betaFeedback

WorksComputer Science

CU-3D: A Scalable Legality-Aware RTL-to-GDS Flow for 3D-IC Physical Design

Abstract

Face-to-face (F2F) hybrid bonds shorten interdie connections, but data, clock, and power-grid (PG) connections must be realized together. CU-3D is a RTL-to-GDS physical design flow based on commercial EDA tools. It derives role-compatible hybrid-bond (HB) sites from a 17-metal ASAP7-based technology model, complete HB footprints, and constructed PG geometry. To coordinate partitioning with placement flexibility, Scaled refines an inherited 2D placement, whereas Fresh places the receiving die first and uses its pin locations and clock loads to guide placement of the other die. Data and clock HB assignments are jointly optimized while retaining alternative data sites and reserving capacity for subsequent PG assignment. The selected HB sites and stacks are preserved during 3D clock-tree synthesis, supply-aware cell insertion, and routing. Across eleven designs, including a 133-macro processor, the flow achieves zero backend DRC violations and passes signal-connectivity checks. Nine paired designs admit lower-footprint, lower-power implementations; four improve setup WNS. On JPEG, early joint planning gains 87 ps and mobility-aware partitioning gains 346 ps. Path analysis explains why cell-delay savings can outweigh net-delay increases.

Three cases illustrating legality decisions and detection stages in a 3D-IC physical design flow

Figure 1. Legality is decided early and detected late. (a) A clear center has an obstructed HB footprint. (b) A DATA request blocks CLOCK access to site h1h_1 although h2h_2 is available. (c) A free row lacks reachable VSS. Filled/open markers show decision/detection stages.

Introduction

Face-to-face (F2F) hybrid bonding connects dies through their interconnect faces [1]. Submicron hybrid bonds (HBs) replace long planar wires [2], promising lower capacitance, shorter critical paths, and smaller footprints. A short vertical connection still needs feeder wires connecting cell pins to HB sites on both dies, power for the surrounding logic, and clocks for its registers. Partitioning and placement identify potential gains; a powered, clocked, and routed layout determines whether those gains survive physical implementation.

Prior flows advance 2D-engine reuse [3, 4, 5], terminal assignment [6, 7], clocks [8], and power delivery [9]. Their integration poses a cross-stage problem: a partitioner sees crossing nets, an allocator terminals, and a router conductor stacks. Optimizing these representations separately can consume resources needed later. A

This work is licensed under a Creative Commons Attribution 4.0 International License.

power grid (PG) can remove a selected landing, and clock construction can require sites already occupied by data bonds. Once placement and interface choices are fixed, ordinary routing has limited freedom to resolve the conflict. The challenge is to coordinate decisions before they become physical commitments.

Figure 1 illustrates three cases where locally feasible decisions lead to downstream conflicts. A clear grid center does not imply that the complete HB footprint fits. A DATA assignment can consume a CLOCK group’s only reachable site even when DATA has another choice. A buffer can fit a legal row yet lack a VSS connection. These are different manifestations of the same issue: the resource accepted by one stage is insufficient for the next. The interface is a shared, role-differentiated resource whose capacity is set by implemented geometry. Legality belongs to the whole decision chain.

CU-3D organizes Define–Decide–Deliver around three principles (Figure 2). Resource realism derives usable sites from complete footprints, macro exclusions, and actual PG landings. Freedom-matched commitment aligns partition capacity with placement mobility: inherited organization is refined where useful, while native placement can expose receiving-die demand before the second die commits. DATA alternatives remain available during CLOCK selection, with capacity reserved for PG. Decision preservation then protects accepted sites, stacks, and clock ownership

Figure 2. Define–Decide–Deliver: a shared F2F model defines resources, placement and allocation decide their use, and physical realization preserves the decisions. Fresh steps 1–6 follow § 7.1.

Three cases illustrating legality decisions and detection stages in a 3D-IC physical design flow

Figure 1 (PDF p. 1). Legality is decided early and detected late. (a) A clear center has an obstructed HB footprint. (b) A DATA request blocks CLOCK access to site h1h_1 although h2h_2 is available. (c) A free row lacks reachable VSS. Filled/open markers show decision/detection stages.

Introduction

Face-to-face (F2F) hybrid bonding connects dies through their interconnect faces [1]. Submicron hybrid bonds (HBs) replace long planar wires [2], promising lower capacitance, shorter critical paths, and smaller footprints. A short vertical connection still needs feeder wires connecting cell pins to HB sites on both dies, power for the surrounding logic, and clocks for its registers. Partitioning and placement identify potential gains; a powered, clocked, and routed layout determines whether those gains survive physical implementation.

Prior flows advance 2D-engine reuse [3, 4, 5], terminal assignment [6, 7], clocks [8], and power delivery [9]. Their integration poses a cross-stage problem: a partitioner sees crossing nets, an allocator terminals, and a router conductor stacks. Optimizing these representations separately can consume resources needed later. A

This work is licensed under a Creative Commons Attribution 4.0 International License.

power grid (PG) can remove a selected landing, and clock construction can require sites already occupied by data bonds. Once placement and interface choices are fixed, ordinary routing has limited freedom to resolve the conflict. The challenge is to coordinate decisions before they become physical commitments.

Figure 1 illustrates three cases where locally feasible decisions lead to downstream conflicts. A clear grid center does not imply that the complete HB footprint fits. A DATA assignment can consume a CLOCK group’s only reachable site even when DATA has another choice. A buffer can fit a legal row yet lack a VSS connection. These are different manifestations of the same issue: the resource accepted by one stage is insufficient for the next. The interface is a shared, role-differentiated resource whose capacity is set by implemented geometry. Legality belongs to the whole decision chain.

