Efficient Labs is live: free, browser-based tools for engineers building on the Electron E1 general-purpose processor. No installs. No logins. Pick a peripheral and get back the exact pins, switch settings, and copy-paste C code. labs.efficient.computer #EmbeddedSystems #EdgeComputing #EnergyEfficiency

We're excited to welcome Diwakar Tundlam to Efficient Computer as Runtime Lead Engineer. Diwakar brings deep experience in software and system engineering, with a background spanning networking and battery-powered products. Most recently a Senior Staff Engineer at Tesla on DOJO Engineering, he previously served as Senior Engineering Manager at Apple, working on battery power and health management, and Senior Manager of Software Engineering at Ciena, building dataplane software for a carrier-grade MPLS platform. He also spent four years as a Senior System Software Engineer at NVIDIA. Welcome to the team, Diwakar!

The Electron E1 is powered by our Fabric architecture, a spatial dataflow design that eliminates the bottlenecks of traditional step-by-step execution. Developers get the programming experience they already know, with dramatically higher energy efficiency. The Cloud EVK is a hosted environment where you can flash firmware, test applications, and run real energy and performance profiling on our Fabric architecture, no hardware required. Available now: Learn more: efficient.computer/announcin… #ElectronE1 #EdgeComputing #EnergyEfficiency

Vision usually gets pulled onto a companion computer because convolving every frame costs more than the airframe can spare. We measured the convolution itself. On Electron E1x, a 5x5 convolution runs 35.2x lower energy and 8.0x faster than an ARM Cortex-M33 MCU. #FabricArchitecture

Computing has an energy problem. Conventional processors waste energy moving data, and fixed-purpose accelerators can not scale as models change. Our Electron E1 general-purpose processor and Fabric architecture deliver up to 100x greater energy efficiency, measured on real applications. #ElectronE1

Welcome Gordon Cooper to Efficient Computer. Gordon was Principal Product Manager at Synopsys, leading the ARC NPX6 neural processing unit IP. He brings 35+ years in processor, DSP, and AI roles at Synopsys, NXP, Analog Devices, and Raytheon, plus service as a US Army Reserve Commanding Officer.

For a drone, every joule spent on compute is a joule not spent flying. SLAM, planning, and MPC dominate the energy budget, not perception, and an NPU does nothing for those. The Fabric architecture runs all of it together, at much lower energy. efficient.computer #FabricArchitecture #Autonomy

Most flight teams want MPC at full loop rate. Power budget usually forces a simpler controller instead. Canonical 50 Hz MPC on Electron E1x vs an ARM Cortex-M4F MCU: 13x lower power. #FabricArchitecture

The core developer story for the Electron E1 general-purpose processor is that there are zero proprietary kernel languages to learn. With standard C and C++ as input, the effcc compiler outputs directly to the Fabric architecture. This allows you to simply recompile between generations instead of re-porting.

Graham Gobieski, CTO and co-founder of Efficient Computer, was named a Most Innovative Person by Embedded Computing Design. Graham, alongside co-founders Brandon Lucia and Nathan Beckmann, developed the Fabric spatial dataflow architecture that powers the Electron E1. issuu.com/opensystemsmedia/d…

We're excited to welcome Victor Tirva to Efficient Computer as Manager of Verification and Emulation. Victor was most recently at Microsoft, where he was Verification Director and led verification of chips for Xbox, HoloLens, and AI. He has over 25 years of experience in silicon verification, validation, design, and manufacturing. Welcome to the team, Victor!

One buoy tells you the reservoir has a problem. Twenty tell you where. On Electron E1, the sensor fusion behind an out-of-range call runs at 18.8x less energy than on an ARM Cortex-M33 based MCU. Multiple seasons on the water, same power. efficient.computer

The shift is already happening. From fixed pipelines to fluid, constantly changing workloads—agentic systems are redefining execution at the edge. What was once predictable is now heterogeneous and dynamic. Compute must adapt. Read here: ow.ly/NeWP50ZKIAX #EdgeComputing #AI #Innovation

Playground shows how your code runs on the Electron E1. Write or paste C code, and the effcc Compiler maps it onto the Fabric, tile by tile, cycle by cycle. Debug at the tile level and see visual energy estimates. Try it: playground.efficient.compute…

A peel-and-stick sensor that watches a pump for five years and only pings you when it is dying. Predictive maintenance is stuck at a fork: dumb dataloggers or wired smart sensors (real intelligence, painful install). Wireless multi-year sensors are the unlock. efficient.computer

Every smart lock has a battery-life number on the box, and tamper detection is usually the feature that loses to it. On Electron E1 that classifier runs at 18.3x less energy than on an ARM Cortex-M4F based SoC. Watches for forced entry, still lasts for years. efficient.computer

New on Efficient Labs: the effprof Visualizer. Drop an effprof recording into your browser and read the run as a flame graph. Timeline, merged and sandwich views, region spans, a Fabric functional unit heat map, and an IPC strip you can brush to rescope everything. Try it: ow.ly/H7c650ZGHxE

Welcome Nazerke Turtayeva to Efficient Computer as Computer Architect. Naz defended her PhD at UC Santa Barbara on programming models and OS support for heterogeneous SoCs with custom accelerators, with internships at Google DeepMind and Microsoft Research.

A fixed gas sensor sits in a hazardous area for years, where battery swaps are costly and risky. Its classifier runs at 38.6x less energy on the Electron E1 than on an ARM Cortex-M33 based MCU. That energy buys years between service visits. efficient.computer

Most processors waste energy moving data back and forth between memory and compute. The Electron E1 general-purpose processor takes a different approach: a Fabric architecture that maps computation spatially, so data moves less and work gets done for less energy. Up to 100x greater energy efficiency than conventional processors. efficient.computer/e1-brief #ElectronE1 #EmbeddedSystems #EdgeComputing #EnergyEfficiency

We are happy to announce that Kim Yaw Tong has joined Efficient Computer as a Lead Physical Design Engineer. Prior to joining Efficient Computer, Kim Yaw was a Hardware Engineer at Tesla, specializing in physical design and SoC development for autonomous vehicles and humanoid robots. He owned memory blocks end-to-end and built in-house design flows for various PnR and Signoff steps. With nearly 20 years of experience in semiconductor physical design and custom memory IP, he previously held senior roles at Qualcomm and AMD. He holds a Ph.D. in Electrical and Computer Engineering from Carnegie Mellon University. Great to have you on the team, Kim Yaw Tong!