Embodiments of the present disclosure generally relate to thermal management in integrated circuit die stacks and integrated circuit packaging of dynamic random-access memory (DRAM) dice, and in particular, to configurations of DRAM dice stacked in integrated circuit packages for improvements in thermal management thereof.
Dynamic random-access memory (DRAM), comprising stacked semiconductor dice, e.g., high bandwidth memory (HBM), in integrated circuit packages, face thermal challenges due to a potentially large number of layers of stacked semiconductor dice that present increasing thermal resistance for heat dissipation from the bottom of the dice stack to the cooling solution at the top of the dice stack. Typically, the majority of the power generation/consumption of a DRAM die stack comes from the bottom layer that may contain the physical electronics (PHY layer) for serializer/de-serializer (SerDes), address, read, write and refresh logic, data and address input receivers and output drivers, and DC voltage regulators. The PHY layer consists of much of the higher-activity/higher-power circuitry which, located at the bottom of the dice stack in the integrated circuit package, is furthest from a cooling solution (e.g., air cooled heat sink with fins, liquid cooling pipes, etc.), generally, at the top of the integrated circuit package. Therefore, the heat generated from the PHY layer die has to travel through the DRAM layer dice.
Excess heat in the DRAM circuitry may cause degradation in performance and require more frequent and higher refresh power to maintain the memory data contents stored therein. Currently, the main solutions dealing with high DRAM temperatures are to either throttle down in speed the memory/computer system in some way to reduce power consumption enough to reduce the amount of excess heat, and/or use of more expensive/exotic cooling solutions (e.g., liquid cooling, cold plate, immersion cooling), neither/both of which are undesirable. Another solution to ensure reliable operability of DRAM at higher temperatures is to increase the refresh rate which eventually leads to lower performance/fewer instructions per cycle (IPC) and increased power overhead.
In one example of the disclosure, an integrated circuit die stack is provided that includes digital device layer, an underlying layer, and a cooling solution. The underlying layer has a lower power consumption relative to digital device layer. The digital device layer disposed closer to the cooling solution.
In another example of the disclosure, a memory stack includes a digital device layer. A cooling solution on a first side of the digital device layer. And a plurality of memory layers on a second side of the digital device layer opposite the first side thereof.
In one example of the disclosure, a memory stack includes a plurality of memory layers stacked one on top of the other. A cooling solution located at one end of the stack of the plurality of memory layers. And a first digital device layer located between two of the plurality of memory layers.
In one example of the disclosure, a memory stack includes a plurality of memory layers stacked one on top of the other. At least two digital device layers attached and electrically interconnected to the plurality of memory layers. And a cooling solution located at one end of the memory stack of the plurality of memory layers and the at least two digital device layers.
In one example of the disclosure, an electronic system includes at least one memory stack. Each memory stack includes a digital device layer, a cooling solution on a first side of the digital device layer, and a plurality of memory layers on a second side of the digital device layer opposite to the first side thereof. And system on a chip (SoC) coupled to the at least one memory stack.
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
A memory interface, in particular, of the digital device layer 104 typically operates at very high speeds to achieve high data rate capacity transfer performance. Memory interface circuits operating at high data transfer rates, can consume a significant amount of power. The power consumption of the DRAM stack 100 generates heat (fire icon 108), which must then make its way up through the rest of the DRAM stack 100 to the cooling solution 106 where the heat can be removed. In an HBM for example, roughly 40 percent of the total power can be dissipated within the digital device layer 104. With an increasingly tall DRAM stack 100 (e.g., eventually 12-16 high), can lead to a sharp increase in temperature of the digital device layer 104 due to the distance and insulating properties of the DRAM layers 102 between the digital device layer 104 and the cooling solution 106. For DRAM circuits especially, this elevated temperature can impact performance (due to the need for more frequent refreshes), power consumption (again due to the need for more frequent refreshes) and reliability.