CU-3D organizes Define–Decide–Deliver around three principles (Figure 2). Resource realism derives usable sites from complete footprints, macro exclusions, and actual PG landings. Freedom-matched commitment aligns partition capacity with placement mobility: inherited organization is refined where useful, while native placement can expose receiving-die demand before the second die commits. DATA alternatives remain available during CLOCK selection, with capacity reserved for PG. Decision preservation then protects accepted sites, stacks, and clock ownership PG: constructed power grid; –: unspecified scope. MoL/LoL: memory-/logic-on-logic. DRC counts use each work’s backend checks.

through construction. Together, these principles connect early optimization to the resources that clocks, power, and routing will actually use.

Our contributions are:

  • Executable F2F resources: complete footprints and dual-die PG landings define role-compatible sites (§ 5).

  • Freedom-matched placement: mobility-aware capacity supports Scaled inheritance and receiving-die-first Fresh placement (§ 6).

  • Scalable joint allocation: DATA alternatives remain movable during CLOCK selection, followed by residual PG matching (§ 8).

  • Legality-preserving realization: receiver-first CTS and powered insertion deliver zero backend DRC and signal-connectivity errors across eleven designs, up to 133 macros (§ 9–§ 10).

Related Work

2D-engine reuse. Shrunk-2D [3], Compact-2D [4], and Snap-3D [5] develop tier-splitting surrogates; Cascade2D [13] and Pin-3D [14] address monolithic 3D physical synthesis. Macro-3D [10] addresses heterogeneous macro-on-logic F2F designs.

Assignment and electrical optimization. BTAssign [6], AnchorGrid [7], and Kim and Park [15] improve terminal assignment; differentiable partitioning [16] penalizes terminal-congested cuts. FlexiCTS [8] coordinates clock buffers and bonds; DPO-3D [9] and Zhu et al. [17] address power delivery. Distill3D [18], timing-driven placement [19], and DiffCCD [20] improve timing prediction and clock–data optimization.

Complete flows. Hetero-3D [11] couples mixed-node floorplanning and power delivery. Open3DFlow [21, 22], Open3DBench [12], and RosettaStone [23] support reproducible evaluation. Patnaik et al. [24] report DRC-clean RDL-based F2F layouts. Open3DBench’s logic-on-logic flow uses ideal clocks and reports residual DRC. CU-3D builds a shared inventory of HB sites, dual-die PG landings, and supply-reachable insertion regions from technology rules and the evolving layout. Constructed clocks, PG, and stacks undergo joint DRC and connectivity checks (Table 1).

MethodInterface representationClock constructionPGMacrosReported DRC
Shrunk-2D [3]Monolithic 3D / MIVs2D CTS, tier split–YesClean
Compact-2D [4]Combined metal stack2D CTS, tier split–Yes–
Snap-3D [5]Tier rows / F2F viasCombined CTS–Yes–
Macro-3D [10]Macro-on-logicMeasured clock tree–Yes–
Hetero-3D [11]Mixed-node F2FCTSBuiltYes–
BTAssign [6]Legal HB mesh––––
FlexiCTS [8]HBT allocationCPPR-aware 3D CTS–––
AnchorGrid [7]Pad arrayRoot-balanced––224 (CPU2)
Open3DBench [12]Vias / virtual HBsMoL: CTS; LoL: ideal–YesResidual (LoL)
DPO-3D [9]F2F power network–Built–Residual
CU-3D6/3-via HB stacksReceiver then ownerBuilt0–1330 (11 layouts)

Table 1. What representative 3D methods physically build and check.

Overview and Objective

Flow at a Glance

Synthesis supplies the netlist; a real-CTS 2D reference supplies timing and area growth. Partitioning assigns tiers, and floorplanning establishes macros, PG, and access reservations. Placement exposes pins for joint allocation and residual PG matching. Stacks, receiver/owner clocks, powered insertion, and per-tier routing realize the selected interfaces (Figure 2).

The shared rules become progressively more specific: partitioning budgets space, allocation tests complete footprints, and insertion checks supply access. Provisional assignments are refined using feedback before physical implementation. After construction, optimization may change feeders but must preserve the assigned HB sites and stack.

Definitions and Objective

The F2F model specifies layers, windows, grid, footprints, and PG/access rules. Roles are DATA (including control/I/O), CLOCK, and PG by supply. A role-compatible site fits the required footprint and access. Physical completion checks whether committed resources coexist (§ 9.4).

We seek tier assignments, placements, HB assignments, clocks, PG, and routes that satisfy physical completion and maximize extracted setup WNS. For a register path i→ji \to j,

Si→j=T+Lj−Li−Dij−tsu−U,Dij=Dijcell+Dijnet.(1)S_{i\to j} = T + L_j - L_i - D_{ij} - t_{\mathrm{su}} - U,\qquad D_{ij} = D_{ij}^{\mathrm{cell}} + D_{ij}^{\mathrm{net}}. \tag*{(1)}

Here Si→jS_{i\to j} is setup slack, TT the period, Li/LjL_i/L_j launch/capture clock latency, tsut_{\mathrm{su}} setup time, and UU uncertainty. Data delay DijD_{ij} comprises cell delay DijcellD_{ij}^{\mathrm{cell}} (including clock-to-output) and net delay DijnetD_{ij}^{\mathrm{net}}. WNS is minimum slack; TNS sums negative setup slacks over reported endpoints. Interconnect changes affect both components through load, slew, buffering, and sizing; clock phase also affects slack.

Executable F2F Resources

F2F Technology Model

CU-3D uses an ASAP7-based technology model [25, 26] to identify sites that fit complete conductor stacks and satisfy spacing, enclosure, and access rules. Two logic dies share one XY database with distinct layer identities and explicit inter-die timing nets. Twelve ordinary and five interface metals form the stack (Figure 3). TOP cell views map local M1–M6 to M12–M7 while retaining pin XY geometry. Layer order, routing direction, PG direction, and instance orientation remain distinct.

Figure 3. F2F model: 17 metals and complete HB footprints. DATA/CLOCK uses six vias; PG uses three. Each stack crosses one bond at 0.5-μ\mum nominal pitch.

