Kandou AI Secures $225M, Revolutionizing AI's Core Connectivity
March 24, 2026, 3:33 pm
Kandou AI, a Swiss-rooted startup, just secured $225 million in a significant funding round. This investment targets the critical bottlenecks plaguing modern artificial intelligence systems. AI's explosive growth demands unprecedented data movement. Traditional networking struggles to keep pace. Kandou AI offers a groundbreaking solution. Its patented copper interconnect technologies promise higher bandwidth, extended reach, and superior energy efficiency. This funding empowers Kandou AI to scale production, forge new partnerships, and accelerate its next-generation chip development. The company aims to enable multi-terabit connectivity for hyperscale and AI infrastructure customers worldwide. This move is poised to simplify complex processor designs, reduce manufacturing costs, and broaden global AI adoption.
The artificial intelligence boom reshapes technology. It also creates immense infrastructure challenges. Large AI models require vast memory and interconnect bandwidth. Current networking solutions, designed for older eras, fall short. This bottleneck limits AI's potential. Kandou AI steps in to solve this problem.
The company recently closed a $225 million funding round. Synopsys Inc. and Cadence Design Systems Inc. participated. These publicly traded giants specialize in chip development software. AIchip Technologies Ltd., SoftBank Group PBC, and Maverick Silicon also joined the investment. The funding round values Kandou AI at $400 million.
Kandou AI focuses on copper-based interconnects. These replace slower, less efficient legacy systems. AI clusters, for instance, rely on GPUs. These GPUs must constantly exchange data. Fiber optic cables typically handle this traffic. Kandou AI offers a cost-efficient, copper-based alternative.
One key product is Glasswing. This interconnect links a chip's internal components. Modern processors often feature multiple computing and memory modules. Each module sits on a separate die. Glasswing moves data between these modules faster. It also uses less memory than competing technologies.
Glasswing operates on CNRZ-5 Chord technology. This proprietary system employs a unique variation of differential signaling. Differential signaling encodes data into two electrical signals. These signals have different voltages. They travel via two separate wires. The system decodes data by measuring the voltage difference.
CNRZ-5 Chord elevates this concept. It utilizes six wires for data transmission. This allows it to send up to five bits simultaneously. It achieves this with half the power of competing technologies. CNRZ-5 also transmits data over longer distances. This provides chip designers greater flexibility. It simplifies processor architecture.
Kandou AI's technology shrinks multi-die processor main chiplets. It eliminates the need for an interposer. Interposers traditionally move data between chiplets. Removing them reduces design complexity. It also lowers manufacturing defects. This makes processors easier and cheaper to produce. These innovations drive down the cost of AI infrastructure. They accelerate AI deployment.
The company shifted its focus to the AI infrastructure market. This strategic pivot occurred after Chief Executive Srujan Linga joined last year. Before this, Kandou AI developed components for connected devices. Industrial computers and consumer electronics were key markets. Their products improved USB and PCIe link reliability. They also offered data exchange switches for peripherals.
Beyond hardware, Kandou AI provides software. An interconnect diagnostics application assists engineers. This tool tests the reliability of network links. It ensures robust connections using Kandou's hardware. This comprehensive approach supports the entire AI system lifecycle.
Kandou AI's early success received vital support. Venture Kick provided foundational funding. This moved the company from simulation to its first silicon. It also established a framework for business case refinement. The structured support forced focus on milestones. It sharpened the company's narrative. It connected them with mentors and investors. This combination transformed a research idea into a competitive semiconductor firm.
The funding will scale production of high-performance connectivity chips. It will expand partnerships with hyperscale and AI infrastructure customers. It also accelerates development of next-generation chips and IP. This aims for multi-terabit connectivity. Kandou AI recently hit significant milestones. These include the tapeout of its next-generation SerDes technology. Its retimer business shows growth. A new design center opened in Hyderabad, India. Industry veterans joined its leadership team.
The interconnect market is dynamic. Several startups have raised substantial funding this year. Xscape Photonics Inc., developing laser-based interconnects, secured $37 million. Ayar Labs Inc. raised $500 million from a consortium including major chipmakers. Kandou AI's investment places it firmly in this competitive landscape. Its copper-based approach differentiates it.
