Compute Solution for Tesla's Full Self-Driving Computer
The DOI matches its Crossref record; 4 of the 4 references it cites were found in a registry or an index.
Abstract
Tesla's full self-driving (FSD) computer is the world's first purpose-built computer for the highly demanding workloads of autonomous driving. It is based on a new System on a Chip (SoC) that integrates industry standard components such as CPUs, ISP, and GPU, together with our custom neural network accelerators. The FSD computer is capable of processing up to 2300 frames per second, a 21× improvement over Tesla's previous hardware and at a lower cost, and when fully utilized, enables a new level of safety and autonomy on the road.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“In the past few years, some companies and universities have presented application specific integrated circuits (ASICs) based on the complementary metal oxide semiconductor (CMOS) technology that are capable to compute and store information in the same unit.”
From Hardware implementation of memristor-based artificial neural networks (2024) · cites this paper as [21]
Citations
Cites
In-Datacenter Performance Analysis of a Tensor Processing Unit
Jouppi, N. P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., Bates, S., Bhatia, S., Boden, N., Borchers, A., et al. (2017) · Proceedings of the 44th Annual International Symposium on Computer Architecture · doi:10.1145/3079856.3080246 · Cited by 4,537 · cited 1×
“Thus, each MAC unit computes the dotproduct locally with no data movement within the MAC array, unlike the systolic array computations implemented in many other processors.(3,4) This results in lower power and less cell area than systolic array implementations.”
Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review
Rawat, W., & Wang, Z. (2017) · Neural Computation · doi:10.1162/neco_a_00990 · Cited by 3,503
Object Recognition with Gradient-Based Learning
LeCun, Y., Haffner, P., Bottou, L., & Bengio, Y. (1999) · Lecture Notes in Computer Science · doi:10.1007/3-540-46805-6_19 · Cited by 947
Floating point HOG implementation for real-time multiple object detection
Komorkiewicz, M., Kluczewski, M., & Gorgoń, M. (2012) · International Conference on Field Programmable Logic and Applications (FPL) · doi:10.1109/fpl.2012.6339159 · Cited by 52
Research works printed but not linked: 1
- horowitz, 1999, Computing's energy problem, IEEE Int Solid-State Circuits Conf Dig Tech Papers, 10
Cited by
Hardware implementation of memristor-based artificial neural networks
Aguirre, F., Sebastian, A., Le Gallo, M., Song, W., Wang, T., Yang, J. J., Lu, W., Chang, M.-F., Ielmini, D., Yang, Y., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-45670-9 · Cited by 564
Huang, M. Q., Ninić, J., & Zhang, Q. B. (2021) · Tunnelling and Underground Space Technology · doi:10.1016/j.tust.2020.103677 · Cited by 370
Survey of Machine Learning Accelerators
Reuther, A., Michaleas, P., Jones, M., Gadepally, V., Samsi, S., & Kepner, J. (2020) · 2020 IEEE High Performance Extreme Computing Conference (HPEC) · doi:10.1109/hpec43674.2020.9286149 · Cited by 163
Perez-Cerrolaza, J., Abella, J., Borg, M., Donzella, C., Cerquides, J., Cazorla, F. J., Englund, C., Tauber, M., Nikolakopoulos, G., & Flores, J. L. (2024) · ACM Computing Surveys · doi:10.1145/3626314 · Cited by 185
“…Processing Units (TPUs), Network Processing Units (NPUs), neuromorphic computing) [Chen et al. 2019b; Dally et al. 2020; Jouppi et al. 2017; Schuman et al. 2017] and custom-designed devices (e.g., Tesla FSD [Talpes et al. 2020]) often including specialized accelerators (e.g., DNN accelerator).”
BackgroundMuralidhar, R., Borovica‐Gajic, R., & Buyya, R. (2022) · ACM Computing Surveys · doi:10.1145/3511094 · Cited by 150
AI and ML Accelerator Survey and Trends
Reuther, A., Michaleas, P., Jones, M., Gadepally, V., Samsi, S., & Kepner, J. (2022) · 2022 IEEE High Performance Extreme Computing Conference (HPEC) · doi:10.1109/hpec55821.2022.9926331 · Cited by 109
AI Accelerator Survey and Trends
Reuther, A., Michaleas, P. W., Jones, M. P., Gadepally, V. N., Samsi, S., & Kepner, J. (2021) · IEEE High Performance Extreme Computing Conference · doi:10.1109/hpec49654.2021.9622867 · Cited by 124
“Finally,a reasonable categorization of accelerators follows their intended application, and the five categories are shown as ellipses on the graph, which roughly correspond to performance and power consumption: Very Low Power for speech processing, very small sensors, etc.; Embedded for cameras, small UAVs and robots,…”
MethodTestability and Dependability of AI Hardware: Survey, Trends, Challenges, and Perspectives
Su, F., Liu, C., & Stratigopoulos, H. (2023) · IEEE Design and Test · doi:10.1109/mdat.2023.3241116 · Cited by 97
“Redundancy-based State-of-the-art AI hardware accelerators for autonomous driving vehicles employ dual-modular redundancy (DMR) to ensure safety for the system [250], [251], which requires substantial hardware resources.”
