Onchip SRAM Vs Registers

Hi all, Can anyone tell me what is the threshold point(memory size) to opt for Onchip SRAM instead of FlipFlops , considering the Power&Area. In a module I have ~5Kbit of memory implemented as verilog registers. I would like to know the Power/Area savings if I switch to SRAM instead of Flipflops. I use .13u process. Any links/documents in this regard are highly welcome. Thanks Deepu John

Reply to
john.deepu
Loading thread data ...

John, your question seems to relate to ASIC design. This newsgroup deals with FPGAs where you select and then use a given architecture. That means the trade-offs are different. Modern FPGAs have SRAM blocks of various sizes, from 512 bits to 18 k bits and even larger. Xilinx offers the RAM option in the Look-up Tables (LUTs), which can even be used as shift registers. The designer's decision is to use available resources, not to design latches or flip-flops. This can lead to peculiar design optimizations when a certain resource is abundant and not all used, e.g. BlockRAMs.

5k bit fits very comfortably in one 18 kbit dual-ported BlockRAM, but would take an inordinate amount of LUTs, even at 16 bits per LUT. You decide. Peter Alfke, Xilinx
Reply to
Peter Alfke

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.