Events

Events​

October 24

GlobalFoundries Tech Conference
Shanghai

Visit our table for the latest validation information & evaluation board demo for GF14/12 EFLX eFPGA. Multiple customers are in design of chips using GF14/12 EFLX eFPGA.

October 23-24

Linley Fall Processor Conference
Santa Clara CA

Cheng Wang will discuss how we make accurate performance estimates for models like YOLOv3 on the InferX X1 Inference Co-Processor: high throughput for large/megapixel models at low power and low cost.

Visit our table for demos of our nnMAX Compiler performance modeler for ONNX and TensorflowLite.

October 29

TSMC China OIP
Shanghai

Visit our booth for the latest information on eFPGA for TSMC 6, 7, 12, 16, 22, 28 & 40nm, including hardware demos for TSMC 16FFC & 28HPC.  Also get information on the new AI Inference IP which is available for integration on TSMC16FFC and will be validated in silicon by March 2020.

November 19

SemIsrael
Tel Aviv, Israel

Attend our presentation on our nnMAX AI Inference IP available now and in fab for validation in TSMC16FFC – best throughput/$ and throughput/watt.

Visit our table for demos of our nnMAX Compiler performance modeler for ONNX and TensorflowLite. And for the latest information on eFPGA for GF 12/14 and TSMC 6/7/12/16/22/28/40nm.

news

News & Press 2019

22 July, 2019, TechCrunch, Powering the Brains of Tomorrow’s Intelligent Machines

10 July, 2019, Semiconductor Engineering, Efficient AI Inference

1 July 2019, EENews Europe, ResNet-50, a misleading machine learning benchmark for megapixel images

2 July 2019, Semiconductor Engineering, TOPS, Memory, Throughput and Inference Efficiency

27 June 2019, Embedded Computing Design, Evolution of eFPGA

6 June 2019, SemiEngineering, Accelerating Endpoint Inferencing

6 June 2019, SemiEngineering.com, Do Large Batch Sizes Always Improve Neural Network Throughput?

5 June 2019, SemiEngineering.com Geoff Tate talks about Improving the Efficiency of Inference Hardware

4 June 2019, Cheng Wang’s Presentation at AI Hardware Summit Beijing on InferX http://flex-logix.com/wp-content/uploads/2019/06/AI-HW-Summit-Asia-2019.pdfX1 – In Mandarin

3 June 2019, D&R Interview with CEO Geoff Tate at DAC

June 1, 2019, Semiconductor Engineering, How to Integrate an Embedded FPGA

News & Press 2018

December 31, 2018, SemiEngineering.com, Top Tech Videos of 2018, including Cheng Wang’s “Inferencing in Hardware”

December 20, 2018, SemiWiki, AI at the Edge

December 6, 2018, SemiEngineering.com, High Neural Inferencing Throughput at Batch=1

December 6, 2018, SemiEngineering.com: Looking Beyond the CPU

December 11, 2018, AI Edge Summit: Updated Technical Slides on NMAX Architecture.

August 30, 2018, EE Times: Harvard takes AI to new low for IoT

August 23, 2018, EE Times China, Flex Logix:eFPGA开发板促进eFPGA应用

August 22, 2018, Quantum Leap Technical Sales Now Representing Flex Logix in North America.

August 16, 2018, Electronic Engineering Journal, Where’s the CNN Synthesis?

August 12, 2018, EE Times China, Flex Logix:eFPGA开发板促进eFPGA应用

August 6, 2018, Semiconductor Engineering, Video Interview: eFPGA for Neural Networks.

August 6, 2018, Embedded Systems Engineering, Optimizing AI and Machine Learning with eFPGAs.

August 3, 2018, Circuit Cellar, FPGA Solutions Evolve to Meet AI Needs (link opens PDF of the august issue; article is on page 43 of the PDF)

August 2, 2018, SemiEngineering, First Public Disclosure on the use of eFPGA in an SoC.

June 13, 2018, Electronics Weekly, Getting the measure of the eFPGA market.

July 18, 2018, SemiWiki, Machine Learning and Embedded FPGA IP.

July 20, 2018, SemiEngineering Roundtable, Architecting for AI.

July 9, 2018, SemiEngineering, FPGAs (and eFPGAs) Drive Deeper Into Cars.

July 19, 2018, Tech Forum, Embedded FPGA Starts to Take Hold in SoCs.

July 20, 2018, Electronic Engineering Journal, Programmable Pile of Parts.

July 10, 2018, Semiconductor Engineering, Reconfigurable AI Building Block for SoCs and MCUs.

July 9, 2018, Semiconductor Engineering, Physical Design Considerations in selecting eFPGA (VIDEO).

July 2, 2018, Electronic Design, How to Validate Embedded FPGA.

March 31, 2018, Barron’s: A New AI Era Dawns for Chip Makers (IP Licensing)

March 27, 2018, SemiWiki, Configurability for Embedded FPGA Hard IP.

March 2, 2018, Semiconductor Engineering Tech Talk: eFPGA Density.

March 1, 2018, Semiconductor Engineering: eFPGA, Think Differently & Experiment.

February 27, 2018, Flex Logix Co-Founder Cheng Wang Awarded Interconnect Patent for Tiling of eFPGA Cores to Creat Wide Range of Larger eFPGA Arrays

February 26, 2018, Semiwiki, The Hierarchical Architecture of an Embedded FPGA.

February 26, 2018, EE News Europe, CEO Interview: Flex Logix’ Geoff Tate on Licensing eFPGA.

February 14, 2018, Electronic Products: 2018 The Year of the Embedded FPGA.

February 15, 2018, DataCenter Journal: Embedded FPGA, Coming to a Data Center Near You.

February 12, 2018, The Information: Amazon is Becoming an AI Chip Maker, May Be Integrating eFPGA.

February 12, 2018, SemiEngineering: Interview with Flex Logix CEO, The Race to Accelerate.

February 1, 2018, SemiEngineering: Introduction to eFPGA Hardware.

News & Press 2017

December 27, 2017, Electronic Products: Six High-Tech Companies to Watch in the New Year.

December 22, 2017, SemiWiki: The Year of the eFPGA, 2017 Recap – embedding RAM in eFPGA

December 21, 2017, All About Circuits: SiFive DesignShare Program.

December 19, 2017, Embedded Computing Design: Using Embedded FPGA as a Reconfigurable Accelerator.

December 19, 2017, Programmable Logic Holds the Key to Addressing Device Obsolence. [applies to eFPGA as well]

December 14, 2017, SemiEngineering: Tony Kozacuzk discussing reconfigurable Accelerators.

December 8, 2017, SemiEngineering: “Get eFPGA with your CPU Now”

December 6, 2017, Geoff Tate Keynote Speech at D&R IP-SoC Grenoble “eFPGA Everywhere (Eventually)” & a 3 minute video interview on the same topic.

December 4, 2017, DataCenter Dynamics: Future-Proofing the Data Center with eFPGAs.

July 2017, COTS Journal: FPGA: It’s all About Integration.

July 27, Semiconductor Engineering: The Secret Life of Accelerators.

July 21, 2017, Flex Logix’s EFLX embedded FPGA Accelerates Processor Performance by 40-100x

July 20, 2017, SemiWiki: Embedded FPGA as Functional Accelerators.

July 18, 2017, Electronic Design: How to Take Advantage of Embedded FPGA.

July 13, 2017, Semiconductor Engineering: Is the IP Industry Healthy?

March 19, 2017, SemiWiki: What’s better than silicon-proven IP? Lab bench proven!

March 16, 2017, Electronic Engineering & Product World China: 为何说嵌入式FPGA改变了芯片和SoC的未来设计方式

March 13, 2017, Electronic Design: 11 Myths About Embedded FPGAs.

January 31, 2017, Embedded Systems Engineering: Extending Battery Life in IoT Devices.

January 30, 2017, Embedded Computing Design: 2017 embedded processor report: At the edge of Moore’s Law and IoT.

January 26, 2017, Defense Systems: DARPA licenses emerging chip technology.

