Computer Sciences

Light-based computing strategy to reduce the amount of power required to mine cryptocurrencies

A new light-based computing method that uses a photonic integrated circuit to save energy for cryptocurrency and blockchain applications has been developed by researchers. Mining cryptographic forms of money like Bitcoin — a course of confirming exchanges and adding new digital currency to the blockchain — consumes up to 1% of the world’s energy. As applications based on cryptocurrencies and blockchain technology become more commonplace, it is anticipated that energy consumption will rise.

Cryptographic forms of money are advanced monetary standards made using encryption calculations. A blockchain—a type of digital ledger that records information like transactions in a way that is hard to change or hack—is necessary for these alternative currencies.

“Currently, cryptocurrency mining is only accessible to those who have access to highly discounted energy—below $0.05/kWh,” stated the study’s first author, Stanford researcher Sunil Pai, who is currently employed by the quantum computing company PsiQuantum. Our low-energy chips will make it workable for people all around the world to productively partake in mining.”

“Our approach to photonic blockchain could also be used for applications other than cryptocurrency, such as securely transferring data for medical records, smart contracts, and voting. This breakthrough paves the door for low-energy optical computing, which has the potential to cut data center energy consumption.”

Author Sunil Pai, who performed the research at Stanford.

The researchers describe their new LightHash scheme, which creates a photonic blockchain by utilizing a photonic integrated circuit, in Optica. The researchers anticipate that this strategy, if implemented on a large scale, could reduce energy consumption by approximately ten times compared to the most advanced digital electronic processors of today. Along with Shanhui Fan and Olav Solgaard, David A.B. Miller co-led the Stanford University research team.

Pai stated, “Our approach to photonic blockchain could also be used for applications beyond cryptocurrency, such as securely transferring data for smart contracts, voting, and medical records.” This work makes ready for low-energy optical processing, which, at last, can diminish server farms’ energy utilization.”

Greener mining with silicon photonics
Developing worries about the enormous measure of energy expected to mine cryptographic forms of money have caused a few famous ones, for example, Ethereum, to move to problematic and possibly uncertain plans to limit their carbon footprint.

To find a more eco-accommodating methodology while maintaining an elevated degree of safety, Pai and partners use silicon photonics to lessen the energy necessities of digital money organizations. LightHash is an improvement on the team’s previous scheme, HeavyHash, which is used in cryptocurrency networks like Optical Bitcoin and Kaspa.

Pai stated, “HeavyHash’s high sensitivity to hardware error was the primary motivation for LightHash.” We designed LightHash to maintain all of HeavyHash’s security features while enhancing its error resistance because analog computers, including photonic ones, struggle to achieve low error rates.

The majority of Bitcoin’s energy consumption comes from the computation of a hash function like SHA256 or Heavyhash, which transforms input data into a single output number in a manner that is too complex to undo. This is necessary for both the secure creation of Bitcoin and the operation of its computing network. Heavyhash was modified by the researchers for use with a silicon photonic chip that was co-designed and contains a 6×6 network of programmable interferometers in the new work. Lighthash’s main computation, matrix multiplications, could now be processed optically with low energy thanks to this.

The researchers constructed an optical rig to control and track the propagation of light by tuning heating elements and imaging grating spots onto an infrared camera in order to evaluate the viability of using LightHash for matrix multiplication. Additionally, they established criteria for scaling the technology’s feasibility and implemented an algorithm for error mitigation.

Precision and lower power consumption in computation The silicon photonic chip’s experimental results were comparable to those derived from simulated error predictions. According to Pai, “Our results suggest that current silicon photonic chip technology can be feasibly used to compute LightHash at scale.” In essence, we have created a method for performing multiplications with analog optical circuits at a precision sufficient for use in a digital encryption scheme while consuming almost no power.”

It is necessary to scale LightHash up to 64 inputs and outputs in order for it to demonstrate significant advantages over digital alternatives. By developing low-power electromechanical tuning elements and energy-efficient converters for converting optical signals into electrical signals, the researchers are also working to further reduce energy consumption.

According to them, the new chip could help make photonic neural network training and application more energy efficient than conventional digital implementations because it accelerates matrix multiplication, the most computationally intensive operation for AI applications.

“It will be fascinating to perceive how cryptographic money innovation advances and how much photonics can add to the undeniably standard job of decentralized records in the public arena today,” said Pai.

More information: Sunil Pai et al, Experimental evaluation of digitally-verifiable photonic computing for blockchain and cryptocurrency, Optica (2023). DOI: 10.1364/OPTICA.476173

Topic : Article