BackgroundAghapour, Z., Sharifian, S., & Taheri, H. (2023) · Computer Networks · doi:10.1016/j.comnet.2023.109577 · Cited by 92
Steering towards safe self-driving laboratories
Leong, S. X., Griesbach, C. E., Zhang, R., Darvish, K., Zhao, Y., Mandal, A., Zou, Y., Hao, H., Bernales, V., & Aspuru‐Guzik, A. (2025) · Nature Reviews Chemistry · doi:10.1038/s41570-025-00747-x · Cited by 95
Alser, M., Lindegger, J., Fırtına, C., Alserr, N. A., Mao, H., Singh, G., Gómez-Luna, J., & Mutlu, O. (2022) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2022.08.019 · Cited by 88
“We are witnessing a paradigm shift to near-data computing with more specialized hardware accelerators for other key applications such as self-driving cars [236], and artificial intelligence with the largest chip ever [237].”
BackgroundAdaptive approximate computing in edge AI and IoT applications: A review
Damsgaard, H. J., Grenier, A., Katare, D., Taufique, Z., Shakibhamedan, S., Troccoli, T., Chatzitsompanis, G., Kanduri, A., Ometov, A., Ding, A. Y., et al. (2024) · Journal of Systems Architecture · doi:10.1016/j.sysarc.2024.103114 · Cited by 66
“This is likely due to them being well-established in both research and production environments (even Google’s TPU and Tesla’s FSD chip execute quantized NNs [113,134,203]) as they are known for bringing very high savings with relatively little impact on model accuracy, as we will see later.”
BackgroundSLO-Aware Inference Scheduler for Heterogeneous Processors in Edge Platforms
Seo, W., Cha, S.-H., Kim, Y., Huh, J., & Park, J. (2021) · ACM Transactions on Architecture and Code Optimization · doi:10.1145/3460352 · Cited by 75
“As a real-world example, the Tesla FSD computer consists of three quad-core CPUs, one GPU, two NPUs, one ISP, and a few more ASIC chips [53].”
MethodGu, M., Zhang, Y., Wen, Y., Ai, G., Zhang, H., Wang, P., & Wang, G. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.106623 · Cited by 55
Shankar, S., & Reuther, A. (2022) · IEEE High Performance Extreme Computing Conference · doi:10.1109/hpec55821.2022.9926296 · Cited by 65
“INTRODUCTION As Artificial Intelligence and Machine Learning (AI/ML) methods evolve with advanced techniques like Deep Neural Networks and Graph Networks, they are increasingly being used across a myriad of applications including materials discovery [1] medicine [2,3], natural language processing [4], driverless cars…”
A Systematic Assessment of Embedded Neural Networks for Object Detection
Verucchi, M., Brilli, G., Sapienza, D., Verasani, M., Arena, M., Gatti, F., Capotondi, A., Cavicchioli, R., Bertogna, M., & Solieri, M. (2020) · IEEE International Conference on Emerging Technologies and Factory Automation (ETFA) · doi:10.1109/etfa46521.2020.9212130 · Cited by 67
“An autonomous car driving application typically detects road users, such as other cars, bicycles and pedestrians [1], [2].”
BackgroundMei, L., Goetschalckx, K., Symons, A., & Verhelst, M. (2023) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca56546.2023.10071098 · Cited by 63
“For HW, five DNN accelerators are selected as the architecture baselines for the case studies: Meta-prototype [28], TPU [14], Edge TPU [38], Ascend [19], and Tesla NPU [31].”
MethodOptimizing Selective Protection for CNN Resilience
Mahmoud, A., Hari, S. K. S., Fletcher, C. W., Adve, S. V., Sakr, C., Shanbhag, N. R., Molchanov, P. A., Sullivan, M. B., Tsai, T. K., & Keckler, S. W. (2021) · Proceedings/Proceedings - International Symposium on Software Reliability Engineering · doi:10.1109/issre52982.2021.00025 · Cited by 69
GPU Devices for Safety-Critical Systems: A Survey
Perez-Cerrolaza, J., Abella, J., Kosmidis, L., Calderon, A. J., Cazorla, F., & Flores, J. L. (2023) · ACM Computing Surveys · doi:10.1145/3549526 · Cited by 43
“GPU devices, and more specifically hard embedded safety GPU devices, have rapidly evolved into new application domains such as safety-critical systems in recent years.”
BackgroundInfluentialTNPU: Supporting Trusted Execution with Tree-less Integrity Protection for Neural Processing Unit
Lee, S., Kim, J.-W., Na, S., Park, J., & Huh, J. (2022) · IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca53966.2022.00025 · Cited by 61
“Recent Apple M1 and Tesla FSD also include one CPU, one GPU, and one or two NPUs [3], [4].”
BackgroundField-Programmable Gate Array Architecture for Deep Learning: Survey and Future Directions
Boutros, A., Arora, A., & Betz, V. (2025) · Proceedings of the IEEE · doi:10.1109/jproc.2025.3623023 · Cited by 55
“For example, Tesla’s full self-driving DL inference chip was custom-designed to meet an aggressive power budget of less than 40 W [28].”
Hosseini, M. J. M., Donati, E., Yokota, T., Lee, S., Indiveri, G., Someya, T., & Nawrocki, R. A. (2020) · Journal of Physics D Applied Physics · doi:10.1088/1361-6463/abc585 · Cited by 46
“machine vision and machine learning, pattern recognition, and low power consumption, for instance as required in selfdriving cars [16].”