DATA/CLOCK traverses six vias from M6 to M7; PG uses the central three between B_TM2 and T_TM2. Both TM2 layers carry horizontal interface PG rails. Ordinary BOTTOM routing uses M2–M6 (three horizontal, two vertical layers); TOP uses M7–M11 (three vertical, two horizontal), with cell-pin access on M1 and M12, respectively. Vertical routing on BOTTOM and horizontal routing on TOP therefore have fewer ordinary layers, motivating direction-specific congestion analysis.

Rail width and via enclosure determine PG clearance, so site extraction uses constructed metal rather than planned stripes. HB connections preserve cross-die net connectivity, while interconnect evaluation accounts for the HB conductor stack and feeder wires on both dies.

Role-Compatible Sites

Within the common die outline, grid H\mathcal{H} contains points ϕ+(kpx,lpy)\phi+ (k p_x,l p_y), where ϕ\phi is its XY origin, px,pyp_x,p_y are pitches, and k,lk,l are integer indices. At site hh, role ρ\rho requires Fρ(h)∩Oρ=∅F_{\rho}(h) \cap O_{\rho} = \varnothing: its complete footprint Fρ(h)F_{\rho}(h) must avoid the applicable obstacle set OρO_{\rho}, including the boundaries, both dies’ macro-access exclusions, and PG clearances. PG stripes can remove DATA/CLOCK eligibility while supplying power landings. For supply net n∈{VDD,VSS}n \in\{\mathrm{VDD},\mathrm{VSS}\},

Pn=H∩(LBn⊖EB)∩(LTn⊖ET).(2)\mathcal{P}_n = \mathcal{H} \cap(L^n_B \ominus E_B) \cap(L^n_T \ominus E_T). \tag*{(2)}

Here Pn\mathcal{P}_n is the eligible PG-site set; B/TB/T denote BOTTOM/TOP. The operator ⊖\ominus erodes same-net landing regions LBn,LTnL^n_B,L^n_T by enclosure footprints EB,ETE_B,E_T, leaving legal centers. Opposite-net clearance and complete-stack containment further restrict this intersection. Both enclosures must fit; a nearby rail alone is not a landing.

Capacity and Assignability

Let S\mathcal{S} be DATA/CLOCK-compatible sites and P=⋃nPn\mathcal{P}=\bigcup_n \mathcal{P}_n. Unique physical capacity and guarded allocation capacity are

Cphys=∣S∪P∣,Calloc=⌊(1−γ)Cphys⌋+g.(3)C_{\mathrm{phys}} = |\mathcal{S}\cup\mathcal{P}|,\qquad C_{\mathrm{alloc}} = \lfloor(1-\gamma)C_{\mathrm{phys}}\rfloor+ g. \tag*{(3)}

Here ∣⋅∣|\cdot| counts distinct sites, 0≤γ≤10 \le\gamma\le1 is the withheld fraction, and integer gg releases part of the reserved capacity, with 0≤g≤Cphys−⌊(1−γ)Cphys⌋0 \le g \le C_{\mathrm{phys}}-\lfloor(1-\gamma)C_{\mathrm{phys}}\rfloor. Thus the guard is deducted once after obstacle filtering. DATA/CLOCK demand must also fit S\mathcal{S}. Role allowances limit counts, not exclusive zones. For a request subset U′U', let N(U′)N(U') be the union of its eligible sites. If ∣N(U′)∣<∣U′∣|N(U')| < |U'|, no matching exists (see Figure 4c): adequate total capacity does not imply assignability.

Figure 4. Role-compatible resources. (a) Full footprints account for BOTTOM (B) and TOP (T) obstacles. (b) PG needs same-net landings on both dies. (c) Unique-site capacity differs from local assignability. (d) Sites h1h_1–h6h_6 serve DATA requests ss, CLOCK groups gg, and PG request pp: relocating s1s_1 admits g1g_1.

Freedom-Matched Placement

Capacity Follows Mobility

Capacity constraints follow the spatial freedom available to downstream placement. For total ordinary-cell area AbA_b in region bb and usable tier areas Cb,B,Cb,TC_{b,B},C_{b,T}, the area assigned to BOTTOM, Ab,BA_{b,B}, must satisfy

max⁡(0,Ab−Cb,T)≤Ab,B≤min⁡(Ab,Cb,B).(4)\max(0,A_b-C_{b,T}) \le A_{b,B} \le\min(A_b,C_{b,B}). \tag*{(4)}

Local bounds suit inherited neighborhoods. Tier-aggregated capacity lets relocatable cells cross old bins without cutting cohesive logic for local balance. Fixed macros, rows, and access exclusions still limit space. Area bounds do not replace legalization or access checks; equal area is only a placement preference. The placer must permit the migration assumed by its partition, or aggregate spare area can hide local crowding.

The partitioner ranks HB overflow, area overflow, non-CLOCK crossing demand, local balance, and weighted cost lexicographically. The final cost is

J=∑e:neHB>0(we+wet−wec+wHBneHB).(5)J = \sum_{e:n_e^{\mathrm{HB}}>0} \left(w_e + w_e^{t} - w_e^{c} + w_{\mathrm{HB}}n_e^{\mathrm{HB}}\right). \tag*{(5)}

For crossing net ee, wew_e is the base cut weight, wetw_e^{t} is the timing penalty, wecw_e^{c} a congestion discount, and wHBw_{\mathrm{HB}} the per-HB weight. Driver, direction, and class determine bond count neHBn_e^{\mathrm{HB}}. Count priorities precede JJ: a favorable cut cannot compensate for exceeded capacity.

Scaled: Inherit and Refine

Scaled contracts 2D coordinate seeds while retaining real cell and macro dimensions. Per-tier legalization and refinement using the complete netlist preserve useful neighborhoods. The risk is a causal chain: rigid local balance splits connected clusters, increases DATA crossings, consumes CLOCK-reachable sites, and concentrates feeders on the remaining routing tracks. Local physical capacity is necessary; a forced 50/50 split in every bin is not.

