Programming Examples

• DRAM Loopback
  • Device initialization
  • Program setup
  • Create buffers in DRAM and L1 (SRAM)
  • Sending real data into DRAM
  • Creating a data movement kernel
  • Setting runtime arguments for the data movement kernel
  • Running the program
  • Download the result and verify output
  • Validation and teardown
• Eltwise binary
  • Program setup
  • Circular buffers
  • Data movement and compute kernels
  • Download the result and verify output
  • Validation and teardown
• Eltwise SFPU
  • Program setup
  • The kernels
  • Set up runtime arguments
  • Program execution and final check
  • Conclusion
• Matmul (Single Core)
  • Device Initialization & Program Setup
  • Data Preparation and Golden Reference
  • DRAM Buffer Allocation
  • Circular Buffer Orchestration for Pipelined MatMul
  • Matmul Kernel Pipeline Breakdown
    • The reader kernel
    • The compute kernel
    • The writer kernel
  • Kernel exexution and result verification
  • Conclusion
• Matmul (Multi Core)
  • Accessing all the cores
  • Splitting the work across cores
  • Using different kernels for reader/writer
  • Compute kernel args
  • Reader/writer kernel runtime args
  • Conclusion
• Matmul (Multi Core Optimized)
  • Data Reuse in matmul_multicore_reuse
    • Fine-Grained Block Size Control
    • Intermediate Circular Buffer Configuration
    • Stride Kernel Arguments
    • Intermediate Results Handling
    • Conclusion
  • Data Multicasting in matmul_multicore_reuse_mcast
    • Additional Compile-Time Argument
    • Configuring Core Ranges for Tile Distribution
    • Circular Buffer Creation for CoreGrid
    • Multicast Reader/Writer Kernel Setup
    • New Compute Kernel: Fused Bias Addition and Activation Functions
    • Semaphores
    • Kernel Runtime Arguments