VPU

The Vector Processing Unit (VPU) is a configurable pipeline that applies activation functions, computes loss derivatives, and calculates activation derivatives. It supports three data pathways for forward pass, transition, and backward pass operations.

Module declaration

module vpu (
    input logic clk,
    input logic rst,
    input logic [3:0] vpu_data_pathway,
    // Inputs from systolic array
    input logic signed [15:0] vpu_data_in_1,
    input logic signed [15:0] vpu_data_in_2,
    input logic vpu_valid_in_1,
    input logic vpu_valid_in_2,
    // Inputs from unified buffer
    input logic signed [15:0] bias_scalar_in_1,
    input logic signed [15:0] bias_scalar_in_2,
    input logic signed [15:0] lr_leak_factor_in,
    input logic signed [15:0] Y_in_1,
    input logic signed [15:0] Y_in_2,
    input logic signed [15:0] inv_batch_size_times_two_in,
    input logic signed [15:0] H_in_1,
    input logic signed [15:0] H_in_2,
    // Outputs to unified buffer
    output logic signed [15:0] vpu_data_out_1,
    output logic signed [15:0] vpu_data_out_2,
    output logic vpu_valid_out_1,
    output logic vpu_valid_out_2
);

Input ports

Systolic array inputs

Port	Width	Description
`vpu_data_in_1`	`signed [15:0]`	Data input from systolic array column 1
`vpu_data_in_2`	`signed [15:0]`	Data input from systolic array column 2
`vpu_valid_in_1`	1	Valid signal for column 1
`vpu_valid_in_2`	1	Valid signal for column 2

Unified buffer inputs

Port	Width	Description
`bias_scalar_in_1`	`signed [15:0]`	Bias value for column 1
`bias_scalar_in_2`	`signed [15:0]`	Bias value for column 2
`lr_leak_factor_in`	`signed [15:0]`	Leak factor α for leaky ReLU (Q8.8 format)
`Y_in_1`	`signed [15:0]`	Ground truth label for loss computation (column 1)
`Y_in_2`	`signed [15:0]`	Ground truth label for loss computation (column 2)
`inv_batch_size_times_two_in`	`signed [15:0]`	Scaling factor: 1/(batch_size × 2)
`H_in_1`	`signed [15:0]`	Cached activation value for derivative (column 1)
`H_in_2`	`signed [15:0]`	Cached activation value for derivative (column 2)

Control signal

Port	Width	Description
`vpu_data_pathway`	`[3:0]`	Module enable bits: `[bias	leaky_relu	loss	leaky_relu_derivative]`

Output ports

Port	Width	Description
`vpu_data_out_1`	`signed [15:0]`	Processed data output for column 1
`vpu_data_out_2`	`signed [15:0]`	Processed data output for column 2
`vpu_valid_out_1`	1	Valid signal for column 1 output
`vpu_valid_out_2`	1	Valid signal for column 2 output

Architecture

Module pipeline

The VPU consists of four processing stages:

Bias (bias_parent): Adds bias to input values
Leaky ReLU (leaky_relu_parent): Applies leaky ReLU activation
Loss (loss_parent): Computes loss derivative (∂L/∂H)
Leaky ReLU Derivative (leaky_relu_derivative_parent): Computes activation derivative

From ~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/src/vpu.sv:115-185:

bias_parent bias_parent_inst (...);           // Z = X + b
leaky_relu_parent leaky_relu_parent_inst (...); // H = LeakyReLU(Z)
loss_parent loss_parent_inst (...);           // ∂L/∂H = (H - Y) / (batch_size × 2)
leaky_relu_derivative_parent lr_d_inst (...); // ∂L/∂Z = ∂L/∂H ⊙ LeakyReLU'(H)

Data pathways

The vpu_data_pathway signal configures the active modules:

vpu_data_pathway[3]

bit

Bias module enable (1 = enabled)

vpu_data_pathway[2]

bit

Leaky ReLU module enable (1 = enabled)

vpu_data_pathway[1]

bit

Loss module enable (1 = enabled)

vpu_data_pathway[0]

bit

Leaky ReLU derivative module enable (1 = enabled)

Operation modes

Forward pass pathway (`4'b1100`)

Systolic Array → Bias → Leaky ReLU → Output

Computes: H = LeakyReLU(Z) where Z = X + b Use case: Hidden layer activations during forward propagation

Transition pathway (`4'b1111`)

Systolic Array → Bias → Leaky ReLU → Loss → Leaky ReLU Derivative → Output

Computes: ∂L/∂Z = (H - Y) / (batch_size × 2) ⊙ LeakyReLU'(H) Use case: Final layer computation that transitions from forward to backward pass. The H matrix is cached internally for use in the derivative calculation.

Backward pass pathway (`4'b0001`)

Systolic Array → Leaky ReLU Derivative → Output

Computes: ∂L/∂Z = ∂L/∂H ⊙ LeakyReLU'(H) Use case: Hidden layer gradients during backpropagation. H values come from H_in_* ports.

Inactive mode (`4'b0000`)

All modules bypassed, no processing occurs.

