Embodiments of the present disclosure generally relate to integrated circuit packaging of at least one digital logic/processor layer and at least one memory layer, and in particular, to configurations of digital logic/processor and memory dice arranged in a stack for use in integrated circuit packages without having to modify the digital logic/processor to supply power and control signals to the at least one memory layer.
Memory and logic (e.g., processors and peripheral) dice, stacked and interconnected vertically, are gaining interest among memory and processor suppliers, and users to further increase memory bandwidth, reduce memory latency, reduce power required to move data and increase component integration density. Placing a stack of memory dice on top of a processor layer requires perforating the processor layer with through-silicon vias (TSV) to deliver signals and power to the memory dice located above the processor layer, which can be expensive to modify the processor layer (likely fabricated in a leading silicon processor node) and foundries might not yet be able to support or enable/offer TSVs in the leading silicon layer processor node, at least not for a while after introduction of the new silicon node.
In one example of the disclosure, an integrated circuit (IC) die stack includes at least one memory layer. A digital device. And at least one bypass chiplet, wherein the at least one memory layer is on a first side of the digital device and the at least one bypass chiplet, and power and signals pass through the at least one bypass chiplet to the at least one memory layer.
In one example of the disclosure, an integrated circuit (IC) die stack includes at least two memory layers, a digital device, and at least two bypass chiplets, wherein the at least two memory layers are on a first side of the digital device and the at least two bypass chiplets, and power and signals pass through the at least two bypass chiplets to respective ones of the at least two memory layers.
In one example of the disclosure, an integrated circuit (IC) die stack includes at least two memory layers, a digital device, and at least one bypass chiplet, wherein the at least two memory layers are on a first side of the digital device and the at least one bypass chiplet, and power and signals pass through the at least one bypass chiplet to the at least two memory layers.
So that the manner in which the above recited features of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to examples, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only typical examples of this invention and are therefore not to be considered limiting of its scope, for the invention may admit to other equally effective examples.
To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures, and a lower-case letter added where the elements are substantially the same. It is contemplated that elements of one embodiment may be beneficially incorporated in other embodiments.
Referring to
Referring to
According to the teachings of this disclosure, bypass chiplets are provided to avoid requiring integration of through-silicon vias (TSV) in a digital device, e.g., processor, for signals and power to a memory stack located above the digital device. Not having to add TSVs in an existing design high-performance digital device may significantly reduce and/or eliminate digital device modification costs, and time to manufacture new products consisting of three-dimensional stacked memory and digital device organizations. This also allows the usage of leading-edge digital device nodes that have not yet been enabled nor have available options for incorporating TSVs in the high-performance digital device components thereof. This allows the manufacture of advanced state of the art products using three-dimensional stacked memory and digital device configurations using existing and unmodified high-performance state of the art digital devices. A digital device may be, for example but is not limited to, one or any combination of a microcontroller, a microprocessor, a mixed signal processor, a central processing unit (CPU), a programmable logic array (PLA), an application specific integrated circuit (ASIC), a digital signal processor (DSP), a graphics processing unit (GPU), a field programmable gate array (FPGA), neural processing unit and tensor processing unit. The terms layer, device, die, silicon layer, silicon die, and silicon device may be used interchangeable herein.
Various features are described hereinafter with reference to the drawing figures. It should be noted that the drawing figures may or may not be drawn to scale and that the elements of similar structures or functions are represented by like reference numerals throughout the drawing figures. It should be noted that the drawing figures are only intended to facilitate the description of the features of the examples. They are not intended as an exhaustive description of the examples below or as a limitation on the scope of the claims. In addition, an illustrated example need not have all the aspects or advantages shown. An aspect or an advantage described in conjunction with a particular example is not necessarily limited to that example and can be practiced in any other examples even if not so illustrated, or if not so explicitly described. Referring now to the drawing figures, the details of examples are representative layouts schematically illustrated. Like elements in the drawing figures will be represented by like numbers, and similar elements will be represented by like numbers with a different lower-case letter suffix.
Referring to
The NoC 306 may be coupled directly to the digital device through the interposer 312. The interposer 312 may also provide interconnections between the at least one MC 308 and the NoC 306. A package substrate 314 may optionally be provided for connecting the active interposer 312 to solder lands on a printed circuit board (PCB) (not shown). The digital device 304 may also include additional logic for operation in combination with the active interposer 312. Of particular note is that the digital device 304 does not require any TSVs to be constructed through it (which might not even be feasible depending on the technology node). The bypass chiplets 310 may be manufactured in an older/more mature process node where TSVs are supported and/or are more cost effective to implement.
Referring to
The digital device 404 is adjacent to and on the same layer as the at least one bypass chiplet 410. The digital device 404 is offset/partially overlapped from the at least one memory layer 402 above it. The interposer 412 is below the digital device 404 and the at least one bypass chiplet 410, wherein the at least one bypass chiplet 410 is between the interposer 412 and the at least one memory layer 402. Furthermore, an additional “dummy” (passive) silicon structure 418 may be located between the digital device 404 and a cooling solution 420. The cooling solution is a thermal dissipation device with heat transfer enhancement structures, for example but is not limited to, a heat sink, a heat sink with fins, liquid cooling tubes, vapor chambers, heat pipes, cold plates and the like. The silicon structure 418 provides for more effective thermal conduction pathways that have better thermal conductivity than when compared to the thermal conductivity path through the at least one memory layer 402, located above the digital device/memory stack 400, as shown in
The at least one passive bypass chiplet 410 (not having active silicon/transistors) may provide electrical interconnections 416 (power and/or signals) (TSVs) to the at least one memory layer 402 from the at least one MC 408. Of particular note is that the digital device 404 does not require any electrical interconnections (TSVs) 416 to be constructed through it (which might not even be feasible depending on the technology node). The at least one bypass chiplet 410 may be manufactured in an older/more mature process node where TSVs are supported and/or are more cost effective to implement.
