feat(hw): add banked CAM pipeline with grouped read/write noise

- Add cam_core_banked.sv with 8-lane banked CAM core
- Add cam_write_noise.sv and cam_read_noise.sv for grouped noise injection
- Add noise_mask_grouped.sv generating grouped flip masks from 128-bit PRNG
- Add match_engine_pipeline.sv with multi-stage pipelined top-1 selection
- Add popcount_pipeline.sv for pipelined popcount operations
- Refactor test_cam_basic.py with parametrized DUT introspection helpers
- Add Python ref_model match_top1_with_read_noise() for read noise verification
- Update Makefile with separate WRITE_NOISE_* and READ_NOISE_* parameter groups
- Add new testbenches: test_cam_core_banked, test_cam_read_noise,
  test_cam_write_noise, test_match_engine_pipeline, test_ref_model_noise
  
breaking change hint: NUM_ROWS default changed from 512→4096, LANES from 16→8

hw/rtl/popcount_pipeline.sv

@@ -0,0 +1,72 @@
+`timescale 1ns / 1ps
+
+module popcount_pipeline #(
+    parameter int WIDTH = 512,
+    parameter int ROW_BITS = 12,
+    parameter int OUT_WIDTH = $clog2(WIDTH + 1)
+) (
+    input  logic                 clk,
+    input  logic                 rst_n,
+    input  logic                 valid_i,
+    input  logic [ROW_BITS-1:0]  row_i,
+    input  logic [WIDTH-1:0]     bits_i,
+    output logic                 valid_o,
+    output logic [ROW_BITS-1:0]  row_o,
+    output logic [OUT_WIDTH-1:0] count_o
+);
+    localparam int GROUP = 8;
+    localparam int NUM_GROUPS = WIDTH / GROUP;
+    localparam int GROUP_COUNT_WIDTH = $clog2(GROUP + 1);
+    localparam int STAGE1_GROUPS = NUM_GROUPS / 4;
+
+    logic [GROUP_COUNT_WIDTH-1:0] group_counts [0:NUM_GROUPS-1];
+    logic [OUT_WIDTH-1:0] partial_q [0:3];
+    logic [OUT_WIDTH-1:0] sum_q;
+    logic [ROW_BITS-1:0] row_s1_q, row_s2_q;
+    logic valid_s1_q, valid_s2_q;
+    logic [OUT_WIDTH-1:0] partial_comb [0:3];
+
+    always_comb begin
+        for (int g = 0; g < NUM_GROUPS; g++) begin
+            group_counts[g] = '0;
+            for (int b = 0; b < GROUP; b++) begin
+                group_counts[g] = group_counts[g] + bits_i[g*GROUP + b];
+            end
+        end
+
+        for (int p = 0; p < 4; p++) begin
+            partial_comb[p] = '0;
+            for (int g = 0; g < STAGE1_GROUPS; g++) begin
+                partial_comb[p] = partial_comb[p] + group_counts[p*STAGE1_GROUPS + g];
+            end
+        end
+    end
+
+    always_ff @(posedge clk or negedge rst_n) begin
+        if (!rst_n) begin
+            valid_s1_q <= 1'b0;
+            valid_s2_q <= 1'b0;
+            valid_o    <= 1'b0;
+            row_s1_q   <= '0;
+            row_s2_q   <= '0;
+            row_o      <= '0;
+            sum_q      <= '0;
+            count_o    <= '0;
+            for (int p = 0; p < 4; p++) partial_q[p] <= '0;
+        end else begin
+            valid_s1_q <= valid_i;
+            row_s1_q   <= row_i;
+            for (int p = 0; p < 4; p++) begin
+                partial_q[p] <= partial_comb[p];
+            end
+
+            valid_s2_q <= valid_s1_q;
+            row_s2_q   <= row_s1_q;
+            sum_q      <= partial_q[0] + partial_q[1] + partial_q[2] + partial_q[3];
+
+            valid_o <= valid_s2_q;
+            row_o   <= row_s2_q;
+            count_o <= sum_q;
+        end
+    end
+endmodule