Mach42 has raised £7 million in a pre-Series A funding round to accelerate the development and commercial adoption of its AI-powered semiconductor simulation technology. IP Group led the round, with participation from existing investors BGF and Foresight Group.
The company plans to use the funding to expand its engineering and business development teams. Mach42 is also seeking to make its platform more widely available to semiconductor companies and electronic design automation providers.
Mach42 develops machine learning technology that accelerates the simulation and verification of analog and mixed-signal circuits. The company says its tools can provide semiconductor engineers with up to 100 times greater verification coverage.
Verification allows chip designers to determine whether a circuit will perform accurately and reliably before it is manufactured. Finding problems during the design process can prevent expensive chip redesigns and reduce delays after fabrication begins.
Analog circuits process continuously varying electrical signals and are used in areas such as power management, sensing and communications. Their behavior can be more difficult and computationally expensive to simulate than many digital circuits.
Mach42 uses advanced machine learning to create high-accuracy surrogate models of complex circuit behavior. These models are designed to reproduce the results of more computationally intensive simulations while requiring less time and computing capacity.
The company’s technology works alongside SPICE simulators, which are widely used to analyze electronic circuits before manufacturing. Mach42 is not seeking to replace these established tools, but to accelerate workflows and increase the number of design conditions engineers can evaluate.
Its models can be generated from relatively limited datasets and exported in Verilog-A, SystemVerilog and C or C++ formats. This allows semiconductor teams to integrate the technology into existing electronic design automation systems and simulation environments.
Mach42 initially plans to focus on power management devices. These components regulate and distribute electrical power within electronic systems and are used across data centers, vehicles, consumer devices and industrial equipment.
Efficient simulation could help power management chip developers shorten design cycles, improve product quality and reduce development costs. Mach42 said these savings can reach millions of dollars for complex semiconductor programs.
The company believes its technology could also support agentic chip design workflows. In these systems, AI agents may perform more design, testing and optimization tasks with limited manual intervention.
Agentic design could significantly increase demand for simulation because AI systems can generate and evaluate far more design variations than human engineering teams. Mach42 expects semiconductor companies to require faster simulation tools as these workflows become more common.
The company is initially targeting power management applications but sees broader potential across the analog semiconductor market. Mach42 estimates that the wider market represents approximately $100 billion across multiple device categories.
The new funding follows early support from Parkwalk, which is part of IP Group. IP Group said its participation reflects the firm’s strategy of supporting science and technology companies from early research through commercial expansion.
KEY QUOTES:
“More complex semiconductor designs are putting real pressure on engineering teams to shorten development cycles without compromising accuracy. Mach42 is addressing that challenge in analog circuit verification, an area where better tools can have a clear impact on productivity and speed to market.”
“This is the epitome of the Group’s business model with IP Group following on with balance sheet capital from Parkwalk’s early support of the business. We’re delighted to lead this round as Mach42 accelerates commercial adoption.”
Dr. Lee Thornton, Partner, Deeptech at IP Group
“The EDA industry has very high expectations of tool performance, having relied on highly accurate simulation for decades. Mach42 makes new agentic approaches to analog chip design viable.”
“Over the next two years we expect the EDA industry to see a surge in compute demand as agentic design flows take hold. Our tools work hand in hand with SPICE simulators, dramatically reducing design cycle times and improving IC design quality. We are successfully demonstrating our modelling capability on very large power management devices, and power management is our focus going forward.”
Tim Haynes, Chair of Mach42
“Mach42 has demonstrated that we can model highly complex non-linear systems with remarkable accuracy and flexibility. This is not a simple problem to solve and requires multiple disciplines working together at the leading edge of machine learning research.”
“We thank our investors and partners for their continued support at a pivotal stage of the company’s development. It enables us to expand access to our platform and push forward our mission to make fast, physics-accurate circuit simulation a reality for engineering teams worldwide.”
Paul Neil, COO of Mach42
