CPA Practice Advisor

NOV 2018

Today's Technology for Tomorrow's Firm.

Issue link: https://cpapracticeadvisor.epubxp.com/i/1051460

Contents of this Issue

Navigation

Page 14 of 33

NOVEMBER 2018 ■ www.CPAPracticeAdvisor.com 15 FROM THE TRENCHES about the same stage that mainframe computers were in 1963-1965 when mainframes (which are Von Neumann machines) were just getting the early compilers and applications were simple, but just starting to work. The difference between mainframes and Quantum Computers is that once building the hardware is practical and the compilers are working properly, the additional speed available will revolutionize Cloud computing providing the horsepower to run applications radically faster. Conversion to Quantum Computing should predominantly mean recompiling applications to take advantage of the new platform. Quantum Computing is in the Early Stages Intel CES 2018 announcements: ■ Intel Labs has developed a neuromorphic research chip, code-named “Loihi,, which is a Quantum Computing enabled artificial intelligence (AI) algorithm inspired by how the brain works. This could unlock exponential gains in performance and power efficiency for the future of AI. ■ Intel’s Mike Mayberry, corporate vice president and managing director of Intel Labs said, “We expect it will be five to seven years before the industry gets to tackling engineering-scale problems, and it will likely require 1 million or more qubits to achieve commercial relevance.” ■ As Quantum Computing matures over the next 10 years, it could threaten blockchain and most cryptography that secures data and digital signatures. Let's Consider Encryption Quantum Computing approaches break encryption. ■ Shor’s algorithm can easily find public key integer factors. This ability would allow a quantum computer to decrypt many of the cryptographic systems in use today. ■ Consider a problem that has these four properties: The only way to solve it is to guess answers repeatedly and check them, ■ The number of possible answers to check is the same as the number of inputs, ■ Every possible answer takes the same amount of time to check, and ■ There are no clues about which answers might be better: generating possibilities randomly is just as good as checking them in some special order ■ For problems with all four properties, the time for a quantum computer to solve this will be proportional to the square root of the number of inputs. Triple DES and AES become easy targets, which are the most popular encryption algorithms today. This stuff gets a little heady, and I certainly wouldn’t want you to think I have mastered all the principles. Like all the emerging technologies we have covered in these emerging technology columns, Quantum Computing has pros and cons. On the positive side: ■ Computing using quantum-mechanical phenomena, such as superposition and entanglement which permits representing much more data in a single position, called a qubit ■ Google says 2018 and IBM says within five years they will achieve quantum supremacy, that is running faster than a classical computer ■ Just 50 qubits can represent 10,000,000,000,000,000 numbers. A classical computer would require a petabyte-scale memory to store that number On the negative side: ■ It is hard currently to scale physically to increase the number of qubits ■ Qubits can be initialized to arbitrary values and read easily ■ Quantum gates need faster decoherence time, currently optimized by cryogenic temperatures, 20 Millikelvins = - 459 Fahrenheit While the current prototype hardware is working, the number of qubits in use is quite small and the current designs require elaborate computing. Note from this picture how IBM is cooling a 56-Qubit design: Why Limits are being approached in the current Von Neumann architectures produced by Intel, AMD and others around the world. The silicon architectures have been miniaturized to 7 nanometers (nm) by TSMC and Intel is currently attempting to transition from 14nm to 10nm. It is becoming more difficult to build processing chips based on these small distances or geometries. Further, the tricks that engineers have used including hardware virtualization, multiple cores and threads have made the chips faster, too. However, there is no significant breakthrough in the engineering thinking that will improve performance notably. While I’m hopeful that some smart engineer will come up with a way to make chips faster and smaller there is nothing that is being reported in the engineering trade journals that shows any promise. While I’m sure significant breakthroughs would be kept secret, most of the methods over the last few years have been evolutionary, not revolutionary change. A breakthrough is needed, and Quantum Computing may be that breakthrough. Recommended Next Steps Quantum Computing is in its very early stages. This technology could prove to be a dead end, but all indications are that we are beyond that stage. The biggest issue is that hardware gates need to be designed that allow a faster decoherence time that operate at room temperature rather than in cryogenics. We’ve tried to give you enough background that you can understand some of the fundamentals of Quantum Computing. While there is no practical choice for most of you today because of the high cost and special environment required by Quantum Computers, you should watch for breakthroughs over the next 5-10 years which will enable your applications to run faster than is possible today. Key Information TECHNOLOGY: Quantum Computing Why is the new technology better? It is a new method of running programs, potentially at higher speeds How can you do this today? D-Wave, IBM Q, Google, Intel, Microsoft, Rigetti Forest Risks: Computation may not be faster Where/when to use: Problems that have to handle vast amounts of data How much? Commercial units are $15M+ for 2,000 qubits When expected in mainstream: Ten+ years Displaced technology or service Traditional computers, mainframes Other resources Waterloo Quantum 101, How QC Work Here's a summary of what you need to know about Quantum Computing:

Articles in this issue

Links on this page

Archives of this issue

view archives of CPA Practice Advisor - NOV 2018