Sr. Staff Software Engineer

Waymo details custom chip built to speed up robotaxi reaction time. Waymo says its new chips give robotaxis a major boost in processing power, part of a compute system that's grown 20 times more powerful in just eight years. On the Dash: * Waymo's new 5nm chip gives its robotaxis a big boost in raw processing power. * The company says Computing power has scaled 20 times over the past eight years. * Waymo still partners with AMD, Nvidia, TSMC and others on hardware. Alphabet's autonomous vehicle unit, Waymo, built its own computer chip to help its robotaxis react faster on the road, the company said Thursday. The chip is called an ASIC, short for application-specific integrated circuit. It handles the flood of data coming in from a robotaxi's cameras, radar and lidar sensors, and processes that data before passing it to Waymo's driving software. Waymo said the goal is faster, sharper reactions in busy city traffic. Waymo said while these ASICs alone deliver over 1,000 TOPS, or trillions of operations per second, dedicated to front-end processing and ML models, they optimize across the full stack to maximize performance. The company said it has scaled its overall compute power 20 times over the past eight years, drawing on more than 200 million fully autonomous miles to guide the design. Waymo built the system around three priorities. It needs to be responsive enough to translate sensor input into a driving decision within milliseconds. It needs to be ruggedized to withstand vibration, shock and extreme temperatures, aided by a direct tie into the vehicle's liquid cooling system. And it needs to be redundant, running as two independent engines in parallel so one can take over instantly if the other fails. The custom ASIC doesn't replace Waymo's outside suppliers. The company said it continues to work with AMD, Micron, Nvidia, Samsung, Sandisk, Socionext and TSMC to build out the rest of its computing stack, pairing its own silicon with third-party CPUs, GPUs and accelerators. Waymo said its latest system processes 13 high-resolution cameras simultaneously in real time, an increase the company credits to co-designing its chips, sensors and algorithms together rather than sourcing them separately. Waymo plans to share more detail on its compute strategy at the Hot Chips conference next week at Stanford's Memorial Auditorium in Palo Alto, California. Hot Chips is an annual conference where chip makers reveal the technical details of their newest processors, usually months before those chips actually ship.

Alphabet (GOOGL) unveils Waymo's custom AI chip and Ojai robotaxi platform. Key highlights. Table of Contents * Waymo developed a proprietary ASIC processor delivering over 1,000 TOPS for autonomous vehicle operations * The processor utilizes TSMC's 5nm manufacturing process, decreasing dependency on Nvidia and AMD hardware * Waymo's Ojai robotaxi platform, based on Zeekr minivan architecture, debuts for public riders in Los Angeles, Phoenix, and San Francisco * The Ojai integrates Waymo's sixth-generation autonomy stack and features Google's Gemini AI for passenger interactions * Chinese manufacturing origins subject Ojai vehicles to US import tariffs, creating cost challenges despite efficiency gains Alphabet's autonomous vehicle subsidiary, Waymo, has introduced a proprietary silicon processor designed specifically for its self-driving taxi operations. The Thursday announcement highlighted the chip's role in enhancing navigation capabilities through dense urban environments. This application-specific integrated circuit (ASIC) delivers computing power exceeding 1,000 trillion operations per second, matching performance benchmarks established by Nvidia's autonomous vehicle platforms. Previously, Waymo's fleet operated on third-party processors from manufacturers including Nvidia and AMD. This proprietary solution enables faster sensor data interpretation and more efficient execution of the artificial intelligence algorithms that enable real-time environmental understanding. Taiwan Semiconductor Manufacturing Company produces the chip using its 5-nanometer fabrication technology. Though not TSMC's most advanced node, this manufacturing standard remains prevalent among leading technology firms. Next-Generation robotaxi platform deployment. Concurrently, Waymo has made its Ojai robotaxi available to the general public across Los Angeles, Phoenix, and San Francisco. These vehicles now incorporate the newly developed custom processor. The Ojai utilizes a minivan chassis produced by Zeekr, a subsidiary of China's Geely Holding Group. Waymo's collaboration with Zeekr began in 2021, with subsequent years dedicated to preparing the platform for commercial deployment. Vehicles arrive in the United States stripped of Chinese connectivity systems. Waymo's Arizona operations center then integrates the company's autonomous driving technology into each unit. The Ojai platform houses Waymo's sixth-generation self-driving architecture, engineered for cross-platform compatibility. Interior redesigns accompany the technical upgrades, while Google's Gemini AI provides conversational assistance to passengers. Approximately 300 Ojai vehicles currently serve Waymo's commercial operations. Passengers requesting rides in the three initial markets may receive an Ojai for their journey. Economic considerations and market expansion. Waymo intends to introduce the Ojai to Denver, Las Vegas, and San Diego before year-end. The company maintains robotaxi services across 11 American cities. The Ojai platform promises reduced production, operational, and maintenance expenses compared to the Jaguar I-Pace models that have anchored Waymo's fleet for years. These economics support Waymo's objective of achieving profitability at scale. Nevertheless, American tariffs on Chinese-manufactured vehicles increase the per-unit import cost for each Ojai. This policy creates headwinds for Waymo's financial optimization efforts. Waymo maintains market leadership among US robotaxi providers, outpacing competitors such as Tesla and Amazon's Zoox division. The pairing of a custom-designed vehicle with proprietary processing hardware demonstrates Waymo's evolution toward greater vertical integration. Alphabet's comprehensive initiative to develop specialized chips for artificial intelligence infrastructure provides additional context. Waymo's processor development aligns with this corporate-wide strategy to minimize reliance on external component suppliers. Limited Time Offer Get 3 free stock ebooks. Discover top-performing stocks in AI, Crypto, and Technology with expert analysis. * Top 10 AI Stocks - Leading AI companies * Top 10 Crypto Stocks - Blockchain leaders * Top 10 Tech Stocks - Tech giants