BackgroundPreventing DNN Model IP Theft via Hardware Obfuscation
Goldstein, B. F., Patil, V. C., Ferreira, V. C., Nery, A. S., França, F. M. G., & Kundu, S. K. (2021) · IEEE Journal on Emerging and Selected Topics in Circuits and Systems · doi:10.1109/jetcas.2021.3076151 · Cited by 40
“Typically, total on-chip cache sizes are in the 10’s of megabytes with even dedicated neural network processing application platforms such as Tesla’s self-driving chip having just 32MB SRAM memory [23].”
BackgroundCho, S., Choi, H., Park, E., Shin, H., & Yoo, S. (2020) · IEEE Access · doi:10.1109/access.2020.3011265 · Cited by 68
“The TPU v1 [37], NVIDIA Turing Titan RTX/Tesla T4 [34], Habana Goya [35], NVIDIA Jetson AGX Xavier [17], Tesla’s NPU in full self-driving chip [62], and OLAccel [47] are some examples of such designs.”
Energy-Efficient Acceleration of Deep Neural Networks on Realtime-Constrained Embedded Edge Devices
Kim, B., Lee, S., Trivedi, A. R., & Song, W. J. (2020) · IEEE Access · doi:10.1109/access.2020.3038908 · Cited by 31
“Examples of CPU and neural accelerator integration in embedded edge devices [1], [4], [18], [27]. to incorporate dedicated hardware accelerators for neural processing.”
BackgroundArchitecture of neural processing unit for deep neural networks
Lee, K. J. (2020) · Advances in computers · doi:10.1016/bs.adcom.2020.11.001 · Cited by 43
Zhang, X., Hao, C., Zhou, P., Jones, A. K., & Hu, J. (2022) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1145/3489517.3530509 · Cited by 28
“Second, recent advanced systems are introducing great heterogeneity by integrating diverse acceleration components with different capabilities to pursue low latency and high energy efficiency.”
Hardware Accelerators for Autonomous Vehicles
Jhung, J., Suk, H., Park, H., & Kim, S. (2023) · Springer eBooks · doi:10.1007/978-3-031-22170-5_9 · Cited by 30
Hsiao, Y.-S., Wan, Z., Jia, T., Ghosal, R., Mahmoud, A., Raychowdhury, A., Brooks, D., Wei, G.-Y., & Reddi, V. J. (2023) · Design, Automation, and Test in Europe · doi:10.23919/date56975.2023.10137246 · Cited by 34
“Prior works adopt redundancy [6] at the hardware or software level to improve autonomous vehicles (AVs) resilience.”
MethodIECA: An In-Execution Configuration CNN Accelerator With 30.55 GOPS/mm² Area Efficiency
Huang, B., Huan, Y., Chu, H., Xu, J., Liu, L., Zheng, L., & Zou, Z. (2021) · IEEE Transactions on Circuits and Systems I Regular Papers · doi:10.1109/tcsi.2021.3108762 · Cited by 27
“CONVOLUTIONAL Neural Networks (CNN) are widely deployed for many artificial intelligence (AI) applications such as object detection [1], face recognition [2], activity recognition [3] and autonomous driving [4].”
BackgroundProbabilistically Guaranteeing End-to-end Latencies in Autonomous Vehicle Computing Systems
Lee, H., Choi, Y., Han, T., & Kim, K. (2022) · IEEE Transactions on Computers · doi:10.1109/tc.2022.3152105 · Cited by 23
“It is also assumed that the AV stack runs on a single-board computer with multi-core processors [11], [12].”
BackgroundDr. DNA: Combating Silent Data Corruptions in Deep Learning using Distribution of Neuron Activations
Ma, D., Lin, F., Desmaison, A., Coburn, J., Moore, D., Sankar, S., & Jiao, X. (2024) · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) · doi:10.1145/3620666.3651349 · Cited by 31
How to Keep Pushing ML Accelerator Performance? Know Your Rooflines!
Verhelst, M., Benini, L., & Verma, N. (2025) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2025.3553765 · Cited by 32
“…from Eyeriss’ 168 and Envision’s 256 MAC units in 2016 and 2017 [4], [12], over Samsung’s and Tesla’s ∼ 10 000 of MACs in 2019 and 2020 [13], [14], to today’s massively parallel near-and in-memory compute engines with 100 000s of parallel MAC operators [15], [16], [17], [18], [19], and often…”
Latency Measurement for Autonomous Driving Software Using Data Flow Extraction
Betz, T., Schmeller, M., Korb, A., & Betz, J. (2023) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv55152.2023.10186686 · Cited by 23
“Many companies are working on realizing autonomous driving technology and have already demonstrated initial results [1].”
Odema, M., Bouzidi, H., Ouarnoughi, H., Niar, S., & Faruque, M. A. A. (2023) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3609386 · Cited by 21
“For instance, commercial SoC platforms, such as the Nvidia Xavier [1] and Tesla FSD [30], have successfully integrated a variety of proven hardware computing units (CUs) and industrial IPs on a single chip to achieve said purpose.”
MethodDACAPO: Accelerating Continuous Learning in Autonomous Systems for Video Analytics
Kim, Y., Oh, C., Hwang, J., Kim, W., Oh, S., Lee, Y., Sharma, H., Yazdanbakhsh, A., & Park, J. (2024) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1109/isca59077.2024.00093 · Cited by 33
“Despite these attempts, the involvement of cloud computing constrains the full potential benefits of edge computing, explaining the increasing development of edge-targeted accelerators in the industry [1], [23], [27], [56], [68], [71].”