CU-3D interrupts this chain through post-CTS insertion headroom, soft balance within hard capacity, congestion weights, role allowances, and a single capacity guard. DATA alternatives and targeted expansion resolve local competition; PG reservations and tier windows preserve access. Fewer cuts help only where released sites match CLOCK reach and PG access.

Fixed-size macros impose a distinct limit on coordinate inheritance: compressed seeds can enter a macro, forcing legalization to reorganize its neighborhood. Old HB targets then become stale. Re-anchoring to current pins restores consistency; releasing inherited coordinates opens a different search space.

Fresh: Expose, Then Commit

Interface-guided Fresh releases inherited ordinary-cell coordinates within assigned tiers, allowing relocatable-capacity partitioning to retain cohesive logic. TOP goes first because its register pins and capacitances define receiving CLOCK demand; BOTTOM macros and I/O already supply fixed boundaries. Figure 5b follows six steps grouped by the information they expose.

Figure 5. Freedom-matched placement. (a) Scaled retains spatial organization while controlling unnecessary crossings. (b) Fresh reveals receiving-die demand, provisionally assigns interfaces, and lets BOTTOM respond before connected refinement. (c) Macros couple placement, interface-access, and supply-service domains.

Steps 1–2 place TOP with inactive-tier attraction isolated and PG access reserved, restore complete connectivity, and export current TOP pins/capacitances plus fixed BOTTOM boundaries. Steps 3–4 jointly plan provisional DATA/CLOCK sites with retained DATA alternatives and compatible PG capacity, then place BOTTOM against these anchors while preserving TOP.

Steps 5–6 refine the connected design BOTTOM then TOP within ownership/protection boundaries. Both feeders are reevaluated; local PG, sites, pins, and loads are refreshed before final allocation. Thus the early plan guides BOTTOM, whereas final choices use both placed dies and actual landings. Costs combine pin spans, criticality, and migration. Because moving an HB can shorten one feeder and lengthen the other, connected timing judges the combined response.

Macro Feasibility Domains

Macros retain dimensions, orientation, pins, obstructions, supply rails, and tier identity. Active BLOCK and inactive COVER views (LEF macro classes) separate placement ownership while preserving geometry. Same-die bodies and rows bound cell placement; two-die footprints bound interface access; both supplies plus signal escape bound new-cell service. Legal cross-die projection overlap can therefore block an HB at the same XY point.

Cell capacity subtracts same-tier exclusion unions; HB capacity filters both dies. Rigid per-tier macro translation can reduce projected exclusions while preserving macro organization. Connected logic and HB targets must follow the moved pins. Shorter matched-pin distances nevertheless accompanied pre-CTS WNS losses of 12.4/27.9 ps on BOTTOM/TOP: proximity omits fanout, load, and collateral nets. Refinement must preserve electrical context.

Joint Interface Allocation

Shared Occupancy

DATA and CLOCK share site occupancy while preserving compatible capacity for residual PG. A CLOCK alternative a=(d,s,h,Ga)a=(d,s,h,G_a) identifies clock domain dd, source ss, HB site hh, and sink set GaG_a. Current pins/caps screen fanout, diameter, reach, latency/skew, and estimated load

C^a=∑i∈GaCi+cw∑i∈Ga∥pi−h∥1.\widehat{C}_a = \sum_{i\in G_a} C_i + c_w \sum_{i\in G_a} \lVert p_i-h\rVert_1.

Here Ci,piC_i,p_i are sink ii’s capacitance and XY location, cwc_w is wire capacitance per unit length, and ∥⋅∥1\lVert\cdot\rVert_1 denotes Manhattan distance. Buffered-tree candidates additionally account for buffer input loads and branch lengths. DATA request uu retains its legal baseline and permitted relocation sites HuH_u. Binary xax_a selects CLOCK alternative aa; binary yu,hy_{u,h} assigns DATA request uu to hh. Writing h(a)h(a) for aa’s site and taking yu,h=0y_{u,h}=0 outside HuH_u, we require

∑a:i∈Gaxa=1∀i,(6)\sum_{a:i\in G_a} x_a = 1 \quad\forall i, \tag*{(6)}
∑h∈Huyu,h=1∀u,(7)\sum_{h\in H_u} y_{u,h} = 1 \quad\forall u, \tag*{(7)}
∑a:h(a)=hxa+∑uyu,h≤1∀h.(8)\sum_{a:h(a)=h} x_a + \sum_u y_{u,h} \le1 \quad\forall h. \tag*{(8)}

Equation 6–8 cover each clock sink ii once, assign each DATA request uu once, and allow one occupant per physical site hh. Exact mode applies clock-count and regional limits, then prioritizes count deviation, CLOCK timing, critical DATA, ordinary DATA, and control/I/O costs.

Retaining DATA Alternatives

Proposition. Feasible assignment sets Ffixed\mathcal{F}_{\mathrm{fixed}} (fixed DATA) and Fjoint\mathcal{F}_{\mathrm{joint}} (movable DATA), under identical CLOCK choices and constraints, satisfy Ffixed⊆Fjoint\mathcal{F}_{\mathrm{fixed}}\subseteq\mathcal{F}_{\mathrm{joint}} if every legal DATA baseline is retained. Setting yu,baseline(u)=1y_{u,\mathrm{baseline}(u)}=1, where baseline(u)\mathrm{baseline}(u) is uu’s retained site, recovers each fixed solution. Solver mode determines how well the enlarged set is searched.

The converse fails in Figure 4d: moving s1s_1 to h2h_2 frees g1g_1’s sole site h1h_1, lets g2g_2 use h3h_3, keeps s2s_2 at h4h_4, and retains h6h_6 for PG. No request

Algorithm 1 Joint allocation with residual PG

Input: Compatible sites, current pins/caps, demand, limits

  1. Build sparse CLOCK/DATA choices; retain legal baselines

  2. Select size-adaptive mode; solve coverage and occupancy

  3. In spatial mode, expand deficient neighborhoods and resolve

  4. If incomplete, return failure category and witness

  5. Enforce guard/count limits; match compatible residual PG

  6. return Accepted plan or failed condition

is dropped or site added. Retaining legal DATA alternatives allows assignment to accommodate CLOCK groups with fewer options.

