Reconfigurable Hardware: ElastixAI and The Future of Fast, Efficient AI Inference
Amelia's Weekly Fish FryShow Notes
Artificial intelligence is moving faster than ever, but as AI models continue to grow in size and complexity, the challenges surrounding inference performance are becoming impossible to ignore. In this week's podcast, ElastixAI CEO Dr. Mohammad Rastegari and I chat about how we can overcome those challenges and why a different approach to AI infrastructure is necessary for the next generation of AI innovation. We also explore the key bottlenecks limiting inference performance, how ElastixAI is tackling these issues, and why FPGAs are emerging as a compelling platform for accelerating large language model inference.
More from Amelia's Weekly Fish Fry
This week we’re diving deep into the world of atomic-scale manufacturing and quantum engineering with Dr. Ludwik Kranz, Head of Quantum Engineering at Silicon Quantum Computing (SQC). Ludwik and I chat about how SQC is engineering quantum processors at the atomic scale, how to measure quantum computing progress beyond qubit count, and what applications could bring quantum computing closer to practical reality.
In this week’s podcast, we’re looking at the future of wireless engineering, and how agentic AI, digital twins, and 6G networks are coming together to change how we design, simulate, and operate connected systems. My guest week is Houman Zarrinkoub, Principal Product Manager of Wireless Communications at MathWorks. Houman and I explore why RF digital twins are moving beyond offline simulations and into live control loops, how engineering tools can keep errors and outdated data from cascading across connected systems, and how specialized AI “skills” can help keep autonomous agents within defined boundaries.
My guest this week is Tom Hazlett from sensiBel. Tom and I explore how sensiBel’s optical MEMS technology is reimagining what’s possible in acoustic performance. Tom and I will dive into how sensiBel’s optical MEMS architecture overcomes the classical constraints of legacy capacitive sensors, real-world applications where being able to detect those subtle sounds and vibrations could make a big difference and we’ll also examine the interesting intersection of acoustic sensing and AI that could open up some pretty interesting possibilities for machine perception.
In this week’s podcast, we're diving deep into the frontier of artificial intelligence and exploring how mimicking the human brain could hold the key to ultra-low-power, highly efficient computing at the edge. My guest is Sean Hehir, CEO of BrainChip. Sean and I explore their groundbreaking Akida neuromorphic processor and how BrainChip's event-based approach processes data only when meaningful changes occur—shifting away from traditional, power-hungry accelerators and paving the way for the future of always-on smart technology.
To celebrate the 700th episode of Amelia’s Weekly Fish Fry, we’re diving big into the world of microscopic tech! Tune in to discover light-powered microbots capable of hunting down bacteria, innovative memristors combining synthetic DNA and semiconductor technology for ultra-low-power computing, and breakthrough nanolasers that promise to halve energy consumption in next-generation processors.