BackgroundExploiting Temporal Data Diversity for Detecting Safety-critical Faults in AV Compute Systems
Jha, S., Cui, S., Tsai, T. K., Hari, S. K. S., Sullivan, M. B., Kalbarczyk, Z., Keckler, S. W., & Iyer, R. K. (2022) · International Conference on Dependable Systems & Networks/Proceedings · doi:10.1109/dsn53405.2022.00021 · Cited by 28
“The algorithms used by the agent are computationally expensive, thereby requiring the use of a heterogeneous computational fabric consisting of CPUs, GPUs, and ASICs [16], [37].”
BackgroundMethodInfluentialHussien, M., Nguyen, K. K., Ranjha, A., Krichen, M., Alshammari, A., & Cheriet, M. (2023) · IEEE Transactions on Consumer Electronics · doi:10.1109/tce.2023.3294832 · Cited by 21
“For example, the computation power of Tesla vehicles can run the auto-pilot that can process 2300 frames per second [39].”
BackgroundA Lightweight Error-Resiliency Mechanism for Deep Neural Networks
Goldstein, B. F., Ferreira, V. C., Srinivasan, S., Das, D., Nery, A. S., Kundu, S. K., & França, F. M. G. (2021) · International Symposium on Quality Electronic Design (ISQED) · doi:10.1109/isqed51717.2021.9424287 · Cited by 20
“Several companies have reported developing their custom set of hardware and sensors (e.g., cameras, LiDAR, radar), from the ground up for autonomous applications, including Google’s Waymo [1], and Tesla [2].”
BackgroundUnderstanding question-answering systems: Evolution, applications, trends, and challenges
Farea, A., & Emmert‐Streib, F. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.110997 · Cited by 21
Efficient Soft-Error Detection for Low-precision Deep Learning Recommendation Models
Li, S., Huang, J., Tang, P. T. P., Khudia, D. S., Park, J., Dixit, H. D., & Chen, Z. (2022) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata55660.2022.10020972 · Cited by 19
“Hardware level redundancy is usually used in safetycritical task such as self-driving [16].”
BackgroundReal-time control of laser materials processing using deep learning
Grant‐Jacob, J. A., Mills, B., & Zervas, M. N. (2023) · Manufacturing Letters · doi:10.1016/j.mfglet.2023.08.145 · Cited by 19
“Neural networks have been used widely, for example in speech recognition, image recognition, image generation and autonomous driving cars [21].”
Heterogeneity-Aware Scheduling on SoCs for Autonomous Vehicles
Amarnath, A., Pal, S., Kassa, H. T., Vega, A. J., Buyuktosunoglu, A., Franke, H., Wellman, J.-D., Dreslinski, R., & Bose, P. K. (2021) · IEEE Computer Architecture Letters · doi:10.1109/lca.2021.3085505 · Cited by 18
“AV applications are composed of compute intensive tasks like perception, planning, control and communication [2], [3].”
BackgroundsNPU: Trusted Execution Environments on Integrated NPUs
Feng, E., Feng, D., Du, D., Xia, Y., & Chen, H. (2024) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1109/isca59077.2024.00057 · Cited by 26
“Due to the high demand for AI applications on mobile devices, current mobile chips [6], [15], [24], [28], [60], [86], [101], 1 The two authors contributed equally to this work and should be considered co-first authors.”
BackgroundInfluentialFunctional Criticality Analysis of Structural Faults in AI Accelerators
Chaudhuri, A., Talukdar, J., Su, F., & Chakrabarty, K. (2022) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2022.3166108 · Cited by 22
“They are being deployed for inferencing in a number of applications, e.g., autonomous driving, manufacturing automation, and navigation [9].”
BackgroundLincoln AI Computing Survey (LAICS) Update
Reuther, A., Michaleas, P. W., Jones, M. P., Gadepally, V. N., Samsi, S., & Kepner, J. (2023) · IEEE High Performance Extreme Computing Conference · doi:10.1109/hpec58863.2023.10363568 · Cited by 22
Shaping the future of nuclear reactors with digital twins: Current developments and perspectives
Huang, Q., Zeng, W., Liu, J., Zhang, Z., Deng, J., Qiu, Z., Xu, L., Wei, Z., Lu, Q., Lanxin, G. O. N. G., et al. (2025) · Applied Energy · doi:10.1016/j.apenergy.2025.126922 · Cited by 21
Special Session: Towards an Agile Design Methodology for Efficient, Reliable, and Secure ML Systems
Dave, S., Marchisio, A., Hanif, M. A., Guesmi, A., Shrivastava, A., Alouani, I., & Shafique, M. (2022) · IEEE VLSI Test Symposium (VTS) · doi:10.1109/vts52500.2021.9794253 · Cited by 20
“NPUs: Examples of NPUs include Google’s Tensor Processing Unit (TPU) [20], tensor cores in NVIDIA A100 Ampere architecture, Samsung NPU [21], Sambanova’s RDU [22], IBM’s AI Accelerator [23], Microsoft Brainwave [24], Tesla’s Self-Driving computer [25], Facebook’s ML accelerator [26], etc.”
