Skip to main content

Blaize Holdings, Inc.

Category: AI Chips / Semiconductors

Blaize provides purpose-built, full-stack AI edge computing solutions with programmable processor architecture and low-code/no-code software platform for real-time AI inference from edge to data center. Blaize Holdings, Inc. was founded in 2011. The company is led by Dinakar Munagala. Based in El Dorado Hills, California, United States. Team size: 201-500. Total funding raised: $365.0M. Latest round: Post-IPO. Key investors include DENSO; Mercedes-Benz Group AG; Magna International; Samsung Catalyst Fund; Temasek; Franklin Templeton; GGV Capital; Bess Ventures; BurTech LP LLC; Rizvi Traverse; Ava Investors; Polar Asset Management Partners; SPARX Group; Wavemaker Partners; SGInnovate; Dado Banatao.

Founded
2011
Headquarters
El Dorado Hills, California, United States
Team size
201-500
Total funding
$365.0M

Value proposition

Blaize delivers energy-efficient, programmable AI computing solutions that enable real-time inference at the edge without cloud dependency, combining custom GSP (Graph Streaming Processor) architecture with a no-code AI software platform to reduce deployment complexity, power consumption, and total cost of ownership.

Products and solutions

Blaize Pathfinder AI Edge Platform; Blaize Xplorer AI Edge Platform; Blaize AI Studio (no-code edge AI software platform); Blaize AI Services Platform (API-delivered AI applications); Blaize Picasso SDK; Blaize GSP (Graph Streaming Processor) architecture; Automotive AI SoCs (chiplet design)

Unique value

Blaize's proprietary GSP (Graph Streaming Processor) architecture is purpose-built for AI workloads, offering high performance at ultra-low power consumption with full programmability — unlike fixed-function NPUs or power-hungry GPUs. The no-code AI Studio platform dramatically reduces edge AI deployment time and complexity.

Target customer

Automotive OEMs and Tier-1 suppliers, defense and critical infrastructure operators, smart vision/security system integrators, industrial automation providers, enterprise edge computing customers, and cloud service providers.

Industries served

Automotive (autonomous driving, EVs); Defense & Critical Infrastructure; Smart Vision / Security; Industrial Automation; Smart Retail; Healthcare; Smart City / Metro; Enterprise Computing

Technology advantage

Graph Streaming Processor (GSP) architecture — a programmable, graph-native processor designed specifically for neural network computation; full-stack hardware + software integration; American-made processor architecture; chiplet design for automotive scalability; low-code/no-code AI Studio platform spanning the entire edge AI operational workflow; energy efficiency enabling offline inference in rugged/defense environments.

How they differentiate

Unlike NVIDIA's GPU approach (high power) or fixed-function NPUs (limited flexibility), Blaize's GSP architecture is fully programmable and graph-native, achieving superior power efficiency for edge inference. The no-code AI Studio platform provides a complete end-to-end workflow from model training to edge deployment, lowering the barrier for non-expert users. Blaize also emphasizes American-made manufacturing and has deep automotive/defense vertical focus.

Main competitors

NVIDIA (Jetson edge AI platform); Intel (edge AI processors); Hailo (edge AI accelerators); Cerebras Systems; Groq; EdgeCortix; DeepX; Ambiq

Key partnerships

DENSO Corporation (strategic investor & automotive partner); Mercedes-Benz Group AG (strategic investor & automotive partner); Magna International; Samsung Catalyst Fund; Nokia (strategic MoU for hybrid AI inference, Jan 2026); NeoTensr (up to $50M contract for edge AI data centers, April 2026); PT Datacomm Diangraha (MoU for hybrid AI inference in Indonesia/SE Asia, April 2026); Winmate Inc. (strategic partnership for rugged defense/critical infrastructure systems, May 2026); GIGABYTE (edge AI solution integration)

Notable customers

DENSO Corporation (joint development agreement); Mercedes-Benz Group AG (multi-phase master services agreement); NeoTensr (up to $50M contract for edge AI data centers); Nokia (strategic MoU for hybrid AI inference); Winmate Inc. (strategic partnership for defense systems)

Major milestones

Founded 2011 as ThinCI; Emerged from stealth and rebranded to Blaize in Nov 2019; Launched Pathfinder and Xplorer edge AI platforms; Announced SPAC merger with BurTech Acquisition Corp. (Dec 2023); Closed SPAC merger and listed on Nasdaq as Blaize Holdings, Inc. (Jan 2025); Signed strategic partnership with Winmate for defense/critical infrastructure AI (May 2026)

Growth metrics

Publicly traded on Nasdaq (BZAI, BZAIW) since Jan 14, 2025. Q1 2026 revenue of $5.8M (missed estimates). FY2026 revenue guidance of $130M reaffirmed. Raised $35M in public offering (May 2026). Cash balance of $33.3M as of Mar 31, 2026. Over $400M qualified pipeline. 200+ employees globally.

Market positioning

Publicly traded (Nasdaq: BZAI) edge AI computing company positioned as a specialized alternative to NVIDIA and Intel in the edge inference market. Valued at ~$1.2B at SPAC merger (Jan 2025). Targets the $34.3B+ AI accelerator market with a focus on automotive, defense, and industrial edge applications where power efficiency, programmability, and offline capability are critical.

Geographic focus

Global, with headquarters in El Dorado Hills, CA (US) and offices in San Jose, CA; Cary, NC; Hyderabad (India); Leeds & Kings Langley (UK); Abu Dhabi (UAE). Strong presence in US, Europe, Japan, and India.

Patents and IP

Multiple patents filed. CEO Dinakar Munagala credited with 5 IP/patent filings; CTO Satyaki Koneru credited with 6 IP/patents and 10 filings; Chief Scientist Ke Yin credited with 6 IP/patents and 6 filings.

About Dinakar Munagala

Ex-Intel (12 years, graphics processor micro-architecture and design); B.E. Electrical & Computer Engineering from Osmania University, India; M.S. Electrical & Computer Engineering from Purdue University, USA

Official website: