ÁùºÏ²ÊÖ±²¥¿ª½±

Specialized Processing

Processors to manage massive and changing compute requirements for machine and deep learning tasks. 

• Neural processor IP supports latest high-performance neural network models, including transformer, CNN & RNN based models
• Efficient ARC Processors provide scalar and vector capabilities and customizable APEX accelerators; tightly coupled memories alleviate bottlenecks; ARC Subsystems enable key AI functions
• ASIP Designer supports custom processing with parallelism and specialized datapath elements

Memory Performance

IP solutions to support efficient architectures for different memory constraints including bandwidth, capacity, cache coherency. 

• DDR IP addresses capacity needs 
• HBM2 IP addresses the bandwidth bottleneck 
• CCIX enables cache coherency with virtualized memory capabilities for heterogeneous computing and reduced latency
• Foundation IP provides a wide array of embedded memories, logic libraries, TCAMs, and multi-port memories, with high density, low leakage, and high performance options 
  

Real-Time Data Connectivity

ÁùºÏ²ÊÖ±²¥¿ª½± provides reliable connectivity to CMOS image sensors, microphones, and motion sensors for AI applications including vision, natural language understanding, and context awareness.

• IP in advanced FinFET technologies reduce power
• Broad portfolio of silicon-proven IP support the most widely used interfaces such as MIPI, USB/Display Port, HDMI, PCI Express, CCIX, Ethernet, etc.