Title: PowerPoint Poster Template
1Rapid Prototyping and Emulation of Many-Core Chip
Multiprocessors with Integrated Hardware
Accelerators
Colin J. Ihrig University of Pittsburgh Email
cji3_at_pitt.edu
ACME Actor Generator
ACME Actor Generator
SuperCISC Compiler
- Problem
- Need to study new architectures
- System design is time consuming
- Software simulators do not scale well
- Orders of magnitude slowdown
pragma HWstart z x y m y ltlt 3 n m
y if ( n lt 0 ) n 0 if ( q 3 ) i z
5 else i z 2 j n i pragma HWend
- C to VHDL automation flow
- Annotate C code with pragmas
- Construct Super Dataflow Graph
- Custom coprocessors in ACME
- ACME
- Graphical design entry
- Uses Ptolemy II environment from UC Berkeley
- Components called actors
- Generate systems targeting FPGAs
- System emulation
- Rapid MPSoC prototyping
- Processor and logic design
- Skeleton code generated for
- Ptolemy Java Actor
- Java Native Interface C and Header files
- ACME VHDL Actor
- Actors automatically incorporated into ACME and
Ptolemy II
- Processor Based Actors
- Design complex components
- Example Switch arbiters
- Describe functionality in C
- Use Java Native Interface within Ptolemy II
- Soft core processors run code on FPGA
Actor Generator GUI
ACME actor library mirrors Ptolemys Java library
Xilinx library contains IP blocks and board
descriptions
Generated skeleton code for actor
Extend Ptolemy II GUI for graphical actor creation
- SuperCISC Compiler
- C to VHDL compiler
- Processor actors are bottleneck
- Hardware acceleration
SoC Generation Tool Flow
Ptolemy II Graphical Model Creation
Xilinx Platform Studio System
Emulation Augmentation
User specified latency and throughput circuit
2x2 Mesh Interconnect Network
2x2 Mesh Interconnect Network
Network Switch
Three cycle throughput
Logic Containing VHDL Actors
- Emulation clock
- Target vs. Host FPGA cycles
- Decouples emulated system from FPGA
- Tracked via hardware counters
One additional latency cycle
Processors set / reset barrier
Microblaze Processor Systems
Fast Simplex Link Bus Connections
Processor / hardware synchronization via a
hardware barrier circuit
Serial Port for PC Communication
Barrier clocks custom logic
Processing Node