High performance interconnect architecture for field programmable gate arrays
View Patent ↗This invention relates to a high performance interconnect architecture providing reduced delay minimized electro-migration and reduced area in FPGAs comprising a plurality of tiles consisting of interconnected logic blocks, that are separated by intervening logic blocks. Each set of interconnected logic blocks is linked by an interconnect segment that is routed in a straight line through an interconnect layer over intervening logic blocks and is selectively connected to the logic block at each end through a connecting segment.
1. A high performance interconnect architecture for a field programmable gate array (FPGA), comprising a plurality of tiles, each tile comprising:
a plurality of logic blocks, each of which can be selectively connected to one or more logic blocks separated from it by one or more intervening logic blocks,
a plurality of interconnect segments, each of which connects two or more logic blocks and is routed in a straight line through an interconnect layer over the one or more intervening logic blocks, and
a connecting segment provided at a periphery within each logic block for selectively connecting the logic block with one of the interconnect segments.
2. The interconnect architecture as claimed in claim 1 , wherein the connecting segments are fabricated on a metallization layer different from a metallization layer used for the interconnect segments to facilitate crossing over intermittent interconnect segments.
3. The interconnect architecture as claimed in claim 1 , wherein the connecting segments are connected to the interconnect segments through vias.
4. The interconnect architecture as claimed in claim 1 , wherein two connecting segments connected to two interconnect segments are provided for each logic block.
5. The interconnect architecture as claimed in claim 1 , wherein:
the two connecting segments are located at specified locations of each logic block; and
the specified locations are shifted in adjacent ones of the logic blocks to emulate a staggering line topology using straight line interconnect segments.
6. The interconnect architecture as claimed in claim 4 , wherein each interconnect segment connects two logic blocks that are separated by at least two intervening logic blocks.
7. A method for providing a high performance interconnect architecture for a field programmable gate array (FPGA), comprising the steps of:
providing a plurality of tiles,
providing a plurality of logic blocks within each tile,
enabling selective connection between a plurality of the logic blocks separated from each other by one or more intervening logic blocks,
routing a plurality of interconnect segments, each interconnect segment routed between two connected logic blocks in a straight line through an interconnect layer over the one or more intervening logic blocks, and
providing a connecting segment at a periphery of each logic block for selective connection of the logic block with one of the interconnect segments.
8. The method as claimed in claim 7 , including fabricating said connecting segments on metallization layers that are different from metallization layers used for fabricating the interconnect segments to facilitate crossing over intermittent interconnecting segments.
9. The method as claimed in claim 7 , including linking of the connecting segments to the interconnect segments through vias.
10. The method as claimed in claim 7 , wherein two connecting segments connected to two interconnect segments are provided for each logic block.
11. The method as claimed in claim 10 , wherein:
the two connecting segments are located at specified locations of each logic block; and
the specified locations are shifted in adjacent ones of the logic blocks to emulate a staggering line topology using straight line interconnect segments.
12. The method as claimed in claim 10 , wherein each interconnect segment connects two logic blocks that are separated by at least two intervening logic blocks.
13. A field programmable gate array comprising a plurality of tiles, at least one of the tiles comprising:
a plurality of logic blocks, each logic block capable of being connected to one or more non-adjacent logic blocks;
a plurality of interconnect segments, each interconnect segment routed in a straight line over one or more intervening logic blocks between two connected logic blocks; and
at least one connecting segment, each connecting segment capable of connecting one of the logic blocks to one of the interconnect segments.
14. The field programmable gate array as claimed in claim 13 , further comprising a plurality of vias, each via connecting one connecting segment with one of the interconnect segments.
15. The field programmable gate array as claimed in claim 13 , wherein each logic block includes two connecting segments connected to two of the interconnect segments.
16. The field programmable gate array as claimed in claim 15 , wherein:
the two connecting segments are located at specified locations of each logic block; and
the specified locations are shifted in adjacent ones of the logic blocks to emulate a staggering line topology using straight line interconnect segments.
17. The field programmable gate array as claimed in claim 13 , wherein each connecting segment is disposed along a periphery of the logic block.
18. The field programmable gate array as claimed in claim 13 , wherein each interconnect segment connects two logic blocks that are separated by at least two intervening logic blocks.
19. The field programmable gate array as claimed in claim 13 , wherein each interconnect segment connects two logic blocks that are separated by three intervening logic blocks.
20. The field programmable gate array as claimed in claim 13 , wherein the plurality of tiles are arranged orthogonally in the field programmable gate array.