Adaptive Solving and Residual PG

Algorithm 1 stores alternatives and sink incidences sparsely; model size selects exact, bounded, or sparse/grouped solving. Spatial matching expands Hall-deficient neighborhoods. Missing candidates, deficient matching, incomplete coverage, infeasibility, and timeout are distinguished: a local lack of alternatives needs candidate expansion, not more solver time.

DATA/CLOCK planning reserves a PG allowance without fixing its final landings. Residual PG matches same-supply requests to unoccupied compatible sites within regional/distance limits. Let NDN_D, NCN_C, NPN_P count assigned DATA/CLOCK/PG interfaces and NP,min⁡N_{P,\min} be the required PG count. Acceptance enforces ND+NC+NP≤CallocN_D+N_C+N_P \le C_{\mathrm{alloc}} and NP≥NP,min⁡N_P \ge N_{P,\min} plus actual matching. Counts alone are insufficient: only reachable same-supply sites can realize the PG allowance.

Legality-Preserving Realization

Interface Invariants

Physical realization preserves each accepted site’s identity, orientation, pins, and conductor stack. DATA/CLOCK interfaces become SRC/Y, a dedicated interdie net nHBn_{\mathrm{HB}}, and DST/A; PG stacks join same-supply landings. Routing refines authorized feeders around this explicit connection (Figure 6a).

Figure 6. Decision preservation: (a) interface identity; (b) measured receiver feedback and owner CTS; (c) dual-supply insertion; (d) routing and checks on one assembled state.

Receiver-First CTS and Skew

TOP trees start at DST/Y. Group gg’s measured receiver latency ℓg\ell_g, skew σg\sigma_g, load, and geometry feed a bounded replan before BOTTOM owner CTS. The owner targets L(src→hg)+ℓg=L∗L(\mathrm{src} \to h_g) + \ell_g = L^* and L(src→s)=L∗L(\mathrm{src} \to s) = L^* for local sink ss. Here hgh_g is the HB root, L(src→hg)L(\mathrm{src} \to h_g) the source-to-root latency, and L∗L^* the common end-to-end latency target. Slower receivers leave smaller owner-side budgets; faster receivers allow longer owner branches. Balancing only HB-root arrivals ignores these unequal downstream delays. Receiver-first construction exposes them while the owner can still adapt.

Let δLi\delta L_i and δLj\delta L_j be launch and capture latency edits. With other terms fixed, the setup/hold slack changes are

ΔSi→jsetup=δLj−δLi,ΔSi→jhold=δLi−δLj.(9)\Delta S_{i\to j}^{\mathrm{setup}} = \delta L_j - \delta L_i,\qquad \Delta S_{i\to j}^{\mathrm{hold}} = \delta L_i - \delta L_j. \tag*{(9)}

In register chain i→j→ki \to j \to k, delaying jj helps incoming setup but hurts outgoing setup; a shared branch affects every descendant. Physical sizing also changes upstream load and slew. Skew optimization must therefore evaluate all affected incoming and outgoing paths, including hold, in the complete clock graph rather than optimize each receiver independently.

Supply-Reachable Insertion

For placement row rr, let R(r)\mathcal{R}(r) be legal cell-origin positions and IVDD(r)I_{\mathrm{VDD}}(r), IVSS(r)I_{\mathrm{VSS}}(r) the positions with reachable supply connections. The insertion service domain is D(r)=R(r)∩IVDD(r)∩IVSS(r)\mathcal{D}(r)=\mathcal{R}(r)\cap I_{\mathrm{VDD}}(r)\cap I_{\mathrm{VSS}}(r), with signal escape additionally required. A legal origin may reach VDD but not VSS (Figure 1c). Checking both supplies prevents a timing repair from creating an unpowered cell. Access reservations protect buffers before insertion; post-insertion checks verify both connections (Figure 6c).

Routing and Completion

Routing obeys tier windows around accepted stacks and PG. Physical completion requires, on one assembled state, zero backend DRC and declared signal/PG opens, intact HB stacks, covered clock sinks, and independent database readback. Timing coverage and electrical qualification accompany these checks; GDS release additionally binds stream maps, cell GDS/CDL, and matched LVS. DRC tests whether the chain’s decisions coexist in the final geometry.

Experimental Evaluation

Setup

Synopsys Design Compiler and Cadence Innovus implement the 2D/3D designs on the ASAP7-based F2F model at TT, 0.70 V, 25 °C. Nominal lengths and areas divide the 4× model geometry by 4 and 16; nominal HB pitch is 0.5 μm\mu\mathrm{m}. Each design uses one synthesized netlist and a real-CTS 2D reference. Both 2D and 3D results follow multiple rounds of placement optimization, CCOpt, post-CTS setup optimization, timing/SI-driven routing, and post-route setup

optimization, with high-effort congestion and path-group settings, then extraction.

Scaled reuses the 2D placement, while Fresh rebuilds the tier placements. We evaluate interface-guided ordering on JPEG. Backend DRC and connectivity checks are performed in Innovus on the assembled two-die layouts. Footprint is reported per die, power is reported by the tool using recorded activity, and wirelength is the total routed length. Table 2 reports nominal wirelengths in mm, while Table 3 uses model lengths.