January 26, 2017, Chip Design: EDA in the year 2017 – Part 2 [embedded FPGA predictions]

January 26, 2017, Electronic Products: DARPA teams with Flex Logix to develop FPGA technology for government agencies designing ICs [NOTE: its actually for any company or agency designing ICs for use by the US government]

January 24, 2017, Chip Design: Make Chips Do More & Last Longer with Embedded FPGA

January 23, 2017, SemiWiki: DARPA, Flex Logix and TSMC!

January 23, 2017, SemiAccurate: Flex Logix Announces a First Customer, DARPA.

January 23, 2017, EE Times: Flex Logix Signs Licensing Deal with DARPA.

January 23, 2017, Electronics Weekly: Flex Logix Signs eFPGA Deal with DARPA.

January 23, 2017, New Electronics: DARPA Licenses Embedded FPGA Technology for US Government Projects.

January 23, 2017: Flex Logix Works with DARPA to Develop Flex Logix Embedded FPGA IP for Government Projects.

January 17, 2017, SemiWiki: The Year of the eFPGA

January 3, 2017, All About Circuits: Embedded FPGA: Changing the Way Chips are Designed

News & Press 2016

December 19, 2016, EDACafe: Flex Logix & TSMC: embedded FPGA cores on 16nm FinFETs

December 13, 2016, Electronic Engineering Journal: Future Proofing your FinFET Design.

December 13, 2016, SemiWiki: Flex Logix is a Different Kind of IP Company!

December 13, 2016, Electronics Weekly: Flex Logix completes 16nm eFPGA core design.

December 13, 2016, EE Times Europe: FPGA fabric offered for TSMC 16nm FinFET.

December 13, 2016: Flex Logix ~1GHz Embedded FPGA IP Core Now Available for TSMC 16FF+ and 16FFC.

December 8, 2016, EE Times: ACE Awards [Flex Logix’ EFLX co-finalist with Intel and TI: great company to be in]

December 5, 2016, EE Times: Can FPGA Fabric and an SoC Co-Exist on the Same Chip?

December 1, 2016, REUSE2016 Keynote Talk: Embedded FPGA for Architects & Physical Designers.

December 1, 2016, Electronics Weekly: Flex Logix’ switchable LUT blocks for SoCs

December 1, 2016, Electronics Weekly: Multiple Vendors Pursue eFPGA

October 31, 2016, Semiconductor Engineering: The Battle to Embed the FPGA.

October 27, 2016, Semiconductor Engineering: Embedded FPGAs Going Mainstream? [NOTE: Mr. Wingard of Sonics says that embedded FPGA, like FPGA chips, want the largest number of metal layers a foundry can provide. However, EFLX embedded FPGAs only require 4-6 metal layers making them compatible with chips using the minimum metal stack!]

October 26, 2016: Embedded Systems Engineering: Microcontroller Architects turn to embedded FPGA for flexibility.

October 24, 2016, Embedded Systems Engineering: Flex Logix mentioned in See the Latest Developments at ARM TechCon.

October 14, 2016, SemiWiki: Adding Reconfigurable DSP hardware shrinks energy for MCU core.

October 14, 2016, Electronic Design: Q&A with Cheng Wang on Embedded FPGA.

October 10, 2016, SemiWiki: CEO Interview, Geoff Tate of Flex Logix.

October 5, 2016, SemiWiki: Microsoft, FPGAs, and the Evolution of the DataCenter.

Flex Logix Technologies Introduces EFLX: an FPGA Core for Upgrading Silicon

Company launches from stealth with founders receiving prestigious Lewis Award for Outstanding Paper at ISSCC. 
Led by Founding Rambus CEO Geoff Tate, Flex Logix is backed by Series A led by Lux Capital

Mountain View, CA, February 23, 2015Flex Logix Technologies, a pioneer in technology that lets chips change, today introduced EFLX, an innovative FPGA (field programmable gate array) core that allows chip makers and equipment manufacturers to update normally fixed functions in silicon at will. The company’s technology platform enables significant new customer benefits by dramatically reducing design and manufacturing risks, accelerating technology roadmaps and bringing greater flexibility to hardware.

