Efficient Computer
Fabless semiconductor designer of energy-efficient general-purpose edge SoCs plus the compiler/toolchain (effcc) that maps standard C/C++ onto them. · United States (Pittsburgh, Pennsylvania)
Pittsburgh fabless chip startup whose Electron E1 spatial-dataflow processor claims 10-100x better energy efficiency than conventional CPUs.
Efficient Computer designs the Electron E1, a general-purpose system-on-chip built on its Fabric spatial-dataflow architecture: a tiled grid of processing elements statically scheduled by the effcc compiler so data flows between tiles instead of being fetched and decoded instruction by instruction. The company claims up to 100x better energy efficiency than conventional low-power CPUs, with roughly 1 TOPS/W (8-bit), 3 MB SRAM and 4 MB MRAM on-die, 21.6 GOPS at 200 MHz, and standard C/C++ programming plus LiteRT/ONNX model import. It sells the chip, an evaluation kit, a cloud EVK and free browser tooling (Efficient Labs) to device makers in infrastructure monitoring, industrial automation, drones/robotics, space and defense, and wearables, and is funding work to scale the same architecture toward datacenter-class parts.
Research mapping
Physical AI / Edge Compute & Control Silicon (L5) / Fabless semiconductor designer of energy-efficient general-purpose edge SoCs plus the compiler/toolchain (effcc) that maps standard C/C++ onto them.
Evidence boundary
Original analysis and independent corroboration remain under development. Commercial scale and durable technical advantage remain under review.
What makes it defensible
Data movement and instruction-fetch overhead, not peak TOPS, set the energy ceiling on physical-AI and edge intelligence.
Position in the stack: Sits in the fabless compute-silicon layer: designs the SoC and compiler that device OEMs design into battery-constrained systems, fabricated at external foundries.
Who buys it
Named customers: BrightAI
Founding team
Brandon Lucia — Co-founder & Chief Executive Officer — Kavčić-Moura Professor of Electrical and Computer Engineering at Carnegie Mellon University; founded CMU's Abstract research group; PhD and MS from University of Washington, undergraduate at Tufts; Sloan Research Fellowship, NSF CAREER Award, Bell Labs Prize.
Graham Gobieski — Co-founder & Chief Technology Officer — PhD '22 at Carnegie Mellon, co-advised by Lucia and Beckmann; research on spatial dataflow architectures and tools for battery-free/energy-harvesting devices; studied computer science at Columbia; Apple AI/ML Scholar.
Nathan Beckmann — Co-founder & Chief Architect (press kit bio also calls him Chief Scientist) — Associate Professor of Computer Science at Carnegie Mellon, founder of the CORGi research group; PhD and MS from MIT, undergraduate degrees at UCLA; Sloan Research Fellowship, NSF CAREER Award, SIGMICRO Hall of Fame.
Alex Hawkinson — Co-founder (commercial) — Serial technology entrepreneur; founder of BrightAI and SmartThings, per company About page.
SourceFinancing detail
- Last round
- $97M+ Series B (Sep 2026)
Investors: TQ Ventures, Eclipse, Union Square Ventures, Triatomic Capital, Giant Ventures, Toyota Ventures, Overlap Holdings, Box Group, RTX Ventures, Overmatch Ventures, Mana Ventures, TO Capital, TF Capital, Borderless
Mapped funds and investors
Direct relationships supported by published evidence.
Reported holdings as-of
No sourced relationships mapped yet.
Disclosed investment history
No sourced relationships mapped yet.
Showing up to four relationships in each group. ETF weights describe the fund’s portfolio, not ownership of this company. Private investment history does not establish current ownership.
Evidence & media
-
Efficient Computer, Taking on AI's Energy Problem, Announces $97M to Scale its Processors from Physical AI to the Datacenter
What this source supports: Company states $97M+ Series B at a $650M valuation, $173M total raised, Electron E1 in volume production with shipments scaling into 2027, and adoption in physical AI, infrastructure observability, space/defense and wearables.
What it does not establish: No revenue, unit volumes, contract values, named customers beyond prior BrightAI disclosure, or audited figures; 10-100x efficiency is a company claim.
Source -
Electron E1 product page and technical specifications
What this source supports: Vendor specs and positioning: up to 1 TOPS/W (8-bit), tiled reconfigurable processing elements, 3 MB SRAM plus 4 MB MRAM, C/C++ and LiteRT/ONNX support via effcc, EVK and cloud EVK availability.
What it does not establish: Vendor-supplied numbers only; no independent bench measurement, no power-at-workload curves, no comparison methodology disclosed.
Source -
Startup Launches General-Purpose Processor for AI (EE Times)
What this source supports: Independent trade-press briefing coverage: 128kB cache, 3MB SRAM, 4MB MRAM, 21.6 GOPS at 200MHz and 5.4 GOPS at 50MHz, 100x claim framed against Arm Cortex-M33/M85, BrightAI as first deployment.
What it does not establish: Specs are supplied by the company in a briefing; no independent measurement or third-party benchmark results.
Source -
Efficient Computer and BrightAI Partner to Unlock Scalable, Intelligent Edge Deployments
What this source supports: First named customer and first real-world deployment: BrightAI integrates Electron E1 into its Stateful platform monitoring water pipelines, gas compressors and power poles.
What it does not establish: No unit counts, revenue, exclusivity, contract length or measured energy savings from the deployment.
Source
Timeline
-
valuation
$650M valuation set by Series B, up from an undisclosed Series A mark seven months earlier
Source -
Financing
More than $97M Series B led by TQ Ventures; total funding $173M, proceeds to ship Electron E1 in volume
Source -
Product
Launches Efficient Labs, free browser-based EVK/energy-profiling tool suite
Source -
Financing
$60M Series A led by Triatomic Capital; total raised $76M
Source -
Product
Electron E1 Evaluation Kit (physical and cloud) made available to early-access developers
Source -
Customer
First customer partnership: BrightAI integrates Electron E1, first real-world deployment
Source -
Product
Electron E1 processor released for developer use with downloadable effcc compiler
Source -
Financing
Emerges from stealth with $16M seed round led by Eclipse
Source