Alpha-particle-tolerant semiconductor die systems, devices, components and methods for optimizing clock rates and minimizing die size
View Patent ↗Systems and methods are disclosed herein for determining the placement of storage and non-storage cells or components, representing a semiconductor component in a design stage, on an integrated circuit die. In one embodiment, regions of a semiconductor die are analyzed with respect to the susceptibility of a region to be exposed to radiation and the distance between a storage component and a local clock buffer. The radiation, for instance, may be alpha particle radiation emitted from lead (Pb) isotopes in solder bumps formed on the integrated circuit die. The distance, spatial positioning and/or physical proximity of a selected local clock buffer and a storage component are preferably selected so that the skew between the storage component and the local clock buffer is about 30 picoseconds or less. Other maximum skews may be employed, however, such as about 100 picoseconds or less, about 90 picoseconds or less, about 80 picoseconds or less, about 70 picoseconds or less, about 60 picoseconds or less, about 50 picoseconds or less, about 40 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less and about 5 picoseconds or less.
1. A semiconductor die, comprising:
(a) a first region comprising at least one solder bump and at least one non-storage element, the first region having high alpha particle radiation;
(b) a second region comprising at least one local clock buffer operably connected to and spatially positioned in respect of at least one storage element such that the maximum skew therebetween does not exceed about 100 picoseconds, the second region having low alpha particle radiation;
wherein the first region contains no storage elements.
2. The semiconductor die of claim 1 , wherein the maximum skew is at least one of about 100 picoseconds or less, about 90 picoseconds or less, about 80 picoseconds or less, about 70 picoseconds or less, about 60 picoseconds or less, about 50 picoseconds or less, about 40 picoseconds or less, about 30 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less, and about 5 picoseconds or less.
3. The semiconductor die of claim 1 , wherein the maximum skew is at least one of about 30 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less, and about 5 picoseconds or less.
4. The semiconductor die of claim 1 , wherein the at least one storage element is at least one of a latch, a memory cell, a memory, a RAM, a DRAM, an SRAM, a CAM, a register, a register and flop, a register array, a register file, a register array, a synchronous register, a register with reset, a register with clear, a register with pre-set, an asynchronous register, a scannable register and a non-scannable register.
5. The semiconductor die of claim 1 , wherein the first region is characterized in generating high alpha particle radiation.
6. The semiconductor die of claim 1 , wherein the first region includes areas in proximity to sources of alpha particle radiation.
7. The semiconductor die of claim 1 , wherein the solder bumps generate alpha particle radiation.
8. The semiconductor die of claim 1 , wherein the solder bumps comprise a metal or metal alloy that emits alpha particles.
9. The semiconductor die of claim 8 , wherein the solder bumps comprise at least one of Lead (Pb), a U 235 isotope, a U 238 isotope and a Th 232 isotope.
10. The semiconductor die of claim 1 , further comprising a third region of intermediate alpha radiation, the third region being located between the first and second regions.
11. A method of designing a semiconductor die, comprising defining first and second regions in the die, placing in the first region at least one solder bump and at least one non-storage element, placing in the second region at least one local clock buffer operably connected to and spatially positioned in respect of at least one storage element such that the maximum skew therebetween does not exceed about 100 picoseconds, and placing no storage elements in the first region.
12. The method of claim 11 , wherein the maximum skew is at least one of about 100 picoseconds or less, about 90 picoseconds or less, about 80 picoseconds or less, about 70 picoseconds or less, about 60 picoseconds or less, about 50 picoseconds or less, about 40 picoseconds or less, about 30 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less, and about 5 picoseconds or less.
13. The method of claim 11 , wherein the maximum skew is at least one of about 30 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less, and about 5 picoseconds or less.
14. The method of claim 11 , further comprising placing a plurality of standard cells in the first and second regions of the semiconductor die, each standard cell being placed on the basis of its corresponding sensitivity to radiation.