Flex Logix is also proud to announce that co-founders Cheng C. Wang, Fang-Li Yuan and UCLA professor of electrical engineering Dejan Markovic have received the prestigious Lewis Award for Outstanding Paper at the International Solid State Circuits Conference (ISSCC), one of the world’s premier chip design conferences, for a paper1 outlining concepts embodied in EFLX. Recent Lewis Award winners include IBM, Toshiba, Nvidia and Sandisk. Other groundbreaking research unveiled at ISSCC include some of the first papers on digital signal processors (Bell Labs, 1980), RISC processors (UC Berkeley, Stanford, 1984) and 1 GHz processors (Intel, 2000).

EFLX is a fully-functional FPGA that, instead of existing in systems as a standalone chip, is integrated into communications chips, microcontrollers and other devices. EFLX cores can be used to upgrade I/O protocols, change encryption algorithms to improve security, enable elements of software-defined radio or accelerate data center algorithms like search. EFLX can be leveraged by manufacturers or end-users to upgrade new products or individual systems already installed in the field.

EFLX FPGA cores can fit into chips selling for hundreds of dollars to less than one dollar.

“We give people a way to change their silicon without changing their silicon,” said Geoff Tate, Flex Logix CEO and co-founder. “FPGA technology has been employed for years in markets where performance requirements are stringent and real-world uses are unpredictable. With EFLX cores, we can bring what’s great about FPGAs to a broader spectrum of customers and users.”

Flex Logix will license EFLX and associated software technology to semiconductor manufacturers. Tate is a recognized leader in semiconductor IP licensing. Prior to Flex Logix, he served as CEO of Rambus for 15 years, leading the company from a four-person startup to a publicly-held company with a market capitalization of over $2 billion dollars. Before that, Tate was a senior vice president at AMD responsible for microprocessors. Lux Capital is the lead Series A investor in Flex Logix.

FPGAs and Flex Logix

Flex Logix’s goal is to empower chip designers and their customers to push the boundaries of performance and technology with far less risk. Over the past several years, chip design costs have been rising sharply, according to analyst reports. Developing new mask, or circuit pattern, to upgrade a single chip can cost $2 million to $5 million. Missing product cycles because of design or technology issues can mean millions in lost revenue. Similarly, end-users often have to replace equipment or endure sub-optimal performance because of unanticipated in-field conditions.

A growing $5 billion market, standalone FPGAs allow hardware to adapt to new demands and technologies, but they add substantial cost: a large percentage are employed in higher-end markets like aerospace, automotive and medical imaging. FPGAs consist of identical circuit blocks connected through an intricate network: increasing their performance typically means increasing costs and board space.

Flex Logix cores are based around a unique hierarchical network that reduces the length of communication links between logic blocks, a breakthrough achieved by co-founder Cheng Wang and explained in the award-winning ISSCC paper. This innovative network architecture reduces overall power consumption and improves performance. It also reduces the area required for interconnects by 50% or more and the number of metal layers, which in turn allows EFLX cores to be integrated into other devices. Standalone FPGAs cannot be easily retrofitted for integration.

Flex Logix has begun to offer its first core, a 2,500-LUT core design in TSMC’s mainstream 28 nanometer HPM process, to manufacturers through licensing contracts. An individual 2,500-LUT (look up table) core is expected to add less than 15 cents to the total manufacturing cost of a device. Designers will also integrate multiple EFLX cores: EFLX cores can be integrated into multiple circuit blocks in the same design or tiled together to create arrays of 7 x 7 or more EFLX cores for greater performance. Flex Logix will also develop smaller EFLX cores that will cost fractions of a cent as well as EFLX cores with more LUTs to serve a greater number of applications. The company is in discussions with a number of companies, especially in communications and networking. 

“Flex Logix has the potential to dramatically reshape the semiconductor industry,” said Peter Hebert, co-founder and Managing Partner of Lux Capital. “Adaptable silicon is a really big idea – and we’re proud to be behind Geoff and his team as they bring their breakthrough to market. If we’re successful, billions of new integrated chips will be based on Flex Logix’s innovative EFLX technology.”

  1. The paper—“A Multi-Granularity FPGA with Hierarchical Interconnects for Efficient and Flexible Mobile Computing”—was authored by Cheng C. Wang, Fang-Li Yuan, Tsung-Han Yu and Dejan Markovic.