DesignRouteWNS (ns)TNS (ns)PP (mW)AA (μm2\mu\mathrm{m}^2)WL (mm)Cell (%)Net (%)
Rocket2D-0.428-321.40042.4135,971.3204.5664.036.0
Scaled-0.410-455.57337.1123,843.0203.7964.135.9
AES2D-0.003-0.0056.9111,672.547.3186.313.7
Scaled0.0000.0006.5731,060.349.1687.512.5
Mini2D-0.290-38.10912.1542,218.962.3575.424.6
Scaled-0.268-51.34010.2941,404.555.5076.923.1
s359322D0.0230.0006.7831,315.517.0778.221.8
Fresh0.0510.0005.851797.619.6961.438.6
Ariane1332D-2.877-10,388.600233.44571,074.41,632.5138.161.9
Fresh-2.876-18,240.000215.13142,640.61,489.8337.162.9

Table 2. 2D/3D PPA and top-20 path composition. PP: tool-reported power; AA: per-die footprint; WL: wirelength. Each row uses one implementation. Cell/net shares pool its 20 lowest-slack data paths; cell delay includes clock-to-output.

VariantWNSTNSWLPP
Bottom-first-1.580-541.766959.252.90
Top-first, DATA only-0.551-212.371723.745.66
Early joint planning-0.464-167.780727.645.63
+ Movable-capacity partition-0.118-54.046573.941.81
Timing-weighted-0.175-139.920606.141.73
Fixed-sink refinement-0.132-140.240578.041.93
Independent initializations-0.219-203.370628.843.83

Table 3. JPEG variants. WNS/TNS: ns; WL: 10310^{3} model μ\mum; raw PP: mW. All report DRC/connectivity 0/0.

Completed Layouts and PPA

Across eleven designs, the post-route layouts exhibit zero backend DRC violations and pass connectivity checks (Figure 7). Table 1 compares implementation scope. Table 2 reports PPA for nine paired CU-3D and 2D implementations. Rocket/Mini Scaled gain 18/22 ps, AES Scaled reaches zero, and s35932 Fresh gains 28 ps. Rocket/Mini trade better WNS for worse TNS. All nine paired designs admit smaller-footprint, lower-power implementations; four improve best-route WNS.

Realized layouts across macro scales showing JPEG, RISC-V32i, Swerv, and Ariane133 designs in BOTTOM and TOP rows, with standard cells, macros, PG routes, DATA, CLOCK, VDD, and VSS

Figure 7. Realized layouts across macro scales. BOTTOM/TOP rows share one legend. All four have zero backend DRC and pass signal connectivity; Swerv and Ariane also pass PG connectivity.

WNS measures the worst endpoint; TNS captures the aggregate setup deficit, distinguishing bottleneck gains from gains across violating endpoints. Mini reduces routed length by 11.0%, but AES/s35932 improve WNS with 3.9%/15.4% more wire. Buffering, clock distribution, and rerouting can increase total length while improving critical paths. Table 2 ties all metrics to the same implementation.

Where the Timing Gain Lands

The 20 lowest-slack paths separate cell delay (including clock-to-output) from inter-instance net delay, excluding clock latency and input arrival. Shares use summed delays, so a larger cell share need not imply larger absolute cell delay. Across nine paired 2D designs, cells contribute 51.6–95.7% of worst-path data delay and 51.3–91.5% pooled over the top 20. Mini/AES reduce mean top-20 data delay by 56.2/19.7 ps; Ariane remains net-dominated in 2D/3D (61.9%/62.9%).

Endpoint-matched analysis aligns pins, edges, type, view, and group, selected by 2D criticality. Nine of ten representative comparisons reduce cell delay by 8.6–71.1 ps; seven increase net delay. Mini/AES save 71.1/63.8 ps in cells despite net increases of 9.2/13.5 ps. Load, slew, sizing, and changed logic chains matter beyond total wirelength. Shorter feeders can reduce driver load and improve downstream slew, benefiting cell delay as well as net delay. Additional buffering likewise changes both components.

Independent cohorts identify current bottlenecks; matched endpoints track the same connection. Rocket’s matched data delay falls 28.9 ps although its top-20 mean grows as input–register paths replace input–output paths. The former s35932 worst path drops to rank 82, illustrating critical-path migration.

Order and Mobility on JPEG

Bottom-first, top-first DATA-only, and early DATA/CLOCK planning reach −1.580, −0.551, and −0.464-ns WNS (Table 3). Early planning gains 87 ps despite 0.53% more wirelength: exposing competition improves more than geometry.

Mobility-aware low-HB partitioning reaches −0.118 ns: WNS improves 346 ps, TNS 113.734 ns, wirelength 21.12%, and raw power 8.38%. DATA (excluding control/I/O) falls 6211 to 561; CLOCK rises 293 to 471; PG is unchanged. Two 50-model-μ\mum samplings reduce peak occupancy from 313/325 to 129/129. Fewer cuts release choices without dictating clock demand.

Fixed-sink refinement and independent initialization lose 14 and 101 ps, respectively. Exposing demand and allowing placement to respond are complementary: an early HB plan needs downstream mobility. Subsequent refinement reaches −0.101 ns; 18/20 worst low-HB paths stay within one die. Once avoidable cuts are removed, cell/load and clock response become the next targets.

Congestion and Macro Scaling

Table 4 compares tier-isolated routing replay. Fresh issues 5.9–14.5× fewer non-CLOCK requests and lowers TOP horizontal overflow in all six designs. Horizontal dominates at all twelve TOP records; vertical is at least horizontal in eleven BOTTOM records, consistent with unequal directional layer counts and PG occupancy. The sum of H/V overflows falls 53 to 26 for JPEG and 178 to 3 for RISC-V32i, but Rocket BOTTOM vertical overflow rises 50 to 83: released capacity can redistribute pressure between dies. The relevant resource is therefore not just a bond count, but reachable sites together with feeder capacity in each tier’s scarce routing direction.