Referring to
The digital device 404 is adjacent to the at least one active bypass chiplet 510. The digital device 404 is adjacent to and on the same layer as the at least one active bypass chiplet 510. The digital device 404 is offset/partially overlapped from the at least one memory layer 402 above it. The interposer 512 is below the digital device 404 and the at least one active bypass chiplet 510, wherein the at least one bypass chiplet 510 is between the interposer 512 and the at least one memory layer 402. Furthermore, an additional “dummy” (passive) silicon structure 418 may be located between the digital device 404 and a cooling solution 420. The silicon structure 418 provides for more effective thermal conduction pathways that has more efficient thermal conductivity.
In other example embodiments, the digital device and the memory are not required to overlap (only that the memory is located at a higher level (layer) than the digital device in the sense that if this was a building, the memory would be on a higher floor or story). Referring to
The memory 402 on top of the bypass chiplet 610 may consist of a single layer (die) of memory, or may consist of multiple layers (dice) of memory (a memory stack). As shown and described for the example embodiments hereinabove, the interposer 612 at the bottom of each digital device/memory stack is depicted as an “active interposer, e.g., comprises active logic/circuits, and electrically connected through the bypass chiplet with electrical interconnections 616 (power and/or signals) (e.g., TSVs). However, alternative example embodiments may also use passive interposers 612, so long as the circuits that would have been the active interposer's logic blocks are implemented elsewhere, e.g., in the digital device 604 and/or in an active bypass chiplet 610.
Referring to
The memory 702, 703 on top of the bypass chiplets 710a and 710b may consist of a single layer (die) of memory, or may consist of multiple layers (dice) of memory (e.g., two or more memory stacks). As shown and described for the example embodiments hereinabove, the interposer 712 at the bottom of each memory stack is depicted as an “active interposer, e.g., comprises active logic/circuits, and electrically connected through the bypass chiplets 710a and 710b with electrical interconnections 716 (power and/or signals) (e.g., TSVs). However, alternative examples may also use passive interposers 712, so long as the circuits that would have been the active interposer's logic blocks are implemented elsewhere, e.g., in the digital device 604 and/or in an active bypass chiplet 710.
A three-dimensional digital device/memory stack, generally represented by the numeral 800, is similar to the aforementioned digital device/memory stack 700 with the exception that a common bypass chiplet 810 may be used in combination with electrical interconnections 816 (power and/or signals) to the memories 702 and 703. Thus, multiple memories could be placed on top of separate bypass chiplets 710a, 710b or a single bypass chiplet 810. The bypass chiplet 810 may be active or passive.
For the examples disclosed above, connections between the vias (TSVs) of the bypass chiplets, memory layer (dice), interposer, and/or digital device may be done with lower resistance metal bonding pads, e.g., hybrid-bonding, copper hybrid-bonding instead of using microbumps in the power delivery paths and may significantly lower resistance of the electrical connections. This solves a significant voltage drop problem associated with using microbumps for electrical power circuit connections. An added benefit is elimination of the die-to-die (D2D) layers between the silicon wafers, allowing direct metal-to-metal electrical connections (hybrid-bonding) between the die layers, thereby further reducing the resistance of connections there between. In addition, the layer stack thickness will be reduced and heat transfer improved there through.
In the examples disclosed hereinabove, the various semiconductor dice are illustrated or otherwise presumed to be “face down” (e.g., back end of line-BEOL metal layers facing toward the bottom of the stack, bulk silicon/backside facing upward toward the top of the stack). However, different examples may utilize one or more chiplets or other silicon components in “face up” orientations as well. It is contemplated and within the scope of this disclosure that the digital device may be located above the memory and closer to the cooling solution. The bypass chiplets disclosed herein may also be used in layer stack organizations with the memory located below the digital device.
The memory can consist of one or more memory technologies including but not limited to dynamic random-access memory (DRAM), static random-access memory (SRAM), serial shift registers, PCM, ferromagnetic RAM, spin-torque transfer RAM, spin-torque transfer STT-MRAM, eDRAM, flash RAM, phase-change memory, resistive RAM and the like. Wherein one or more memory layers may include additional functionality, such as compute or processing (e.g., processing in memory or PIM). Advantages of utilizing any one or more of the examples disclosed above avoids having to add connections (e.g., TSVs) through the digital device silicon which can increase manufacturing fabrication speed, reduce manufacturing costs and enable faster development of new products to market.
As will be appreciated by one skilled in the art and having the benefit of this disclosure, the embodiments disclosed herein may be embodied as a system, method, apparatus, or computer programmed product. Accordingly, aspects may take the form of an entirely hardware embodiment, an entirely software embodiment (including firmware, resident software, micro-code, etc.) or an embodiment combining software and hardware aspects that may all generally be referred to herein as a “circuit,” “module” or “system.” Furthermore, aspects may take the form of a computer program product embodied in one or more computer readable medium(s) having computer readable program code embodied thereon.
While the foregoing is directed to embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.
This Application is a continuation of U.S. Non-Provisional application Ser. No. 18/526,298, filed on Dec. 11, 2023 of which is incorporated herein by reference in its entirety.
| Number | Date | Country | |
|---|---|---|---|
| Parent | 18526298 | Dec 2023 | US |
| Child | 18540257 | US |