1

Physical Design via

Place-and-Route: RTL to GDS
Edward Wang
April 10, 2018
 2
RTL
● Stands for Register Transfer Level
● An abstraction for digital circuits, consisting of
○ Combinational logic
○ Registers (state elements)
○ Modules (hierarchical and “blackbox” - e.g. analog macros, SRAM macros, etc) and ports/nets
● Described in terms of a hardware description language (HDL)
 3
Hardware description languages (HDLs)
● An HDL is a language for describing circuits using the RTL abstraction.
● Includes facilities for describing combinational logic, registers/state, and
modules.
● Common HDLs: Verilog, VHDL.
● Research-y HDLs: FIRRTL, CoreIR.
module add_one :
input clock : Clock
input reset : UInt<1>
output io : {flip in : UInt<4>, out :
UInt<4>}

io.out <= tail(add(io.in,

UInt<1>("h01")), 1)
 4
RTL in Action (FIRRTL)
circuit HelperDelayedAdd2 :
module HelperDelayedAdd2 :
input clock : Clock
Module/port
input reset : UInt<1> description
output io : {flip in : UInt<4>, out : UInt<4>}

reg my_reg : UInt, clock Registers/state

my_reg <= tail(add(io.in, UInt<1>("h01")), 1) (Combinational)

io.out <= tail(add(my_reg, UInt<1>("h01")), 1) logic
 5
RTL in Action (Verilog)
module HelperDelayedAdd2 (
input clock,
input reset,
Module/port
input [3:0] in, description
output [3:0] out);

reg [3:0] my_reg; Registers/state

always @ (posedge clock) begin (Combinational)

my_reg <= in + 4’h1; logic
end
assign out = my_reg + 4’h1;

endmodule
6
 7
Digression: HDL vs HCL
● Chisel (strictly speaking) isn’t an HDL...
● What’s the difference?

HDL HCL
source source
 8
Other HCLs/HDLs
● PyMTL
● Bluespec
● Magma
● Lava
● Netlist
● etc
 9
RTL Design Is Only Part of the Picture
 10
RTL Design Is Only Part of the Picture
 11
What Makes Creating Hardware Difficult?
● What makes the design cycle long and expensive?
○ Architectural Design Space Exploration
○ RTL Development [J. Bachrach et al, DAC 2012]
○ Physical Design and Implementation
○ Verification - “is it correct?”
○ Validation - “is it the right problem to solve?”
● Compilers and Generators
○ Having reliable, re-usable, and robust tools is the name of the
game
○ BAG, Chisel, etc.
12
 13
Why Physical Design Matters
 Physical Design is HARD - CAD Tools Aren’t Automatic
Legend
source code generated file
“WhichVerilog
logic gates do I need?”
Instances

CAD Tools “Can you try toTCL Scripts

place them in this way?”

Chip gates end up?”

“Where did the logic
Agile RTL is slowed by non-Agile Physical Design
RTL Generator

Chisel + FIRRTL

Verilog Instances

CAD Tools
TCL Scripts

Chips
 16
Digression: Why Agile Physical Design?
1. Analog/Mixed-Signal (AMS) Systems
2. Improved Usability for Faster Design Space Exploration
3. Technology Portability
4. Hierarchical Design
 17

Research Plug
 18
HAMMER/CICL: A Modular Platform to Encode
Expertise and Intent
● Physical design is a collection of many difficult problems
○ No silver bullet
● Need to lower barrier to solving these problems
○ Other tapeouts solve these problems, but their solutions are not general or reusable
○ Get designers to encode solutions in a more reusable way, so future tapeouts can leverage
previous work (even with different technologies, CAD tools, or designs)
● Provide collection of API’s that designers leverage to build these tools
○ Higher-level and CAD-tool independent directives
○ Directly manipulate/introspect on RTL
○ Higher-level technology abstractions
 19
What HAMMER means for this class
1. Re-use of other research tapeouts’ efforts
2. Faster flow development
3. Abstractions to reduce the complexity of VLSI flows and make them more
accessible
4. Encoding designer knowledge/expertise in a robust way
a. There’s a ton of info that ends up in people’s heads as you do stuff, and it’s hard to write stuff
down in a productive way
b. Reducing pain for future tapeout students like yourselves
Big Picture Overview (Simplified) 20
Chisel SRAM Cache