DesignRequests S/FBOTTOM S →\rightarrow FTOP S →\rightarrow F
Rocket5,532/4687/50 →\rightarrow 4/83150/28 →\rightarrow 62/11
JPEG6,230/5780/1 →\rightarrow 0/647/5 →\rightarrow 20/0
RISC-V32i2,257/35495/62 →\rightarrow 0/019/2 →\rightarrow 3/0
AES1,593/2691/10 →\rightarrow 0/658/2 →\rightarrow 30/0
Dynamic1,998/3100/3 →\rightarrow 0/928/2 →\rightarrow 19/4
Mini1,926/1330/2 →\rightarrow 0/530/1 →\rightarrow 28/0

Table 4. Scaled (S) to Fresh (F): non-CLOCK interface requests and tier-isolated horizontal/vertical (H/V) routing overflow. Same-stage global-route replay; entries give H/V overflow.

Figure 7 spans 0–133 macros. Swerv has 3,544 DATA/CLOCK bundles, 6,302 PG bonds, and 56 macro PG terminals; Ariane integrates 4,379 bundles and 7,218 PG bonds. Both pass signal/PG connectivity with zero backend DRC. Swerv’s tall macro blocks leave narrow channels, whereas Ariane’s repeated macro banks distribute access boundaries across the footprint. These layouts test the same interface rules under distinct constraints on cell placement, bond access, and power delivery.

Realized Structures and Scale

Figure 8 contrasts direct five-DFF and buffered seven-DFF receivers. In panel (a), 27/96 denotes straight wire sections visible in the window out of 96 in the complete BOTTOM owner net. Panel (b) uses two parallel buffers, one per sink path: allocation selects the root, whereas receiver CTS realizes the fanout. Both receiver structures use the same interface and legal-site model. Of 144 sites, 56 support DATA/CLOCK, 48 PG, and 40 are blocked; 33 are occupied. Capability differs from occupancy. Sparse allocation spans 1,661 variables (gcd) to 3,043,683 (JPEG Scaled), versus 528,181 for JPEG Fresh. These are assignment variables, not physical bonds: candidate sites and clock groupings represent alternative uses of the same resources. Retained alternatives let assignment resolve competition without changing the interface lattice.

Figure 8. Realized JPEG structures. (a,c,d) Actual clock routes, PG, and HB sites in one 6-by-6-μm\mu\mathrm{m} window; XY is measured and Z illustrates layer order. (b) A second receiver’s connection topology: two parallel buffers drive seven DFFs, one buffer per path. Open/filled site markers denote eligibility/occupancy.

Conclusion

CU-3D makes legality a property of the decision chain: resource realism ties capacity to geometry, freedom-matched commitment exposes demand before fixing choices, and decision preservation carries interfaces through powered clock construction and routing. Eleven designs, including 28- and 133-macro processors, demonstrate zero backend DRC and passing signal connectivity. JPEG links retained alternatives and placement mobility to timing gains; matched paths show cell savings outweighing net-delay increases.

Cell-dominated logic calls for load- and slew-aware reorganization, whereas Ariane’s net-dominated paths motivate coordinated costs for both feeders. Retained sites make alternatives physically available; completed timing determines which alternatives are useful. Thus neither fewer bonds nor shorter total wirelength is a universal 3D objective. Future work will feed matched-path cell, feeder, and clock-phase costs into partitioning, HB selection, and macroneighborhood optimization, and extend the common physical basis to multicorner timing, power integrity, and thermal constraints.

Acknowledgments

This work is partially supported by The Research Grants Council of Hong Kong SAR (No. CUHK14211824 and T46-415/25-R).