MethodDiVa: An Accelerator for Differentially Private Machine Learning
Park, B., Hwang, R., Yoon, D., Choi, Y., & Rhu, M. (2022) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1109/micro56248.2022.00084 · Cited by 21
“While the details of the underlying microarchitecture and its dataflow are not publicly available, Tesla’s Full SelfDriving (FSD) computer [7], [91] hints at the adoption of an outer-product dataflow in conducting GEMMs.”
BackgroundML Processors Are Going Multi-Core: A performance dream or a scheduling nightmare?
Verhelst, M., Shi, M., & Mei, L. (2022) · IEEE Solid-State Circuits Magazine · doi:10.1109/mssc.2022.3201783 · Cited by 15
“Over the past years, several scheduling space exploration frameworks have been proposed for this, such as Interstellar [37], Timeloop [26], Union [15], MAGNet [35], Mind Mapping [10], or our own ZigZag framework [22], [23], [32].”
BackgroundOptimizing Monocular Driving Assistance for Real-Time Processing on Jetson AGX Xavier
Nguyen, H.-H., Tran, D. N., Pham, L. H., & Jeon, J. W. (2024) · IEEE Access · doi:10.1109/access.2024.3402239 · Cited by 15
“In addition to software development, progressively more powerful embedded platforms targeting autonomous driving have been launched or planned, including NVIDIA Jetson and Drive, Intel Mobieye EyeQ, Tesla Hardware [2], [3].”
MethodMaity, B., Yi, S., Seo, D., Cheng, L., Lim, S.-S., Kim, J.-C., Donyanavard, B., & Dutt, N. D. (2021) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3477005 · Cited by 19
“Based on our observations, it is likely that effective deployment of self-driving workloads will ultimately necessitate custom hardware [48], but designers can identify bottlenecks and candidates for optimization if realistic algorithms are evaluated on real embedded platforms.”
BackgroundHsiao, Y.-S., Wan, Z., Jia, T., Ghosal, R., Mahmoud, A., Raychowdhury, A., Brooks, D., Wei, G.-Y., & Reddi, V. J. (2023) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2023.3332293 · Cited by 19
“There are multiple error mitigation techniques, including dynamic verification [12] and redundancy [13] at the hardware or software level to improve AVs’ resilience.”
MethodA Design Framework for Generating Energy-Efficient Accelerator on FPGA Toward Low-Level Vision
Zhou, Z., Duan, X., & Han, J. S. (2024) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2024.3409649 · Cited by 12
“Digital Object Identifier 10.1109/TVLSI.2024.3409649 neural processing unit (NPU) [6], [7], [8], [9], [10] has been considered to keep pace with rapid updates.”
BackgroundInfluentialVisualizing laser ablation using plasma imaging and deep learning
Grant‐Jacob, J. A., Mills, B., & Zervas, M. N. (2023) · Optics Continuum · doi:10.1364/optcon.495923 · Cited by 13
“It has found diverse applications across many fields, including image [21,22] and speech recognition [22], natural language processing [23], and autonomous driving [24].”
Emergent Bioanalogous Properties of Blockchain-based Distributed Systems
Abramov, O., Bebell, K. L., & Mojzsis, S. J. (2021) · Origins of Life and Evolution of Biospheres · doi:10.1007/s11084-021-09608-1 · Cited by 16
“…Hirayama et al. 2019), smart manufacturing (e.g., Kusiak 2018; Wang et al. 2018), autonomous driving (e.g., Li et al., 2018; Elluswamy et al. 2020; Talpes et al. 2020), and in conjunction with global, satellite-based internet (e.g., Su et al. 2019) make it possible in the near future to close the…”
BackgroundFKeras: A Sensitivity Analysis Tool for Edge Neural Networks
Weng, O., Meza, A., Bock, Q., Hawks, B., Campos, J., Tran, N. V., Duarte, J. M., & Kastner, R. (2024) · ACM Journal on Autonomous Transportation Systems · doi:10.1145/3665334 · Cited by 13
“It provides a sensitivity score that ranks the NN parameters based on their sensitivity to faults and supports modeling single-and multiple-bit fault injection campaigns on NN weights.”
Hardware Platforms Enabling Edge AI for Space Applications: A Critical Review
Mystkowska, G., Monopoli, M., Nannipieri, P., Zulberti, L., Codinachs, D. M., & Fanucci, L. (2025) · IEEE Access · doi:10.1109/access.2025.3596326 · Cited by 15
“Tesla used it in the Full self-driving (FSD) computer Neural Network Accelerator (NNA), which is dedicated to accelerating NN for autonomous driving [24].”
Techniques for detecting and masking faults in semantic segmentation applications
Burel, S., Evans, A., & Anghel, L. (2024) · Microelectronics Reliability · doi:10.1016/j.microrel.2024.115397 · Cited by 11
Fixed-point Encoding and Architecture Exploration for Residue Number Systems
Deng, B., Nadendla, B. S., Suo, K., Xie, Y., & Lo, D. C.-T. (2024) · ACM Transactions on Architecture and Code Optimization · doi:10.1145/3664923 · Cited by 13
“Appropriate RNS real-number encoding could be one of the strategies to further raise the efficiency of AI accelerators (e.g., Google TPU [40, 41], Sam-sung NPU [38, 54], Tesla FSD Chip [11, 66], Nvidia NVDLA [25, 28], Arm MLP [16]).”