According to the teachings of this disclosure, a plurality of memory layers, e.g., DRAM, static random-access memory (SRAM), serial shift registers, eDRAM, Flash, phase-change memory, resistive RAM, ferromagnetic RAM, spin-torque transfer RAM and the like, and a digital device layer, e.g., a memory interface for address, command, and data of the plurality of memory layers may be configured in a three-dimensional memory stack. This memory stack may have the digital device layer and the plurality of memory layers inverted from the construction of the memory stack of the prior art. The inverted memory stack has a first surface (side) of the digital device layer located closest to the cooling solution and the plurality of memory layers located on a second surface (side) of the digital device layer opposite to the first surface (side) thereof. This inverted memory stack configuration substantially alleviates and/or mitigates the thermal challenges of efficiently removing heat from the digital device layer since now it is closest to the cooling solution and not being thermally insulated by the memory layers. It is contemplated and within the scope of this disclosure that one or more of the digital device layer(s), in addition to a memory interface, may comprise 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, tensor processing unit, and the like.
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 cooling solution 206 is adapted to receive and dissipate heat from the digital device layer 204 and the plurality of memory layers 202. The digital device layer 204, proximate to the cooling solution 206, comprises interface circuits to external IC package connections, e.g., memory interface signals (address, command, and data) to the memory layers 202, and may also include other circuits and connections for debug, test, control, etc., of the plurality of memory layers 202. Power delivery to the plurality of memory layers 202 and digital device layer 204 may be provided directly from the bottom of the memory stack 200, e.g., via a subset of connections at the bottom of the memory layer 202h.
The higher-power dissipation circuitry (digital device layer 204) may be placed proximate to the cooling solution 206 in the memory stack 200, instead of having the plurality of memory layers 202 there between. This significantly reduces the thermal resistance between the circuitry in the digital device layer 204 and the cooling solution 206, thereby enabling the plurality of memory layers 202 to maintain lower operating temperatures. The majority of heat is generated by the digital device layer 204 and has a much shorter distance to travel to the cooling solution 206, as represented by the fire arrow 208. The digital device layer 204 die typically is not thinned (or not thinned nearly as much as the other memory layers 202) since it does not need to provide for integration of through-silicon vias (TSV). This in turn improves spreading of heat laterally by reducing thermal resistance which is also a key attribute for reducing the severity of localized semiconductor die hotspots. There are a few considerations for this organization. First, the connections from the external signal paths (address/command/data) have to be routed to the digital device layer 204 through the plurality of memory layers 202, which may require more TSVs that in turn increases the die area of each of the plurality of memory layers 202. A large increase in die area overhead is not required because only enough additional areas for signal paths are required and not for all of the power delivery (power and ground) TSVs. The power delivery TSVs already exist in the prior art DRAM stack 100 (
Referring to
Referring to
Flexibility in the selection of the locations for the at least two digital device layers 204 enables optimizing the thermal implications for better distribute of heat generated by the at least two digital device layers 204 and may allow for a more even thermal profile of the memory stack 400. This provides for some tradeoffs between how far the at least two digital device layers 204 are from the cooling solution 206, at one end of the memory stack 400, and from the other end of the memory stack 400 where external interfacing of signal connections are located. Where signal integrity is critical, being closer to the external integrated circuit (IC) package signal connections may take precedence over optimal cooling performance of the memory stack 400. This may somewhat increase the distance that the heat generated by the at least two digital device layers 204 has to travel to the cooling solution 206, as represented by the fire arrows 408. Operating the at least two digital device layers 204 at a reduced voltage (lower power dissipation) may help offset the increased thermal resistance (poorer heat flow) caused by the at least two digital device layers 204 having some memory layers 202 located between them and the cooling solution 206.
Referring to
Referring to
It is contemplated and within the scope of this disclosure the memory layers 202 and digital device layer(s) may be electrically coupled together using micro-bump or micro-pillar stacking technologies. In addition, other types of memory technologies and/or other types of connection/stacking technologies, e.g., hybrid bonding, may be used. Similarly, TSVs may be coupled to external connections on the interposer 512.
Preliminary analysis demonstrates that moving the digital device layer 204 closer to the cooling solution 206 provides significant thermal benefits as shown in the table below.
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.