References

  1. [1]M.-F. Chen, F.-C. Chen, W.-C. Chiou, and D. C. H. Yu, “System on integrated chips (SoIC) for 3D heterogeneous integration,” in Proceedings of ECTC. Piscataway, NJ, USA: IEEE, 2019, pp. 594–599.DOI
  2. [2]S. A. Chew, B. Zhang, K. Vanstreels, E. Chery, J. De Messemaeker, L. Witters, K. Van Sever, S. Iacovo, S. Dewilde, M. Stucchi, J. De Vos, G. Beyer, A. Miller, and E. Beyne, “The challenges and solutions of Cu/SiCN wafer-to-wafer hybrid bonding scaling down to 400nm pitch,” in Proceedings of IEDM. Piscataway, NJ, USA: IEEE, 2023, pp. 1–4.
  3. [3]S. Panth, K. Samadi, Y. Du, and S. K. Lim, “Shrunk-2-D: A physical design methodology to build commercial-quality monolithic 3-D ICs,” IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, vol. 36, no. 10, pp. 1716–1724, 2017.DOI
  4. [4]B. W. Ku, K. Chang, and S. K. Lim, “Compact-2D: A physical design methodology to build commercial-quality face-to-face-bonded 3D ICs,” in Proceedings of ISPD. New York, NY, USA: Association for Computing Machinery, 2018, pp. 90–97.DOI
  5. [5]P. Vanna-Iampikul, C. Shao, Y.-C. Lu, S. Pentapati, and S. K. Lim, “Snap-3D: A constrained placement-driven physical design methodology for face-to-face-bonded 3D ICs,” in Proceedings of ISPD. New York, NY, USA: Association for Computing Machinery, 2021, pp. 39–46.
  6. [6]S. Liu, J. Jiang, Z. He, Z. Wang, Y. Lin, B. Yu, and M. Wong, “Routing-aware legal hybrid bonding terminal assignment for 3D face-to-face stacked ICs,” in Proceedings of ISPD. New York, NY, USA: Association for Computing Machinery, 2024, pp. 75–82.
  7. [7]P. Vanna-iampikul, J. Yoon, C. Park, G. Yeap, and S. K. Lim, “Placement-aware 3D net-to-pad assignment for array-style hybrid bonding 3D ICs,” in Proceedings of ISPD. New York, NY, USA: Association for Computing Machinery, 2025, pp. 200–208.
  8. [8]S. Li, L. Jin, S. Liang, Z. Zhuang, R. Chen, B. Yu, and T.-Y. Ho, “FlexiCTS: CPPR-aware 3D clock tree synthesis for face-to-face bonded ICs,” in Proceedings of DAC. New York, NY, USA: Association for Computing Machinery, 2026.DOI
  9. [9]Z. Zhuang, Z. Yang, Y. Zhao, J. Hu, B. Yu, S. K. Lim, and T.-Y. Ho, “DPO-3D: Differentiable power delivery network optimization via flexible modeling for routability and IR-drop tradeoff in face-to-face 3D ICs,” in Proceedings of ASP-DAC. Piscataway, NJ, USA: IEEE, 2026, pp. 1273–1279.DOI
  10. [10]L. Bamberger, A. García-Ortiz, L. Zhu, S. Pentapati, D. E. Shim, and S. K. Lim, “Macro-3D: A physical design methodology for face-to-face-stacked heterogeneous 3D ICs,” in Proceedings of DATE. Piscataway, NJ, USA: IEEE, 2020, pp. 37–42.
  11. [11]L. Zhu, J. Hu, G. Murali, and S. K. Lim, “Hetero-3D: Maximizing performance and power delivery benefits of heterogeneous 3D ICs,” in Proceedings of ISLPED. New York, NY, USA: Association for Computing Machinery, 2024, pp. 1–6.DOI
  12. [12]Y. Shi, C. Gao, W. Ren, P. Xie, S. Xu, K. Xu, M. Yuan, C. Qian, and Z.-H. Zhou, “Open3DBench: Open-source benchmark for 3D-IC backend implementation and PPA evaluation,” 2026, version 2, April 5, 2026. [Online]. Available: https://arxiv.org/abs/2503.12946v2
  13. [13]K. Chang, S. Sinha, B. Cline, R. Southerland, M. Doherty, G. Yeric, and S. K. Lim, “Cascade2D: A design-aware partitioning approach to monolithic 3D IC with 2D commercial tools,” in Proceedings of ICCAD. New York, NY, USA: Association for Computing Machinery, 2016, pp. 1–8.DOI
  14. [14]S. S. K. Pentapati, K. Chang, V. Gerousis, R. Sengupta, and S. K. Lim, “Pin-3D: A physical synthesis and post-layout optimization flow for heterogeneous monolithic 3D ICs,” in Proceedings of ICCAD. New York, NY, USA: Association for Computing Machinery, 2020.
  15. [15]G. Kim and H. Park, “A unified design flow for homogeneous and heterogeneous 3D integration with fine-pitch hybrid bonding,” in Proceedings of ICCAD. Piscataway, NJ, USA: IEEE, 2025, pp. 1–9.
  16. [16]L. Jiang, X. Shi, R. Xiao, C. Yan, R. Jiang, Z. Wang, and K. Zhu, “Differentiable partitioning with dual-max 3D wirelength model and hybrid bonding terminal awareness for optimized 3D placement,” arXiv:2609.06408v2, 2026. [Online]. Available: https://arxiv.org/abs/2609.06408v2
  17. [17]L. Zhu, C. Jo, and S. K. Lim, “Power delivery solutions and PPA impacts in micro-bump and hybrid-bonding 3D ICs,” IEEE Transactions on Components, Packaging and Manufacturing Technology, vol. 12, no. 12, pp. 1969–1982, 2022.DOI
  18. [18]M. Wang, L. Zou, G. Li, F. Gu, Y. Lu, T. Yang, B. Han, J. Mu, H. Li, S. K. Lim, and B. Yu, “Distill3D: Cross-stage knowledge distillation for early-stage 3D IC timing prediction,” in Proceedings of ICCAD, 2026, to appear. [Online]. Available: https://lanchengzhou.github.io/
  19. [19]Y. Ji, Y. Shi, T. Hou, Y. Zhao, C. Zhao, P. Liao, Z. Guo, C. Qian, Y. Lin, and B. Yu, “Integrated timing-driven placement for hybrid-bonding-based face-to-face 3D ICs,” in Proceedings of DAC. New York, NY, USA: Association for Computing Machinery, 2026.
  20. [20]Y. Ji, Y. Lu, Z. Zhang, Z. Guo, Y. Lin, and B. Yu, “DiffCCD: Differentiable concurrent clock and data optimization,” in Proceedings of ICCAD. Piscataway, NJ, USA: IEEE, 2025, pp. 1–9.
  21. [21]Y. Zhu, D. Feng, Z. Luan, L. Ren, W. Chen, and Z. Tan, “Open3DFlow: An open-source EDA platform for 3D chip design with AI enhancement,” in Proceedings of ICCAD. Piscataway, NJ, USA: IEEE, 2025, pp. 1–9.DOI
  22. [22]Open3DFlow Authors, “Open3DFlow: Open-source project and demonstrations,” n.d., project artifact; accessed October 28, 2026. [Online]. Available: https://github.com/bzhang24/Open3DFlow
  23. [23]L. Jiang, A. B. Kahng, Z. Wang, and Z. Zheng, “Invited: Toward sustainable and transparent benchmarking for academic physical design research,” 2026. [Online]. Available: https://arxiv.org/abs/2601.17520
  24. [24]S. Patnaik, M. Ashraf, O. Sinanoglu, and J. Knechtel, “Best of both worlds: Integration of split manufacturing and camouflaging into a security-driven CAD flow for 3D ICs,” in Proceedings of ICCAD. New York, NY, USA: Association for Computing Machinery, 2018, pp. 1–8.DOI
  25. [25]L. T. Clark, V. Vashishta, L. Shifren, A. Gujja, S. Sinha, B. Cline, C. Ramamurthy, and G. Yeric, “ASAP: A 7-nm FinFET predictive process design kit,” Microelectronics Journal, vol. 53, pp. 105–115, 2016.DOI
  26. [26]V. Vashishta, M. Vangala, and L. T. Clark, “ASAP7 predictive design kit development and cell design technology co-optimization: Invited paper,” in Proceedings of ICCAD. Piscataway, NJ, USA: IEEE, 2017, pp. 992–998.

Paper details

vlnc:W_pv9rq2m7
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Computer Science

Not recorded for this paper

View statement
Looking for related papers…
Contents