Panopticus: Omnidirectional 3D Object Detection on Resource-constrained Edge Devices
Lee, J., Jung, C. R., Kim, J., & Cha, H. (2024) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3636534.3690688 · Cited by 13
Hardware accelerator systems for embedded systems
Song, W. J. (2020) · Advances in computers · doi:10.1016/bs.adcom.2020.11.004 · Cited by 15
Vision Control Unit in Fully Self Driving Vehicles using Xilinx MPSoC and Opensource Stack
Chakaravarthy, R. V., Kwon, H. J., & Jiang, H. L. (2021) · Asia and South Pacific Design Automation Conference (ASP-DAC) · doi:10.1145/3394885.3431616 · Cited by 11
“While companies like Tesla, Google [1] [2] have developed proprietary stack to build FSD vehicles, organizations like Autoware Foundation and Baidu [3] [4] are developing open source stacks to enable self-driving mobility.”
BackgroundJigsaw: Taming BEV-centric Perception on Dual-SoC for Autonomous Driving
Sun, L., Li, C., Hou, X., Huang, T., Xu, C., Wang, X., Bao, G., Sun, B., Rui, S., & Guo, M. (2024) · IEEE Real-Time Systems Symposium (RTSS) · doi:10.1109/rtss62706.2024.00032 · Cited by 9
“Some SoCs are equipped with both GPU and AI accelerators [7], [72].”
InfluentialFoundation Models for Autonomous Driving Systems: An Initial Roadmap
Wu, X., Cheng, M., Ren, X., Hu, Q., Chen, J., Huang, Y., Cordy, M., Zhang, Y., Xie, X., Ma, L., et al. (2026) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3807455 · Cited by 18
“…a m e - w o r k P y T o r c h [ 153 ] , T e n s o r F l o w [ 5 ] , J A X [ 20 ] , P a dd l e P a dd l e [ 1 8 ] T r a i n i n g t i m e , a cc u r a c y , i n f e r e n c e t i m e , C P U / G P U u t i l i z a t i o n , m e m o r y u s a g e , c o m - m u n i t y s u pp o r t O n l y o ff e r b…”
BiSTAHL: A Built-In Self-Testable Soft-Error-Hardened Scan-Cell
Holst, S., Ma, R., Wen, X., Yan, A., & Xu, H. (2023) · IEEE European Test Symposium (ETS) · doi:10.1109/ets56758.2023.10174154 · Cited by 8
“Self-driving cars and other advanced machine-learning based automation requires advanced technology nodes being used in safety-critical applications [1].”
BackgroundHardware-aware NAS by Genetic Optimisation with a Design Space Exploration Simulator
Hendrickx, L., Symons, A., Ranst, W. V., Verhelst, M., & Goedemé, T. (2023) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw59228.2023.00222 · Cited by 8
“However, compared to the NPU results described in Table 1, the reduction factor is smaller, around 2 times.”
MethodZhao, W., Yang, G., Xia, T., Chen, F., Zheng, N., & Ren, P. (2023) · IEEE Transactions on Very Large Scale Integration (VLSI) Systems · doi:10.1109/tvlsi.2023.3327110 · Cited by 8
“To address this problem, researchers have proposed several architectural innovations [15], [16], [17], [18], [19], [20], [21], [22], [23].”
BackgroundHighly Efficient Self-checking Matrix Multiplication on Tiled AMX Accelerators
Mummidi, C. S., Ferreira, V. C., Srinivasan, S., & Kundu, S. K. (2023) · ACM Transactions on Architecture and Code Optimization · doi:10.1145/3633332 · Cited by 12
“The Tesla Full Self-driving (FSD) chip also employs double-modular redundancy at the chip level [50].”
BackgroundAuto-Healer: Self-Healing Hardware for Perception Stage Faults in Autonomous Driving Systems
Suvizi, A., & Venkataramani, G. (2025) · International Conference on Supercomputing · doi:10.1145/3721145.3725755 · Cited by 11
A cognitive digital twin approach to improving driver compliance and accident prevention
Gu, Y., Li, S., Qi, J., Fu, B., Tang, R., Yang, L., Tian, S., & Jiang, Z. (2025) · Accident Analysis & Prevention · doi:10.1016/j.aap.2024.107913 · Cited by 7
Reliability of Automotive Multidomain Controllers: Advancements in electronics cooling technologies
Korta, J. A., Skruch, P., & Holon, K. (2020) · IEEE Vehicular Technology Magazine · doi:10.1109/mvt.2020.3038588 · Cited by 7
“Some examples of electronic devices designed for automotive MDCs are the NVIDIA Xavier and the Tesla Full Self-Driving systems on chip (SoCs), which alone can dissipate 30 W [3] and up to 40 W [4] (often quoted as 36 W) on dies of 3.”
MethodResearch Challenges for Energy-Efficient Computing in Automated Vehicles
Dick, R. P., Aitken, R., Mogill, J., Strachan, J. P., Bresniker, K., Lu, W. D., Nakahira, Y., Li, Z., Marinella, M. J., Severa, W., et al. (2023) · Computer · doi:10.1109/mc.2022.3180987 · Cited by 7
“Such research will need to build on existing designs 15 and extend them to improve capability, power, and performance.”
