Blog
Engineering updates and perspectives from our team
Interview with Semidynamics CTO Iakovos Stamoulis: AI sovereignty and Europe’s opportunity / Inside the Semidynamics approach to AI inference
Semidynamics is developing AI accelerator racks for inference, bringing CPU, vector and tensor compute together in a single architecture. CTO Iakovos Stamoulis explains how that integration simplifies programming, how Gazillion Misses keeps computation supplied with data, and how the company is building its software stack around familiar AI frameworks.
For someone coming to Semidynamics for the first time, what are you developing?
Stamoulis: Semidynamics is developing AI accelerator racks aimed at inference. We are primarily targeting LLMs
Semidynamics Rack: Why you need efficient memory, plus TOPS
TOPS are easy to print on a slide, easy to compare and a useful indication of how much work an accelerator can do. But at Semidynamics we think that number is becoming less useful on its own as a way to understand real rack performance. It does not matter how many tensor operations a chip can theo
Elevating Verification Rigour: Why Semidynamics Partnered with LUBIS EDA
At Semidynamics, the engineering team is dedicated to building high-performance AI inference systems, from terabyte-scale AIPUs to custom CPU and NPU cores. Operating at this level of complexity demands an uncompromising approach to quality. As the architecture of these systems evolves, so does the difficulty of verifying that the logic is flawless. While traditional functional verification remains a cornerstone of the process, there are areas where this rigorous simulation becomes less effective.These challenges typically arise within the most complex design blocks—the intricate logic paths where functional verification struggl
Your AI Model is Already Free
Most deployments follow the same pattern: you build and train on GPUs using a de-facto standard environment based on CUDA for R&D. Then, almost by habit, you may assume deployment also has to be CUDA-centric if you want high performance. That assumption is expensive. It nudges you into architecting y
Cervell and the Changing Shape of AI Infrastructure
AI infrastructure is no longer defined by how much raw compute you can deliver, but by how efficiently you can run it at scale. Hyperscalers are under pressure from ballooning inference costs, memory bottlenecks, and the thermal and power ceilings of their datacenters. Adding TOPS alone doesn’t solve the problem. What matters is keeping pipelines full, racks utilized, and workloads flexible across different tiers of deployment.