15. The method of claim 11 , wherein the first region is exposed to a high intensity of radiation and the second region is exposed to a low intensity of radiation.
16. The method of claim 11 , wherein the first region includes areas in proximity to sources of radiation.
17. The method of claim 16 , wherein the sources of radiation are the solder bumps.
18. The method of claim 11 , wherein the solder bumps comprise a metal or metal alloy that emits alpha particles.
19. The method of claim 18 , wherein the solder bumps comprise at least one of Lead (Pb), a U 235 isotope, a U 238 isotope and a Th 232 isotope.
20. The method of claim 11 , further comprising defining a third region on the die, the third region being located between the first and second regions, each of the first, second and third regions representing different intensity levels of alpha particle radiation.
21. A method of making a semiconductor die, comprising providing first and second regions for the die, fabricating in the first region at least one solder bump and at least one non-storage element, fabricating in the second region at least one local clock buffer operably connected to and spatially positioned in respect of at least one storage element such that the maximum skew therebetween does not exceed about 100 picoseconds, and fabricating no storage elements in the first region.
22. The method of claim 21 , wherein the maximum skew is at least one of about 100 picoseconds or less, about 90 picoseconds or less, about 80 picoseconds or less, about 70 picoseconds or less, about 60 picoseconds or less, about 50 picoseconds or less, about 40 picoseconds or less, about 30 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less, and about 5 picoseconds or less.
23. The method of claim 21 , wherein the maximum skew is at least one of about 30 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less, and about 5 picoseconds or less.
24. The method of claim 21 , further comprising forming a plurality of standard cells in the first and second regions of the semiconductor die, each standard cell being placed and formed on the basis of its corresponding sensitivity to radiation.
25. The method of claim 21 , wherein the first region is exposed to a high intensity of radiation and the second region is exposed to a low intensity of radiation.
26. The method of claim 21 , wherein the first region includes areas in proximity to sources of radiation.
27. The method of claim 26 , wherein the sources of radiation are the solder bumps.
28. The method of claim 21 , wherein the solder bumps comprise a metal or metal alloy that emits alpha particles.
29. The method of claim 28 , wherein the solder bumps comprise at least one of Lead (Pb), a U 235 isotope, a U 238 isotope and a Th 232 isotope.
30. A computer system, comprising:
memory for storing a placement engine, and
a processor in communication with the memory, the processor configured to execute the placement engine stored in the memory;
wherein the placement engine comprises:
logic configured to determine the location of solder bumps on an integrated circuit die;
logic configured to determine a first set of regions including the areas on the integrated circuit die in proximity to the solder bumps;
logic configured to determine a second set of regions including the areas on the integrated circuit die outside the first set of regions;
logic configured to identify whether a standard cell is a storage element or a non-storage element;
logic configured to place the standard cell in one of the second set of regions and in proximity to a local clock buffer such that the maximum skew therebetween is about 100 picoseconds or less if the standard cell is identified to be a storage element; and
logic configured to place the standard cell anywhere on the integrated circuit die if the standard cell is identified to be a non-storage component.
31. The method of claim 30 , wherein the maximum skew is at least one of about 100 picoseconds or less, about 90 picoseconds or less, about 80 picoseconds or less, about 70 picoseconds or less, about 60 picoseconds or less, about 50 picoseconds or less, about 40 picoseconds or less, about 30 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less, and about 5 picoseconds or less.
32. The method of claim 30 , wherein the maximum skew is at least one of about 30 picoseconds or less, about 20 picoseconds or less, about 10 picoseconds or less, and about 5 picoseconds or less.
33. The computer system of claim 30 , wherein the integrated circuit die is a flip-chip die.
34. The computer system of claim 30 , wherein the first set of regions are more susceptible to alpha particle radiation than the second set of regions.
35. The computer system of claim 30 , wherein the logic configured to determine the first set of regions and the logic configured to determine the second set of regions determine the regions such that storage elements placed in the second region experience fewer soft errors.