FIRRTL PDK and

MacroCompiler Analog Macros Standard Cell
Library

Physical Design
SRAMs FIRRTL
Power Timing & I/O
Floorplan
Strategy Constraints
Verilog

TCL

GDS Note: in real life chips need to go on

boards with packages, etc.
 21
Chisel -> FIRRTL
● Recap: Chisel is a HCL embedded in Scala
● That is to say - every Chisel design is a Scala program, which when executed,
emits a concrete instance of a circuit in FIRRTL.
● We are using digital top (place and route tool will manage the top level), so we
will instantiate analog macros in the digital top.
● A brief note on scan chains: we will use a scan chain generator written in
Chisel
 22
MacroCompiler
● In Chisel, we specify memories using an abstract Mem()/SyncReadMem()
construct:
class SRAMTest extends Module {
val io = IO(new Bundle {
val in = Input(UInt(32.W))
val en = Input(Bool())
val out = Output(UInt(32.W))
})
val counter = Counter(1024)
val mem = SyncReadMem(1024, UInt(32.W))
when (io.en) {
mem.write(io.in, counter.value)
counter.inc()
}
io.out := mem.read(io.in)
}
 What it looks like in Verilog (i.e. giant
23
MacroCompiler bank of flip flops):

circuit HelperSRAMTest : reg [31:0] mem [0:1023];

module HelperSRAMTest :
input clock : Clock
input reset : UInt<1>
output io : {flip in : UInt<32>, flip en : UInt<1>, out : UInt<32>}

reg value : UInt<10>, clock with : (reset => (reset, UInt<10>("h00"))

smem mem : UInt<32>[1024]
when io.en :
write mport _T_10 = mem[bits(io.in, 9, 0)], clock
[...]
node _T_16 = bits(io.in, 9, 0)
read mport _T_17 = mem[_T_16], clock
io.out <= _T_17
 24
MacroCompiler
● However, by default, these memories would compile to standard cell
flip-flops, which is very area-inefficient for implementing memories in contrast
to SRAM macros. Example:

SRAM macro implementation

Flip-flop implementation
 25
MacroCompiler
● Solution: FIRRTL compiler passes that identify the generic memories from
Chisel/FIRRTL (ReplSeqMem) and replace them with modules which use
collections of BlackBox SRAM memories (MacroCompiler) given a cache of
technology SRAMs.

● ReplSeqMem: Replace mem => mem_ext (create blackboxes)

● MacroCompiler: Create the mem_ext module which uses technology SRAMs
 26
Timing and I/O Constraints
● Clock constraints - tells the tool about clock frequencies, uncertainty/jitter,
etc.
○ Can also specify related clocks
● I/O constraints - specifies input and output delays, capacitances for external
pins
● I/O types/cells - specifies I/O types (input, output, tri-state) and corresponding
cells to drive pins
 27
Bumps vs Wirebond Pads
● Bumps: metal (e.g. Cu) bumps on top of the chip which we flip over and bond to a
substrate/board
● Wirebond pads: wires are used to bond exposed metal on top of the chip to a
substrate/package/board
● In this class, we will use wirebond pads

source source
 28
False Paths
● False paths
○ A logically impossible path that appears with a naive analysis.
○ Look at the timing report and declare it as a false path.
○ Dangerous if misused

source
 29
Synthesis
● Maps RTL (Verilog) to a post-synthesis netlist (structural Verilog).
● Standard cells come in different sizes and drive strengths.
● The synthesis tool uses the previously-mentioned constraints to select
standard cells appropriately.
● Synthesis will also perform optimizations to simplify the RTL.
○ E.g. if all of a module’s inputs are constants, it may optimize away the module entirely by
precomputing its outputs.