Kandou AI's solutions make AI systems more efficient. They are also more cost-effective. This supports wider adoption of AI infrastructure. The company rethinks data movement across copper interconnects. This enables a future where AI's immense processing power is unconstrained by slow connections. This investment signals a major step forward for the entire AI industry. It underscores the critical role of advanced connectivity in the age of artificial intelligence.
The artificial intelligence boom reshapes technology. It also creates immense infrastructure challenges. Large AI models require vast memory and interconnect bandwidth. Current networking solutions, designed for older eras, fall short. This bottleneck limits AI's potential. Kandou AI steps in to solve this problem.
The company recently closed a $225 million funding round. Synopsys Inc. and Cadence Design Systems Inc. participated. These publicly traded giants specialize in chip development software. AIchip Technologies Ltd., SoftBank Group PBC, and Maverick Silicon also joined the investment. The funding round values Kandou AI at $400 million.
Kandou AI focuses on copper-based interconnects. These replace slower, less efficient legacy systems. AI clusters, for instance, rely on GPUs. These GPUs must constantly exchange data. Fiber optic cables typically handle this traffic. Kandou AI offers a cost-efficient, copper-based alternative.
One key product is Glasswing. This interconnect links a chip's internal components. Modern processors often feature multiple computing and memory modules. Each module sits on a separate die. Glasswing moves data between these modules faster. It also uses less memory than competing technologies.
Glasswing operates on CNRZ-5 Chord technology. This proprietary system employs a unique variation of differential signaling. Differential signaling encodes data into two electrical signals. These signals have different voltages. They travel via two separate wires. The system decodes data by measuring the voltage difference.
CNRZ-5 Chord elevates this concept. It utilizes six wires for data transmission. This allows it to send up to five bits simultaneously. It achieves this with half the power of competing technologies. CNRZ-5 also transmits data over longer distances. This provides chip designers greater flexibility. It simplifies processor architecture.
Kandou AI's technology shrinks multi-die processor main chiplets. It eliminates the need for an interposer. Interposers traditionally move data between chiplets. Removing them reduces design complexity. It also lowers manufacturing defects. This makes processors easier and cheaper to produce. These innovations drive down the cost of AI infrastructure. They accelerate AI deployment.
The company shifted its focus to the AI infrastructure market. This strategic pivot occurred after Chief Executive Srujan Linga joined last year. Before this, Kandou AI developed components for connected devices. Industrial computers and consumer electronics were key markets. Their products improved USB and PCIe link reliability. They also offered data exchange switches for peripherals.
Beyond hardware, Kandou AI provides software. An interconnect diagnostics application assists engineers. This tool tests the reliability of network links. It ensures robust connections using Kandou's hardware. This comprehensive approach supports the entire AI system lifecycle.
Kandou AI's early success received vital support. Venture Kick provided foundational funding. This moved the company from simulation to its first silicon. It also established a framework for business case refinement. The structured support forced focus on milestones. It sharpened the company's narrative. It connected them with mentors and investors. This combination transformed a research idea into a competitive semiconductor firm.
The funding will scale production of high-performance connectivity chips. It will expand partnerships with hyperscale and AI infrastructure customers. It also accelerates development of next-generation chips and IP. This aims for multi-terabit connectivity. Kandou AI recently hit significant milestones. These include the tapeout of its next-generation SerDes technology. Its retimer business shows growth. A new design center opened in Hyderabad, India. Industry veterans joined its leadership team.
The interconnect market is dynamic. Several startups have raised substantial funding this year. Xscape Photonics Inc., developing laser-based interconnects, secured $37 million. Ayar Labs Inc. raised $500 million from a consortium including major chipmakers. Kandou AI's investment places it firmly in this competitive landscape. Its copper-based approach differentiates it.
Kandou AI's solutions make AI systems more efficient. They are also more cost-effective. This supports wider adoption of AI infrastructure. The company rethinks data movement across copper interconnects. This enables a future where AI's immense processing power is unconstrained by slow connections. This investment signals a major step forward for the entire AI industry. It underscores the critical role of advanced connectivity in the age of artificial intelligence.