Today's $9.5m funding round announcement is a significant moment for Mach42. It will enable us to take our world-leading physics-driven surrogate modelling technology from the AI lab to the semiconductor engineer’s desktop and will usher in a new era of advanced verification tools for the world’s most complex semiconductors.
Whilst the electronics world is seeing a plethora of Agentic based companies addressing digital chip design in a bid to reduce design cycles by 50%, the verification bottleneck for analog circuits threatens both time-to-market and product quality and remains a headache for design teams worldwide. We predict that over the next two years we will see a huge asymmetry appearing between digital and analog design without a big push in the analog design tools.
With our surrogate modelling technology, verification engineers can generate high fidelity surrogate models of analog circuits capable of executing up to 2,500 times faster than conventional techniques. Our models can be hosted and interrogated in the Discovery Platform, enabling either interactive or script-based verification of analog blocks. They can also be generated and instanced within the Cadence® Virtuoso environment, giving verification engineers access to our powerful surrogates within the industry-standard design environment.
Our tools can handle a broad range of complexity, from simple designs with a few tens of analog and digital ports, up to state-of-the-art analog subsystems in excess of a thousand ports. This gives designers the ability to use our capabilities both at the lowest level of the design hierarchy or using a top-down approach for the most demanding block verification tasks.
We’re also thinking about the broader future of EDA at Mach42. Agentic AI is delivering productivity dividends in the digital space today, and we anticipate its imminent arrival in the analog realm. Those productivity dividends are associated with the offload of design exploration or infrastructure tasks but still rely on the underlying compute engines – simulation, synthesis, place-and-route – to deliver the answers. Orchestration enables the selection of the “best” results, but the time to achieve those results is inexorably linked to those compute engines. For Mach42, there is a natural fit between the use of our surrogate models and the acceleration of the underlying simulation tasks. This is the focus that stimulated our most recent funding round.
We have a number of exciting innovations to reveal in the coming months, but if you’d like a sneak peek at our latest technology, come along to booth 659 at DAC in Long Beach next week and ask for a demo. Alternatively, if you fancy being part of the verification revolution, check out the vacancies on our jobs page. Finally, if you’d like to become part of the Mach42 community, please reach out to us at info@mach42.ai and sign up to our newsletter.