RTL

Standard cell library Synthesis

Constraints
 30
Synthesis Example

module adder(a, b, c);

input [1:0] a, b;
output [1:0] c;
wire [1:0] a, b;
module adder ( wire [1:0] c;
input [1:0] a, wire n_0, n_1, n_3, n_4, n_5, n_6;
NAND2X54_P0 g80__7837(.A (n_6), .B (n_5), .Z (c[1]));
input [1:0] b, NAND2X3_P0 g81__7557(.A (n_3), .B (n_4), .Z (n_6));
output [1:0] c OR2X8_P0 g82__7654(.A (n_4), .B (n_3), .Z (n_5));
); NAND2X54_P0 g84__8867(.A (n_0), .B (n_1), .Z (c[0]));
assign c = a + b; XNOR2X6_P0 g83__1377(.A (b[1]), .B (a[1]), .Z (n_3));
endmodule NAND2AX3_P0 g86__3717(.A (a[0]), .B (b[0]), .Z (n_1));
NAND2AX3_P0 g85__4599(.A (b[0]), .B (a[0]), .Z (n_0));
AND2X8_P0 g87__3779(.A (b[0]), .B (a[0]), .Z (n_4));
endmodule
 31
Floorplanning
● Recall: RTL says what to put (logic, state, macros) but doesn’t say where to
put stuff
● Floorplanning is the art of specifying placement constraints.
● Main types of placement constraints:
○ Chip size - tells the place and route tool how large the chip is and how much padding there is
○ Module placement - tells the place and route tools to put cells from a certain module within a
certain boundary
○ Macro placement - tells the place and route tool where to put macros (e.g. analog blocks,
SRAMs, etc)
 32
Standard Cell Layout
● Digital layout typically uses
standard cells (as opposed to
fully custom layouts in analog).
● Standard cells are transistor-level
implementations of CMOS logic
gates.
● Typical structure of a standard
cell includes power/ground rails
and pins.

source source
 33
Standard Cell Layout
● Standard cells are assembled into
layouts in tracks by placing them
next to each other.
● Signals are routed in layers above
the standard cells.
● Power is routed to the rails (vdd
and gnd) via a power plan (e.g.
power grid and vias).
● Each row is typically mirrored
(vdd->gnd, gnd->vdd, etc)
● Overlap rails, not abut them

source
 34
Other Aspects of Floorplanning
● Power planning - defines the power strategy for the chip. For example, a
power plan for the chip can involve creating grids for VDD and GND on each
layer.
● Tap cells - technologies require the substrate/body to be “tapped” to a known
voltage. Standard cells exist to perform this body tapping. Some stdcell
architectures come with built-in taps.
● Filler cells - in order to meet density requirements, unused space must be
filled, typically with decap.
 35
Floorplan Visualization (Example)
 36
Place and Route
● Given a post-synthesis netlist and floorplanning/physical design constraints,
create a physical layout by placing standard cells on the chip and creating
wires to route between the different cells.
● Performs standard cell placement and routing while respecting the
floorplanning/physical design constraints and routing to macros (e.g. analog
macros, SRAMs).
● The final result is a GDS file which can be sent to the fab.
 37
DRC
● Design Rule Check (DRC) is the process of checking that the geometry in the
GDS file follows the rules given by the fab.
● Digital standard cell layouts must still obey design rules.
● Errors often happen when designs/layouts are integrated together.
● DRC rules in advanced technologies are extremely complex and confusing.
● Sometimes CAD tools can do stupid things!

(Edward Wang, June 2017)

 38
LVS
● Layout vs. Schematic (LVS) is process of checking that the geometry/layout
matches the schematic/netlist.
● CAD tools can export netlists for digital designs.
● As before, LVS errors can often arise when blocks are integrated together.
● They are confusing since a shorted net can mess up the entire check!
(Lydia Lee, June
2017)
 39
Verification
● We can run simulations on post-synthesis and post-place and route netlists
(RTL) in order to check that the system still functions as intended.
● In industry, they run these checks with timing annotations so that setup and
hold times aren’t violated.