Chris Lattner

Career

Lattner worked at Apple from July 2005 to January 2017 in several positions, including managing the Developer Tools department responsible for Xcode, Instruments, Swift Playgrounds, compilers and debuggers. After briefly running the Tesla Autopilot team in early 2017, Lattner joined Google's TensorFlow team in August 2017 and led its infrastructure work, including hardware support for CPUs, GPUs and TPUs, the low-level runtime and language efforts.¹

From January 2020 to January 2022 Lattner led the engineering and product teams at SiFive, which builds RISC-V processors. In January 2022 Lattner and Tim Davis launched Modular to rebuild machine learning infrastructure, with Lattner as CEO. Modular's July 2026 announcement that Qualcomm had completed its acquisition said Lattner would become Executive Vice President of Advanced AI Software and Platforms, and that Mojo, MAX and Modular Cloud would continue as products.¹²

Compilers and languages

Lattner is the original author of the LLVM compiler infrastructure, an open-source umbrella project for compilers, debuggers, JIT systems and related tools, and also started the Clang front end. LLVM is now overseen by the nonprofit LLVM Foundation, on whose board Lattner serves.¹

Lattner began work on the Swift programming language in July 2010; it became a major focus of Apple's Developer Tools group in July 2013 and was released as open source in December 2015. Lattner remained on the Swift Core Team until 2021. At Google, Lattner created the MLIR (Multi-Level Intermediate Representation) compiler infrastructure, starting from a whitepaper in April 2018; MLIR is now part of the LLVM project.¹

Modular

Modular develops Mojo, a language for writing code that runs across CPUs, GPUs and accelerators, and MAX, a framework for serving AI models through an OpenAI-compatible endpoint. Modular released Mojo 1.0 as part of its 26.5 release in August 2026, and its open-source code is published in the modular/modular repository.⁴³⁵

Work and contributions

• MAX

  Co-founded Modular, which develops MAX. Its quickstart serves a model on a local server at localhost port 8000 and queries it with the OpenAI Python client; the guide says consumer-grade systems, including Macs, can follow along with fewer compatible models.¹³⁵

• Mojo

  Co-founded Modular, which develops Mojo, a language Modular describes as being for writing high-performance code across CPUs, GPUs, and accelerators; Mojo 1.0 was announced in August 2026.¹⁴⁵

• LLVM

  Original author of the LLVM compiler infrastructure; also started Clang, Swift, and MLIR.¹

Timeline

1. Jul 2005Joined Apple, where Lattner later managed the Developer Tools department.¹
2. Jul 2010Began work on the Swift programming language.¹
3. Dec 2015Swift was released as open source.¹
4. Aug 2017Joined Google's TensorFlow team to lead its infrastructure efforts.¹
5. Apr 2018Started the MLIR compiler infrastructure from a whitepaper at Google.¹
6. Jan 2020Joined SiFive to lead its engineering and product teams.¹
7. Jan 2022Launched Modular with Tim Davis and became its CEO.¹
8. Jul 29, 2026Qualcomm completed its acquisition of Modular, naming Lattner Executive Vice President of Advanced AI Software and Platforms.²
9. Aug 11, 2026Modular released Mojo 1.0 as part of its 26.5 release.⁴

Other documented roles

Roles the sources document that are not confirmed as current.

• Led Engineering and Product teams (January 2020 to January 2022), SiFive¹
• TensorFlow infrastructure lead, including hardware support, runtime and MLIR (August 2017 to January 2020), Google¹
• Briefly ran the Autopilot team (early 2017), Tesla¹
• Several roles including managing the Developer Tools department (July 2005 to January 2017), Apple¹

Honors and recognition

• ACM Software System Award (for LLVM) · Association for Computing Machinery¹

Where they work

• Modular 2 people in the wiki

Related people

• Tim DavisAlso at Modular
• Pete WardenInference and serving · ML frameworks and libraries
• Gavin UbertiInference and serving · ML frameworks and libraries
• Dmytro DzhulgakovInference and serving · ML frameworks and libraries
• Jeff JohnsonInference and serving · ML frameworks and libraries
• Kaichao YouInference and serving · ML frameworks and libraries
• Lin QiaoInference and serving · ML frameworks and libraries
• Patrick von PlatenML frameworks and libraries · Inference and serving
• Johannes GäßlerInference and serving · ML frameworks and libraries
• Georgi GerganovInference and serving · ML frameworks and libraries