Nandan Nayampally is Chief Commercial Officer of Baya Systems. He is a semiconductor executive with nearly three decades of experience in engineering, product and business roles. With an extensive background in processing, systems and software and in particular IP, after an initial start on Athlon processor team at AMD, he spent 16 years at Arm building the processor, fabric, vertical product businesses, and was General Manager for the CPU business and first GM of the Client business. He led Amazon’s 3rd party Alexa effort, and had key roles at Denali, and most recently at BrainChip prior to his role at Baya Systems, where he focused on bringing disruptive technologies to market.
Tell us a little bit about your background and what brought you to Baya Systems?
I have had nearly 30 years of experience in silicon design, and in particular processing. Being part of the original Athlon design team at AMD, to developing EDA tools IP at various startups, led me to a very successful career at Arm where I was able to help drive Arm processors, system IP and systems from traditional feature phones to smartphones and laptops all the way to large scale systems in automotive, infrastructure and datacenter. At most points in the last two decades, I have been lucky to have the opportunity to participate in disruptive trends, including, most recently, neuromorphic computing for AI.
What I did see however, was that there was a lot of emphasis on processing, but there was substantial gap emerging in data movement, which seems to be core challenge for AI, communications, data center and automotive markets. The need for an end to end solution of performance modelling, software and IP was clear.
However, it is a very tough market because very few experts have the vision, the technical chops and the execution muscle to do something disruptive. However, when I reconnected with Dr. Sailesh Kumar, CEO of Baya Systems, industry luminary Jim Keller and the Baya team, it was clear that this was a special team with a keen focus on doing just that. And 9 months in, I’m even more excited with how far we’ve come and where we plan to go.
What is Baya Systems for those who may not be familiar? What is its backstory?
Baya Systems started with a vision to enable semiconductor system design to efficiently transition to the chiplet era and match the exponentially increasing demands of intelligent compute (the tight integration of different compute elements such as CPUs, GPUs and accelerators that enable the diverse computational requirements of AI).
Our cofounders, Dr. Sailesh Kumar, Eric Norige and Joji Philip, previously worked together at NetSpeed and later at Intel following its acquisition. At Intel, they particularly focused on the Xeon family of processors, creating a SoC builder and other tools to address then-present challenges of scaling computational capabilities. With their unique perspective, they recognized that communication bandwidth between compute, memory, and input/output (IO), which was becoming a critical bottleneck to the advancement of AI and other data-intensive compute functions, could not effectively addressed by existing system architectures.
They started Baya Systems to offer a holistic, software-based approach to designing, building and analyzing these complex, highly performant multi-die systems. Today, Baya Systems has two products. The WeaverPro™ software, which does advanced performance modelling and system analysis enabling continuous refinement of system design from concept to post-silicon tuning. WeaverPro’s performance analysis capability, allows customers to visualize and test multiple design scenarios in real-time, allowing them to quickly make adjustments and reduce risk.
Baya’s WeaveIP™ portfolio delivers modern network-on-chip (NoC) needs for more efficient data movement. Unlike proprietary NoCs, Baya’s technology is protocol-agnostic and adaptable, making it compatible with a range of industry standards, including UCIE and AMBA—and also making its technology more accessible to a wider range of businesses. It’s unique in its ability to handle multiple protocols on the same transport, unifying complex fabric and making it more efficient. It is also designed from the ground up for the design and development of chiplets.
Our team is determined to enable seamless and efficient communication within computing systems to unlock their full potential.
What is the biggest challenge facing the semiconductor industry right now?
The biggest challenges facing the industry today are efficient data movement and scale of computation.
Due these advanced compute systems that are rapidly growing into from tens to thousands of compute nodes and accelerators in tight concert, the semiconductor industry is shifting away from traditional system-on-chip (SoC) designs to “system-off-chip” or “system-of-chips” models. Instead of packing all functions onto a single chip, manufacturers are now seeking to design systems composed of multiple specialized chips (or chiplets) within a single package. This allows for scaling performance while optimizing power and cost.
The resultant challenge, however, is the exploding complexity and performance readiness of these large systems. It requires a new development model, and IP to design, connect these chiplets seamlessly, and deliver the critical Key Performance Indicators (KPIs) such as high data throughput, minimum latency, reliability and safety while ensuring efficient memory and cache management. This is where Baya’s holistic solution comes in, combining advanced hardware IP with powerful software tools to solve these challenges.
To what do you attribute Baya Systems’ rapid growth and success since coming out of stealth last year?
