echoloc
Get started
echoloc/ Companies/ SambaNova
Company profile

SambaNova Tech Stack

Full-stack AI inference infrastructure, from custom dataflow chips to models

Computer Hardware Manufacturing San Jose, California 201–500 employees Founded 2017 Privately Held
sambanova.ai LinkedIn

SambaNova builds AI inference systems spanning custom silicon, on-prem stacks and a cloud service. The stack pairs chip-design tooling (SystemVerilog, AXI, DDR5, Ansys HFSS) with compiler and serving layers (MLIR, LLVM, vLLM, SGLang, TensorRT-LLM), and hiring skews heavily toward senior and principal engineers, pointing to deep investment in next-generation RDU silicon and super-pod networking alongside the SambaStack serving platform.

Tech Stack 88 technologies

Core StackPython PyTorch TensorFlow C++ Docker Jenkins SystemVerilog AXI HubSpot CMS Google Analytics 4 Google Tag Manager vLLM TensorRT-LLM SGLang NVIDIA AMD Cerebras Groq MLIR InfiniBand RDMA JAX DDR5 Ansys HFSS Bazel GCC LLVM CMake DeepSeek Llama+57 more

What SambaNova Is BuildingChallenges and projects from their job posts

◆Challenges

  • Optimizing AI inference performance and memory bottlenecks
  • Bridging multi-cloud and on-prem environments consistently
  • Scaling platform for internal and external customers
  • Closing reliability gaps
  • High demand for hardware engineers
  • Reducing financial close cycle time

▲Active Projects

  • SambaStack inference serving platform
  • Next-generation RDU chip development and verification
  • Super-pod scale ML hardware networking architecture
  • Production GenAI applications on SN40L
  • Emulation, FPGA prototyping and PCIe Gen5/6 validation
  • Joint go-to-market and partner marketing campaigns

Hiring Activity

Steady75 roles · 20 in 30d

Department

Engineering
54
Marketing
7
Product
3
Sales
3
Finance
2
HR
2
Manufacturing
2
Support
2

Seniority

Senior
38
Principal
23
Mid
6
Director
4
Manager
4
Lead
1
VP
1

Notable leadership hires: Director Technical Accounting, Support Lead

Companies like SambaNova

100+ Computer Hardware Manufacturing companies running Python

See the full list

About SambaNova

SambaNova is a San Jose, California-based company founded in 2017 that develops AI infrastructure covering chips, systems, software and foundation models. It sells to developers, enterprises, governments, data centers and cloud service providers. Customers can use a hosted cloud offering with OpenAI-compatible APIs serving open models such as Llama and DeepSeek, deploy SambaStack on-premise, or run SambaManaged as a ready-to-deploy AI cloud. The privately held company employs 201-500 people and hires across the United States, United Kingdom, India, Singapore, Japan and Sweden.

HeadquartersSan Jose, California
Company Size201–500 employees
Founded2017
Hiring MarketsUnited States, United Kingdom, India, Singapore, Japan, Sweden

Frequently Asked Questions

What tech stack does SambaNova use?

SambaNova's stack spans hardware design (SystemVerilog, AXI, DDR5, InfiniBand, RDMA), ML frameworks (PyTorch, JAX, TensorFlow), compilers (MLIR, LLVM, GCC) and inference serving tools like vLLM, SGLang and TensorRT-LLM.

What is SambaNova working on?

Current work includes the SambaStack inference serving platform, next-generation reconfigurable dataflow unit (RDU) chips and verification, super-pod scale ML networking, and production GenAI applications on the SN40L.

What countries does SambaNova hire in?

SambaNova hires in the United States, United Kingdom, India, Singapore, Japan and Sweden, with demand concentrated in senior and principal engineering roles.

Similar Companies in Computer Hardware Manufacturing

Other companies in the same industry, closest in size

L
Lightmatter
Computer Hardware Manufacturing
E
Etched
Computer Hardware Manufacturing
A
Ayar Labs
Computer Hardware Manufacturing
X
Xanadu
Computer Hardware Manufacturing

How this profile is built

SambaNova's technology stack, projects, and hiring signals are inferred from public hiring and company data — career pages, public listings, and company web presence — then clustered and de-duplicated. Figures are estimates that refresh over time. Read our full methodology →

This is not an official vendor or customer list. It is a technology-adoption signal inferred from public data, intended for B2B research.