TL;DR:
- AVCC and MLCommons announce the industry’s first Automotive Benchmark.
- The benchmark aims to translate specifications and benchmarks into open-source software and certification.
- Initial participants include Arm, Bosch, cTuning Foundation, KPIT, NVIDIA, Red Hat, Qualcomm, Inc., Samsung Electronics, and other industry leaders.
- ML adoption in autonomous vehicles has increased, necessitating common ML benchmarks for fair comparisons.
- The partnership aims to develop an industry-standard Automotive Benchmark Suite.
- The Suite will leverage existing benchmark reports and accelerate the development of new capabilities.
- The automotive benchmark suite will be developed in phases, with open-sourced software solutions planned by year-end.
- The collaboration seeks to foster innovation, encourage standards, and enhance decision-making for OEMs and automotive suppliers.
- Broad participation from the industry is encouraged, including AVCC and MLCommons members.
- The Automotive Benchmark initiative invites organizations using automotive AI technology to participate in developing the new Standard Automotive Benchmark Suite.
Main AI News:
AVCC and MLCommons have joined forces to introduce a groundbreaking Automotive Benchmark, paving the way for enhanced machine learning (ML) capabilities in the automotive industry. This collaboration marks the first-of-its-kind effort to translate specifications and benchmarks into open-source software and certification, setting a new standard for excellence. Notably, the benchmark’s initial contributors include prominent industry leaders such as Arm, Bosch, cTuning Foundation, KPIT, NVIDIA, Red Hat, Qualcomm, Inc., and Samsung Electronics.
In recent years, the utilization of ML, particularly in autonomous vehicles’ perception systems, has experienced a remarkable surge. This rise in ML adoption has facilitated transformative innovations like automatic lane-keeping, significantly enhancing road safety. As vehicles continue to evolve into highly intelligent machines, the industry faces the imperative need for a universally accepted set of ML benchmarks. These benchmarks will enable fair and accurate comparisons among different technologies, fostering healthy competition and driving further advancements.
The primary objective of this partnership is to develop an industry-standard Automotive Benchmark Suite specially designed for original equipment manufacturers (OEMs) and automotive suppliers utilizing AI/ML Deep Neural Network (DNN) technology. Building upon the AVCC AI/ML Benchmark Technical Reports and the MLPerf™ benchmark suites developed by MLCommons, this Suite will serve as a comprehensive resource for the industry.
It will guide collective engineering efforts across various platforms, ultimately expediting the development of cutting-edge capabilities. To achieve this, the development of the automotive benchmark suite will follow a carefully planned, multi-phase approach, with a targeted release of open-sourced software solutions by the end of the year.
David Kanter, the executive director of MLCommons, expressed his enthusiasm about this collaboration, stating, “We are thrilled to leverage our machine learning expertise in the automotive industry by collaborating with AVCC. This partnership presents a remarkable opportunity to foster innovation and establish standards for increasingly intelligent and capable vehicles.”
Addressing the existing challenges faced by OEMs and automotive suppliers, Armando Pereira, president of AVCC, explained, “Currently, these industry players struggle to assess a solution’s compute performance and system resource requirements. Through the concerted efforts of our joint task force, we aim to provide the industry with a readily available and certified source of information. This will empower decision-makers in selecting suppliers and making substantial project investments.“
The technical groundwork has already commenced, and the collaborative project welcomes broad participation from the industry. The Automotive Benchmark initiative actively seeks input from AVCC and MLCommons members, as well as other stakeholders in the automotive sector. If your organization utilizes or plans to leverage automotive AI technology, we encourage you to join this joint initiative. Together, we can drive the development of the new Standard Automotive Benchmark Suite, shaping the future of the industry.
Conlcusion:
The introduction of the Automotive Benchmark by AVCC and MLCommons signifies a significant milestone in the market. This collaboration sets the stage for standardized specifications, benchmarks, and open-source software in the automotive industry. With the participation of key industry players, the benchmark aims to facilitate fair and accurate comparisons between ML technologies. This will encourage healthy competition, spur innovation, and drive advancements in the field.
Additionally, the development of the industry-standard Automotive Benchmark Suite will provide OEMs and automotive suppliers with a certified source of information, enabling them to make informed decisions regarding supplier selection and project investments. Ultimately, this initiative will contribute to the overall growth and maturation of the market, fostering a more robust and efficient ecosystem for AI-driven automotive solutions.