Although we came out of stealth in June 2024, we began in March 2023, which gave us the advantage of taking the time to build best-in-class solutions before unveiling ourselves to the general public. The more important aspect is the domain expertise of the team that is doing ground up development of software driven system for the third time and has learned all the pitfalls and rapidly developed the framework for the product.
The inception of the company itself was driven by a key board member recognizing the widening gap in the market that the founding team could solve. With that mission, we’ve focused on building the best solutions for our customers’ challenges, and giving them capabilities that were previously inaccessible. Building the right, fit-for-purpose solution, without any legacy issues and thereby enabling us to scale quickly.
Nine months after inception, we completed our first license and product delivery. After we emerged from stealth, our growth accelerated and we secured multiple customers and partnerships. Over the past year, we’ve entered into a licensing agreement for our WeaveIP fabric with Tenstorrent, launched a combined chiplet-optimized Network on Chip (NoC) and Physical Layer (PHY) interconnect IP solution with Blue Cheetah Analog Design and entered a strategic partnership with IPro Silicon IP Ltd to expand our global reach to EMEA markets, among other developments.
From a headcount perspective, we’re also approaching 50 employees and expect to significantly grow beyond that this year. Our goal is to quadruple bookings and revenue by the end of 2025.
Can you tell us about your recent round of funding and the new board members that have joined?
On January 23, we announced our $36M+ Series B funding round. It was led by Maverick Silicon, with strategic investment from Synopsys, and our previous investors, including Matrix Partners and Intel Capital, also reinvested.
The funding will be used to support Baya Systems’ operational growth, accelerate the development and deployment of our software-driven IP technology portfolio, and further enhance our capabilities in the chiplet economy to meet the growing demands for hyper-efficient data movement in intelligent compute systems.
In our product portfolio, we will advance our flagship products like WeaverPro™ and WeaveIP™, optimizing performance and efficiency in multi-die system designs. We will enhance our ability to support leading-edge designs in AI acceleration, automotive, and data center infrastructure, focusing on scalable, modular solutions that address data movement and connectivity bottlenecks. We are seeing emerging needs in communications infrastructure and the growth in scale up and scale out as new areas for our solutions.
In conjunction with the funding, we added two industry experts to our board: Manish Muthal, Senior Managing Director at Maverick Silicon, and Siva Yerramilli, SVP of the Corporate Incubation Group at Synopsys, Inc. They join our existing board, which includes industry luminary and CEO of Tenstorrent Jim Keller, Matrix Capital General Partner Stan Reiss and our CEO, Dr. Sailesh Kumar. Combined, their guidance and insights are invaluable to support our ongoing growth and commitment to advancing high-performance system design.
How does Baya Systems differentiate itself and stay ahead of competitors in this space?
We identified certain gaps in the current market, including very high-performance, large core count coherent processing systems, high-bandwidth, petabyte-scale AI acceleration systems with advanced capabilities and complex processing requirements. All need efficient designs without compromising performance, and most importantly, systems built from the ground up for multi-chiplet scale.
Fundamentally, the system development is workload-based and software driven, allowing the rapid refinement from start to finish with full clarity on KPIs. The unique transport allows us to build highly efficient compact fabric with extremely high performance. The hardware-based coherency model and scale is designed for large clusters and chiplets, and the fabric is fully featured from the ground up for reliability (RAS) and Functional Safety (FuSA).
Established players are certainly formidable, however, the demand we’re seeing is for solutions that don’t currently exist in the market. In this class of performance, only a handful of the leading silicon vendors have been able to create solutions, and they’ve spent hundreds of man-years developing them in-house.
We’re growing our operations to meet this demand. Emerging ecosystems are developing around RISC-V, a popular open CPU architecture, both for general purpose processing and AI, which also have a significant need for our solutions and an interest in them.
As we establish our value in these leading edge markets, we will see a natural flow through to other markets downstream.
What do you envision for the future of chips and chiplets designed for next-generation computing requirements, such as AI?
Chiplets will soon become a mechanism to delivery optimized IP for compute elements or “sub-systems”. AI requires multiple compute elements, from traditional CPUs, GPUs and specialized NPUs or accelerators working in a highly optimized way, using a very capable, high-bandwidth fabric. The nature of AI, especially Generative AI, and multi-modal compute is the need for highly scalable systems, where modularity is critical.
The notion of a “tile” that can be at the granule of a self-contained compute subsystem that can be scaled by putting down many instances is how the industry can meet the consistently growing demand while addressing Time to Market (TTM) needs. Baya’s platform is designed exactly for the purpose of highly efficient tiling, with a physically-aware design platform that allows for rapid and localized customization.
Scalable systems are the way of the future. Thinking of systems from the understanding what are the workloads, to the best possible ways to solve it.
