Smart Cars May Pack 100GB+ DRAM, Up to 1.5TB NAND
Memory shortages and rising prices have become so severe that even Apple has raised the prices of some Macs. Yet beyond smartphones, PCs, and AI data centers, the automotive industry is quietly emerging as a major consumer of memory. So, how much memory does a modern vehicle actually need? Micron previously projected that Level 4 autonomous vehicles would require more than 300GB of memory. But a closer look at today’s most advanced vehicles suggests that 300GB is not an exaggerated estimate—in fact, it may even be conservative.
According to TechNews, Tesla’s HW 4.0 autonomous driving system is equipped with 32GB of GDDR6/LPDDR memory, while this year’s AI 4+ doubles that to 64GB. NVIDIA’s Drive Orin platform is likewise available with 32GB and 64GB configurations. However, these account for only part of a vehicle’s total memory capacity.
How Smart Vehicles Consume DRAM
As the report highlights, next-generation smart vehicles are increasingly equipped with powerful infotainment systems. Mercedes-Benz’s MBUX platform features an eight-core processor and 24GB of memory to support streaming, UI animations, head-up display graphics, and other visual effects. The BMW iX3 and future Neue Klasse models are also estimated to use 16–24GB of memory for their panoramic display and Operating System X.
Beyond autonomous driving and infotainment, automakers are increasingly integrating AI assistants. The report indicates that Mercedes-Benz has demonstrated localized AI models running directly on the vehicle through its MB.OS platform, requiring at least 4–12GB of DRAM. Taken together, the report estimates that today’s digitally advanced vehicles require an additional 4–8GB of DRAM for various electronic control units (ECUs). Combined with the major systems above, a modern vehicle is estimated to use at least 40GB of DRAM, with high-end configurations potentially exceeding 100GB.
The Other Memory Challenge: NAND Flash
However, the figures above account only for DRAM. NAND flash used for storage has also been affected by the ongoing memory shortage. As the report indicates, over-the-air (OTA) software updates have become an essential feature of modern smart vehicles. Routine OTA updates are typically around 1–2GB, while major releases can exceed 10GB. Because a failed update could pose safety risks, vehicles require redundant storage, with roughly 50GB reserved for OTA updates alone.
In addition, autonomous driving systems must store massive amounts of image data, requiring at least 100–300GB of storage. Infotainment systems, AI models, and electronic control units (ECUs) further increase NAND demand. Overall, the report estimates that a Tesla vehicle requires around 500GB of NAND flash, while a Mercedes-Benz model may require as much as 1–1.5TB.
Notably, as the report points out, these estimates are based on vehicles available in 2026. Tesla’s upcoming AI5 chip is rumored to feature up to 192GB of memory. If adopted in future mass-produced vehicles, total DRAM capacity could exceed 200GB, driving storage requirements even higher.