BackgroundA Review of Key Technologies for Environment Sensing in Driverless Vehicles
Huo, Y., & Zhang, C. (2024) · World Electric Vehicle Journal · doi:10.3390/wevj15070290 · Cited by 8
Custódio, T., Alves, C. P., Silva, P., da Silva, J. V. L., Rodrigues, C., Lourenço, R., Pessoa, R., Moreira, F., Marques, R., Tomé, G., et al. (2021) · Electronics · doi:10.3390/electronics11010021 · Cited by 7
Accelerating Object Detection Models Inference within Deep Learning Workbench
Demidovskij, A. V., Tugaryov, A., Fatekhov, M., Aidova, E., Stepyreva, E., Shevtsov, M., & Gorbachev, Y. (2021) · International Conference on Engineering and Emerging Technologies (ICEET) · doi:10.1109/iceet53442.2021.9659634 · Cited by 7
Lincoln AI Computing Survey (LAICS) and Trends
Reuther, A., Michaleas, P. W., Jones, M. P., Gadepally, V. N., & Kepner, J. (2025) · IEEE High Performance Extreme Computing Conference · doi:10.1109/hpec67600.2025.11196427 · Cited by 7
Cano, M., Guillén, A., Tasic, I. A. B., & Villafranca, A. (2024) · International Conference on Control, Robotics and Cybernetics (CRC) · doi:10.1109/crc60659.2023.10488509 · Cited by 7
“…learning have led to new achievements, such as solving a largely open problem on protein folding, obtaining highly intelligent language models such as GPT-3 [58], making autonomous machines that walk, think, and learn [59], all at an unprecedented rate, or even fully autonomous driving [60].”
BackgroundYi, X., Deng, Y., Antonio, R. A. G., Kong, F., Paim, G., & Verhelst, M. (2025) · ACM/IEEE Design Automation Conference (DAC) · doi:10.1109/dac63849.2025.11133141 · Cited by 7
“To meet the stringent performance and energy consumption requirements of DNN deployment, numerous contemporary domain-specific dataflow accelerators have been developed [2], [7]–[11], incorporating extensive optimizations in the processing element (PE) array design and dataflow strategies.”
GPU-Accelerated Timing Simulation of Systolic-Array-Based AI Accelerators
Holst, S., Bumun, L., & Wen, X. (2021) · Proceedings - Asian Test Symposium/Proceedings · doi:10.1109/ats52891.2021.00034 · Cited by 6
Robot-on-Chip: Computing on a Single Chip for an Autonomous Robot
Jeong, Y. W., Go, K. H., & Lee, S. E. (2022) · International Conference on Consumer Electronics · doi:10.1109/icce53296.2022.9730399 · Cited by 6
Verification and Validation of Autonomous Systems
Shetiya, S. S., Vyas, V., & Renukuntla, S. (2024) · International Journal of Current Research in Science Engineering & Technology · doi:10.30967/ijcrset/sneha-sudhir-shetiya/153 · Cited by 6
Compilation and Optimizations for Efficient Machine Learning on Embedded Systems
Zhang, X., Chen, Y., Hao, C., Huang, S., Li, Y., & Chen, D. (2023) · Springer eBooks · doi:10.1007/978-3-031-39932-9_3 · Cited by 7
9.1 D1: A 7nm ML Training Processor with Wave Clock Distribution
Fischer, T. C., Nivarti, A. K., Ramachandran, R., Bharti, R., Carson, D., Lawrendra, A., Mudgal, V., Santhosh, V., Shukla, S., & Tsai, T.-C. (2023) · Digest of technical papers/Digest of technical papers - IEEE International Solid-State Circuits Conference · doi:10.1109/isscc42615.2023.10067658 · Cited by 6
“DOJO applications include training ML networks that run on the Tesla Full Self-Driving (FSD) computer in Tesla vehicles [3,4].”
MethodUltra-Dense 3D Physical Design Unlocks New Architectural Design Points with Large Benefits
Srimani, T., Radway, R. M., Kim, J., Prabhu, K., Rich, D., Gilardi, C., Raina, P., Shulaker, M. M., Lim, S. K., & Mitra, S. (2023) · Design, Automation, and Test in Europe · doi:10.23919/date56975.2023.10137051 · Cited by 6
Игоревич, А. В., Tominov, R. V., Vakulov, Z., Rodriguez, D. J., Polupanov, N. V., & Smirnov, V. A. (2025) · Nanomaterials · doi:10.3390/nano15010075 · Cited by 5
“…in robotics; for example, Tesla cars utilize a fully self-driving computer, and their chip contains a neural network accelerator used to detect a predefined set of objects, including lane lines, pedestrians, and different types of vehicles at a very high frame rate and with a modest budget [9].”
Robotic computing system and embodied AI evolution: an algorithm-hardware co-design perspective
Yan, L., Zhao, X. J., Yang, B., Wu, Y., Dai, G. L., Li, J., Tsui, C.-Y., Cheng, K.-T. T., Zhang, Y., & Tu, F. (2025) · Journal of Semiconductors · doi:10.1088/1674-4926/25020034 · Cited by 8
Kumar, D., Suthar, N., Rodriguez, R. V., & Hemachandran, K. (2025) · Journal of Information Communication and Ethics in Society · doi:10.1108/jices-07-2025-0174 · Cited by 6
Zhang, F., Shen, A., & Huang, S. (2026) · Results in Engineering · doi:10.1016/j.rineng.2026.113673 · Cited by 18
Detecting Numerical Deviations in Deep Learning Models Introduced by the TVM Compiler
Xia, Z.-C., Chen, Y., Nie, P., & Wang, Z. (2024) · Proceedings/Proceedings - International Symposium on Software Reliability Engineering · doi:10.1109/issre62328.2024.00018 · Cited by 5
“The benefits also stimulate the uses of the compilers in practice [7–9].”
