Throughput‑focused inference
Inference pipeline optimized for maximum throughput across large datasets
Cervell™ NPU
C32
Clustered IP to Scale Throughput
When one engine isn’t enough, Atrevido Cluster multiplies throughput under the same ISA and memory‑first design (CPU/Vector, optionally Tensor‑equipped instances)
Key Benefits
Effortless Scale-out
Scale‑out without software rewrites
Balanced by design
Compute kept busy by Gazillion™‑driven bandwidth across the fabric
Ready Today, Future-Proof
Silicon‑ready today; aligned with future boards/chiplets
Architecture Highlights
Cluster Integration Reference
Reference cluster topologies and coherency/NoC integration notes
Sustained Data Flow
Gazillion™ across the fabric to sustain data flow
Unified Programming Model
Unified programming model across single and multi‑engine systems
Software Path
Same RISC‑V ISA from
single‑engine to multi‑engine deployments
single‑engine to multi‑engine deployments
Typical Deployments
Batch analytics
Accelerating large, complex data analyses performed on scheduled, high-volume data sets
Multi‑stream vision
Enabling simultaneous, low-latency data processing and inference across multiple data flows
Find out more
Our IP is silicon-ready, and in silicon implementations. Speak to us about reference designs
Learn more