Partitioning to optimize AI inference for multi-core platforms
By Rami Drucker, Ceva
EDN (January 8, 2024)
Not so long ago, artificial intelligence (AI) inference at the edge was a novelty easily supported by a single neural processing unit (NPU) IP accelerator embedded in the edge device. Expectations have accelerated rapidly since then. Now we want embedded AI inference to handle multiple cameras, complex scene segmentation, voice recognition with intelligent noise suppression, fusion between multiple sensors, and now very large and complex generative AI models.
Such applications can deliver acceptable throughput for edge products only when run on multi-core AI processors. NPU IP accelerators are already available to meet this need, extending to eight or more parallel cores and able to handle multiple inference tasks in parallel. But how should you partition expected AI inference workloads for your product to take maximum advantage of all that horsepower?
To read the full article, click here
Related Semiconductor IP
- DeWarp IP
- 6-bit, 12 GSPS Flash ADC - GlobalFoundries 22nm
- LunaNet AFS LDPC Encoder and Decoder IP Core
- ReRAM NVM in DB HiTek 130nm BCD
- UFS 5.0 Host Controller IP
Related Articles
- MultiVic: A Time-Predictable RISC-V Multi-Core Processor Optimized for Neural Network Inference
- AI Edge Inference is Totally Different to Data Center
- Breaking the HBM Bit Cost Barrier: Domain-Specific ECC for AI Inference Infrastructure
- Making Strong Error-Correcting Codes Work Effectively for HBM in AI Inference
Latest Articles
- VolTune: A Fine-Grained Runtime Voltage Control Architecture for FPGA Systems
- A Lightweight High-Throughput Collective-Capable NoC for Large-Scale ML Accelerators
- Quantifying Uncertainty in FMEDA Safety Metrics: An Error Propagation Approach for Enhanced ASIC Verification
- SoK: From Silicon to Netlist and Beyond Two Decades of Hardware Reverse Engineering Research
- An FPGA-Based SoC Architecture with a RISC-V Controller for Energy-Efficient Temporal-Coding Spiking Neural Networks