Jeong, Y., Park, J., Kim, R., & Lee, S. E. (2024) · Integration · doi:10.1016/j.vlsi.2024.102337 · Cited by 5
CDA.AI for OpenCDA: AI pathways for cooperative driving automation research
Han, X., Zheng, Z., Zhou, Z., Yun, Z., Cai, T., Liu, Y., Xiang, H., Correa-Jullian, C., Meng, Z., Huang, Z., et al. (2025) · Artificial Intelligence for Transportation · doi:10.1016/j.ait.2025.100002 · Cited by 4
Misbehavior Detection With Collective Perception in V2X Networks: A Survey
Yüce, M. F., Ertürk, M. A., & Aydın, M. A. (2025) · Transactions on Emerging Telecommunications Technologies · doi:10.1002/ett.70267 · Cited by 4
“For instance, Tesla incorporates neural network (NN) ASIC boards in their vehicles [143], greatly enhancing collective perception and awareness capabilities.”
Hosseini, M. J., Yang, Y., Bamford, S. A., Bartolozzi, C., Indiveri, G., & Nawrocki, R. A. (2026) · npj Flexible Electronics · doi:10.1038/s41528-025-00512-6 · Cited by 4
Software-Defined Vehicles: Architecting the Future of Intelligent and Connected Mobility
Ahmad, A., Hemanth, K., Kumar, O., Kumar, D. G. R. S., & Haregaonkar, R. S. (2026) · SAE technical papers on CD-ROM/SAE technical paper series · doi:10.4271/2026-26-0691 · Cited by 4
SOPHGO BM1684X: A Commercial High Performance Terminal AI Processor with Large Model Support
Gao, P., Liu, Y., Wang, J., Cai, W., Shen, G., Hong, Z., Qu, J., & Wang, N. (2024) · IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1109/micro61859.2024.00104 · Cited by 4
“…vehicles, UAVs, robots, etc. [10], [13], [19], [23], [31], [50]; Chips and cards for data centers, which typically include multiple CPUs, GPUs, programmable FPGA solutions, and programmable accelerators [11], [24], [28], [35]–[37], [40], [44], [48], [53]; and data center systems [4], [16].”
KINDRED: Heterogeneous Split-Lock Architecture for Safe Autonomous Machines
Gan, Y., Leng, J., Bo, Y., & Zhu, Y. (2025) · ACM Transactions on Architecture and Code Optimization · doi:10.1145/3711924 · Cited by 4
“We notice that currently multiple different accelerators, such as GPU, DSP, and NPU, have been utilized in building hardware for autonomous machines [4, 34, 53–55 , 11 Discussion and Future Work Lock-step CPU design methodology.”
Performance Implications of Multi-Chiplet Neural Processing Units on Autonomous Driving Perception
Odema, M., Chen, L., Kwon, H., & Faruque, M. A. A. (2025) · Design, Automation, and Test in Europe · doi:10.23919/date64628.2025.10993228 · Cited by 4
“Yet, the non-uniform distribution of performance gains, the complex dependencies of the FE and BFPN networks, and the concurrent execution requirement for 8 FE+BFPN models impose strict requirements on accelerator chiplets assignment process.”
Evaluating Energy Efficiency of Ai Accelerators Using Two Mlperf Benchmarks
Ferdaus, F., Wu, X., Taylor, V., Lan, Z., Shanmugavelu, S., Vishwanath, V., & Papka, M. E. (2025) · IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid) · doi:10.1109/ccgrid64434.2025.00035 · Cited by 5
In Search of the Performance- and Energy-Efficient CNN Accelerators
Sedukhin, S. G., Tomioka, Y., & Yamamoto, K. (2021) · IEICE Transactions on Electronics · doi:10.1587/transele.2021lhp0003 · Cited by 4
Metadata
- Authors
- Talpes, E., Sarma, D. D., Venkataramanan, G., Bannon, P., McGee, B., Floering, B., Jalote, A., Hsiong, C., Arora, S., Gorti, A., & Sachdev, G. S.
- Affiliations
- Tesla (United States) (US)
- Venue
- IEEE Micro
- Published
- 2020-03-01
- Volume
- 40
- Issue
- 2
- Pages
- 25-35
- ISSN
- 0272-1732 (ISSN-L)
- 1937-4143
- OpenAlex
- W3007788310
- Cited by
- 193
- References
- 4
- Citations by year
- 2026: 18 · 2025: 51 · 2024: 31 · 2023: 32 · 2022: 29 · 2021: 22 · 2020: 10
- FWCI
- 8.60
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE Micro (publishedVersion)
- Topics
- Physical Sciences › Engineering › Automotive Engineering › Autonomous Vehicle Technology and Safety
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
- Keywords
- system-on-chip, neural network accelerators, graphics processing units, image sensor interface
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.