Power-aware FPGA design (Part 3)
pldesignline.com (February 17, 2009)
This three-part article covers several aspects of FPGA power consumption; it also provides a new look at power dissipation numbers, and questions the traditional methods of estimating and measuring power.
Part 1
– Abstract
– Introduction
– FPGA Power Components and System Profile
– Fighting Static Power
– Fighting Dynamic Power
Part 2
– Fighting Dynamic Power (continued)
Part 3
– Fighting Dynamic Power (continued)
– Proposed Power Reduction Methodology
– Conclusions
– References
Fighting logic and nets' dynamic power
Several papers have been dedicated to this subject alone [BASZ2000, Belhadj 2002]. Due to lack of space, the focus of the following sections will be dedicated only to a power profiling of some arithmetic DesignWare elements, and a brief discussion of techniques to eliminate or reduce the propagation of unnecessary glitches.
Power-aware synthesis options/constraints setting
Because of the pressure to meet timing requirements, many FPGA designers do give much attention to the synthesis process and use stringent global timing constraints. This usually leads to larger utilization of the FPGA logic elements and the associated routing resources.
Logic resources could be reduced by studying the slack distribution for each clock domain, which is the number of paths violating the required timing specification and the overall severity of these violations. For the blocks that have relaxed timing or sufficient margin in all internal paths, careful area-oriented synthesis will lead to a sizeable reduction of the needed logic resources and thus lower power dissipation for these blocks. The caveat is the need to analyze the result of the area-oriented mapping by performing a timing analysis and being alert for glitch propagation paths due to an artificially higher number of logic levels.
To summarize, setting synthesis options and constraints should be the focus, applied appropriately to timing-critical blocks or sub-blocks. Analysis of slack distribution and avoidance of global synthesis settings are keys to improving logic dynamic power dissipation.
14. Slack distribution analysis and synthesis goals setting.
(Click this image to view a larger, more detailed version)
To read the full article, click here
Related Semiconductor IP
- RVA23, Multi-cluster, Hypervisor and Android
- 64 bit RISC-V Multicore Processor with 2048-bit VLEN and AMM
- NPU IP Core for Mobile
- RISC-V AI Acceleration Platform - Scalable, standards-aligned soft chiplet IP
- H.264 Decoder
Related White Papers
- Power-aware FPGA design (Part 1)
- Power-aware FPGA design (Part 2)
- Achieving FPGA Design Performance Quickly
- How to tackle serial backplane challenges with high-performance FPGA designs
Latest White Papers
- QiMeng: Fully Automated Hardware and Software Design for Processor Chip
- RISC-V source class riscv_asm_program_gen, the brain behind assembly instruction generator
- Concealable physical unclonable functions using vertical NAND flash memory
- Ramping Up Open-Source RISC-V Cores: Assessing the Energy Efficiency of Superscalar, Out-of-Order Execution
- Transition Fixes in 3nm Multi-Voltage SoC Design