DLL-based clocking architecture with programmable delay at phase detector inputs

1. A delay-locked loop (DLL) on a chip, the DLL comprising:

a first programmable delay element configured to output a first clock;

a second programmable delay element configured to output a second clock; and

a phase detector including a first clock input and a second clock input, the first and second programmable delay elements further configured, in combination, to introduce a controllable skew between the first and second clocks, the DLL configured to input the first and second clocks to the first and second clock inputs of the phase detector, respectively, the DLL unable to compensate for the controllable skew introduced by the first and second programmable delay elements.

2. The DLL of claim 1 , wherein the chip includes at least one critical path and at least one non-critical path and wherein the controllable skew is configured to improve frequency of the chip by reducing latency of the at least one critical path and increasing latency of the at least one non-critical path.

3. The DLL of claim 1 , wherein respective programmable delays of the first and second programmable delay elements are configured to be programmed via control signals generated external to the DLL.

4. The DLL of claim 1 , wherein:

the first and second programmable delay elements are Joint Test Action Group (JTAG) enabled and include respective JTAG interfaces; and

respective programmable delays of the first and second programmable delay elements are configured to be programmed via the respective JTAG interfaces.

5. The DLL of claim 1 , wherein respective programmable delays of the first and second programmable delay elements are configured to be programmed via at least one Chip Select Register (CSR) of the chip.

6. The DLL of claim 1 , wherein the phase detector is configured to compare respective phases of the first and second clocks and to generate a phase difference based on the respective phases compared and wherein the DLL further comprises:

a fixed delay element;

a variable delay element;

a first clock distribution circuit interposed between the fixed delay element and the first programmable delay element;

a second clock distribution circuit interposed between the variable delay element and the second programmable delay element; and

a finite state machine (FSM) configured to generate a delay control based on the phase difference, a variable delay of the variable delay element configured via the delay control.

7. The DLL of claim 6 , wherein:

the DLL is coupled to a clock generator, the clock generator configured to output a root clock (RCLK) to the fixed and variable delay elements;

the fixed delay element is configured to output a first delayed RCLK to the first clock distribution circuit, the first clock distribution circuit configured to output a reference RCLK to the first programmable delay element; and

the variable delay element is configured to output a second delayed RCLK to the second clock distribution circuit as a function of the delay control, the second distribution circuit configured to output a controlled RCLK to the second programmable delay element.

8. The DLL of claim 7 , wherein the first and second programmable delay elements are further configured, in combination, to introduce the controllable skew to offset a systematic skew between the reference RCLK and the controlled RCLK, the systematic skew due to clock distribution differences of the first and second clock distribution circuits.

9. The DLL of claim 7 , wherein the fixed delay element is an input clock fixed delay element and wherein the DLL further comprises:

a first internal fixed delay element configured to delay the first clock and output the first clock delayed;

a second internal fixed delay element configured to delay the second clock and output the second clock delayed;

a falling phase detector (PDL) configured to generate a PDL output as a function of the first clock and the second clock delayed, wherein the PDL output is input to the FSM; and

a rising phase detector (PDR) configured to generate a PDR output as a function of the second clock and the first clock delayed, wherein the PDR output is input to the FSM.

10. The DLL of claim 1 , wherein the DLL further comprises a first clock distribution circuit and a second clock distribution circuit, wherein the controllable skew is configured to offset a systematic skew between respective clocks input to the first and second programmable delay elements, the systematic skew due to clock distribution differences of the first and second clock distribution circuits.

11. The DLL of claim 1 , wherein the DLL further comprises:

a first internal fixed delay element configured to delay the first clock and output the first clock delayed;

a second internal fixed delay element configured to delay the second clock and output the second clock delayed;

a PDL configured to generate a PDL output as a function of the first clock and the second clock delayed; and

a PDR configured to generate a PDR output as a function of the second clock and the first clock delayed.

12. A method for improving frequency of a chip, the method comprising:

outputting a first clock from a first programmable delay element of a delay-locked loop (DLL) on a chip to a first clock input of a phase detector of the DLL;

outputting a second clock from a second programmable delay element of the DLL to a second clock input of the phase detector; and

introducing a controllable skew between the first and second clocks by controlling respective programmable delays of the first and second programmable delay elements, in combination, the DLL unable to compensate for the controllable skew introduced by the first and second programmable delay elements.

13. The method of claim 12 , wherein the chip includes at least one critical path and at least one non-critical path and wherein the introducing reduces latency of the at least one critical path and increases latency of the at least one non-critical path to improve the frequency.

14. The method of claim 12 , further comprising configuring the respective programmable delays of the first and second programmable delay elements via control signals generated external to the DLL.

15. The method of claim 12 , wherein the first and second programmable delay elements are Joint Test Action Group (JTAG) enabled and include respective JTAG interfaces and wherein the method further comprises configuring the respective programmable delays of the first and second programmable delay elements via the respective JTAG interfaces.

16. The method of claim 12 , wherein the method further comprises configuring the respective programmable delays of the first and second programmable delay elements via at least one Chip Select Register (CSR) of the chip.

17. The method of claim 12 , further comprising:

comparing respective phases of the first and second clocks and generating a phase difference based on the comparing;

generating a delay control based on the phase difference; and

and controlling a variable delay via the delay control.

18. The method of claim 12 , wherein the DLL includes a first clock distribution circuit and a second clock distribution circuit and wherein the method further comprising:

offsetting a systematic skew between respective clocks input to the first and second programmable delay elements via the controllable skew, the systematic skew due to clock distribution differences of the first and second clock distribution circuits.

19. The method of claim 12 , wherein the DLL includes a falling phase detector (PDL) and a rising phase detector (PDR) and wherein the method further comprises:

delaying the first clock and outputting the first clock delayed;

delaying the second clock and outputting the second clock delayed;

generating a PDL output from the PDL as a function of the first clock and the second clock delayed; and

generating a PDR output from the PDR as a function of the second clock and the first clock delayed.

20. A delay-locked loop (DLL) on a chip, the DLL comprising:

a first programmable delay element configured to output a first clock;

a second programmable delay element configured to output a second clock; and

a phase detector including a first clock input and a second clock input, the first and second programmable delay elements further configured, in combination, to introduce a controllable skew between the first and second clocks, the DLL configured to input the first and second clocks to the first and second clock inputs of the phase detector, respectively, the controllable skew introduced to increase frequency of the chip, the DLL unable to compensate for the controllable skew introduced by the first and second programmable delay elements.