Tesla's AI5 SoC tapes out at Samsung Foundry with 2nm-class node; production starts soon after TSMC tape-out.

Tesla's AI5 chip has reached tape-out at Samsung Foundry and is scheduled for mass production using the company's 2nm-class process technology at the Taylor fab, according to James Kim, a principal engineer at Samsung Foundry, who disclosed the milestone in a LinkedIn post.

"The Tesla-Samsung AI5 chip has reached tape-out," Kim wrote. "It is scheduled to be manufactured at the Taylor fab using our latest 2nm process and will soon be integrated into Tesla's newest products. It has been an honor to collaborate with the outstanding engineers at Tesla Palo Alto and Austin over the past several months."

Elon Musk first demonstrated the AI5 in mid-April and revealed that Tesla plans to manufacture the processor concurrently at both TSMC and Samsung Foundry. The TSMC version reached tape-out several months earlier than the Samsung variant.

The AI5 processor module demonstrated by Musk in April integrates a relatively compact accelerator die—roughly half a reticle in size—alongside 12 SK Hynix memory packages that appear to be standard GDDR6 or GDDR7 devices. The package uses an organic substrate with memory components labeled similarly to conventional discrete DRAM chips.

Tesla has not disclosed the width of the AI5's memory subsystem, but the presence of 12 memory packages suggests a relatively broad external memory interface. Assuming the module uses 12 GDDR6 or GDDR7 ICs, the processor would feature a 384-bit memory bus, delivering memory bandwidth ranging from 768 GB/s to 1.536 TB/s, depending on the memory technology and transfer rates employed.

While Tesla has not disclosed peak compute performance or other detailed specifications, Musk has claimed that in certain workloads, AI5 can deliver performance improvements of up to 40X compared to its predecessor. Musk expects AI5 to become one of the most-produced chips ever, which is why Tesla is using two foundries for manufacturing. The company plans to deploy AI5 in Tesla vehicles, Tesla robots, and Tesla data centers.