qubits in a neuromorphic architecture operating at room temperature. Their roadmap beyond Apollo envisions a series of further versions that will increase this capability by multiple orders of magnitude, with a long-term goal of reaching 1 million physical qubits by 2034.
- Next steps for Finserv experts
Our mission at FSE is to help our clients design, develop, and deploy new business models that would not be possible without the new capabilities enabled by emerging technologies.
We are not quantum scientists and are not qualified to comment on whether the Dynex approach constitutes a “true” simulation of quantum computing, or whether their planned hardware release constitutes “real” qubits.
What we can comment on, what we are qualified to assess, ultimately what we really care about is evidence-based results. This study shows to our satisfaction that Dynex is fully capable of delivering the results they claim to deliver. Furthermore, there is reason to believe that the level of capability Dynex has demonstrated will have an impact on certain target use cases.
Our next step is to begin working with our customers to explore how these new features can be deployed to deliver new capabilities that were previously impossible. ————————————————Copyright Statement: This article is an original article by CSDN blogger "Aibit" and follows the CC 4.0 BY-SA copyright agreement. Please attach the original source link and this statement when reprinting. Original link: https://blog.csdn.net/u010876122/article/details/151117714
