Regional FPGA Power Control for Flexible Low-Power Logic
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional programmable semiconductor devices like FPGAs and PLDs are less power efficient, limiting their flexibility and increasing costs due to the need for dedicated custom integrated circuits or ASICs.
Innovation Solution
A programmable semiconductor device with dynamically controlled power regions, allowing selective programming of logic functions and power management between regions through regional power control ports and inter-chip connections, enabling dynamic power supply and conservation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If dedicated custom integrated circuits or ASICs are used to implement desirable functions, then power efficiency is improved, but flexibility and cost are worsened
Solution Approach 1:
The programmable semiconductor device is divided into multiple independently controllable power regions, each capable of being powered up or down separately. This segmentation allows the device to dynamically activate only the regions needed for current operations, achieving power efficiency similar to ASICs while retaining the reconfigurability of programmable devices.
Solution Approach 2:
The device implements dynamic power management by allowing regions to be selectively powered up or down during operation based on computational needs. This dynamic control enables the system to adapt its power consumption to actual workload requirements, combining the power efficiency of dedicated circuits with the flexibility of programmable devices.
2Adaptability or versatility
If conventional programmable semiconductor devices are used, then flexibility is improved, but power efficiency is worsened
Solution Approach 1:
By dividing the programmable device into separate controllable regions, the system can isolate and power down inactive areas while keeping active regions operational. This maintains the full flexibility of the programmable device when needed while reducing power consumption during partial operation.
Solution Approach 2:
Different regions of the device can have different power states simultaneously, with some regions active and others in low-power mode. This local differentiation allows the device to maintain flexibility in active regions while achieving power efficiency through selective deactivation of unused regions.
3Use of energy by moving object
If entire programmable device is powered down to conserve power, then power consumption is reduced, but configuration data is lost and device cannot resume operation
Solution Approach 1:
By segmenting the device into independent power regions, the system can power down only the regions that are not currently in use while keeping other regions active with their configuration data intact. This allows aggressive power saving without losing the ability to quickly resume operations by simply powering up the dormant regions.
Solution Approach 2:
Configuration data is maintained in regions even when powered down, preparing them for rapid reactivation. This preliminary preservation of configuration state allows regions to be quickly brought back online without requiring full reconfiguration, enabling fast resumption of operations after power-down periods.
Data Source
AI summary
A programmable semiconductor device capable of being selectively programmed to perform one or more logic functions includes a first region, second region, first regional power control (“RPC”), and second-to-first power control connection. The first region, in one embodiment, contains first configurable logic blocks (“CLBs”) able to be selectively programmed to perform a first logic function. The second region includes a group of second CLBs configured to be selectively programmed to perform a second logic function. The first RPC port or inter-chip port which is coupled between the first and second regions facilitates dynamic power supply to the first region in response to the data in the second region. The second-to-first power control connection is used to allow the second region to facilitate and/or control power to the first region.