Activation caching

The VPU includes an internal cache for H matrices (activation outputs): From ~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/src/vpu.sv:333-348:

always @(posedge clk or posedge rst) begin
    if (rst) begin
        last_H_data_1_out <= 0;
        last_H_data_2_out <= 0;
    end else begin
        if (vpu_data_pathway[1]) begin  // Loss module enabled
            last_H_data_1_out <= last_H_data_1_in;
            last_H_data_2_out <= last_H_data_2_in;
        end
    end
end

When the loss module is active (transition pathway), the leaky ReLU outputs are stored and used for the derivative calculation in the same pass.

Signal routing logic

The VPU uses combinational logic to route signals through the enabled modules: From ~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/src/vpu.sv:187-330:

always @(*) begin
    if(vpu_data_pathway[3]) begin      // Bias enabled
        bias_data_1_in = vpu_data_in_1;
        b_to_lr_data_in_1 = bias_z_data_out_1;
    end else begin                     // Bias bypassed
        b_to_lr_data_in_1 = vpu_data_in_1;
    end
    
    if(vpu_data_pathway[2]) begin      // Leaky ReLU enabled
        lr_data_1_in = b_to_lr_data_in_1;
        lr_to_loss_data_in_1 = lr_data_1_out;
    end else begin                     // Leaky ReLU bypassed
        lr_to_loss_data_in_1 = b_to_lr_data_in_1;
    end
    
    // Similar logic for loss and leaky_relu_derivative modules
end

Timing behavior

Bias: 1 clock cycle latency
Leaky ReLU: 1 clock cycle latency
Loss: 1 clock cycle latency
Leaky ReLU Derivative: 1 clock cycle latency

Total latency depends on active pathway:

Forward pass: 2 cycles (bias + leaky ReLU)
Transition: 4 cycles (all modules)
Backward pass: 1 cycle (leaky ReLU derivative only)

Example instantiation

From ~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/src/tpu.sv:157-184:

vpu vpu_inst (
    .clk(clk),
    .rst(rst),
    .vpu_data_pathway(vpu_data_pathway),
    
    // From systolic array
    .vpu_data_in_1(sys_data_out_21),
    .vpu_data_in_2(sys_data_out_22),
    .vpu_valid_in_1(sys_valid_out_21),
    .vpu_valid_in_2(sys_valid_out_22),
    
    // From unified buffer
    .bias_scalar_in_1(ub_rd_bias_data_out_0),
    .bias_scalar_in_2(ub_rd_bias_data_out_1),
    .lr_leak_factor_in(vpu_leak_factor_in),
    .Y_in_1(ub_rd_Y_data_out_0),
    .Y_in_2(ub_rd_Y_data_out_1),
    .inv_batch_size_times_two_in(inv_batch_size_times_two_in),
    .H_in_1(ub_rd_H_data_out_0),
    .H_in_2(ub_rd_H_data_out_1),
    
    // To unified buffer
    .vpu_data_out_1(vpu_data_out_1),
    .vpu_data_out_2(vpu_data_out_2),
    .vpu_valid_out_1(vpu_valid_out_1),
    .vpu_valid_out_2(vpu_valid_out_2)
);

TPU - Top-level integration
Unified Buffer - Data source and destination
Bias parent (~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/src/bias_parent.sv)
Leaky ReLU parent (~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/src/leaky_relu_parent.sv)
Loss parent (~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/src/loss_parent.sv)
Leaky ReLU derivative parent (~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/src/leaky_relu_derivative_parent.sv)

Testing

See test files:

~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/test/dump_vpu.sv - Waveform dump configuration
~https://github.com/tiny-tpu-v2/tiny-tpu/blob/main/test/test_vpu.py - Python test suite

Core Modules

VPU Components

Module declaration

Input ports

Systolic array inputs

Unified buffer inputs

Control signal

Output ports

Architecture

Module pipeline

Data pathways

Operation modes

Forward pass pathway (`4'b1100`)

Transition pathway (`4'b1111`)

Backward pass pathway (`4'b0001`)

Inactive mode (`4'b0000`)

Activation caching

Signal routing logic

Timing behavior

Example instantiation

Testing

Build docs developers (and LLMs) love

Core Modules

VPU Components

Documentation Index

​Module declaration

​Input ports

​Systolic array inputs

​Unified buffer inputs

​Control signal

​Output ports

​Architecture

​Module pipeline

​Data pathways

​Operation modes

​Forward pass pathway (4'b1100)

​Transition pathway (4'b1111)

​Backward pass pathway (4'b0001)

​Inactive mode (4'b0000)

​Activation caching

​Signal routing logic

​Timing behavior

​Example instantiation

​Related modules

​Testing

Build docs developers (and LLMs) love

Module declaration

Input ports

Systolic array inputs

Unified buffer inputs

Control signal

Output ports

Architecture

Module pipeline

Data pathways

Operation modes

Forward pass pathway (`4'b1100`)

Transition pathway (`4'b1111`)

Backward pass pathway (`4'b0001`)

Inactive mode (`4'b0000`)

Activation caching

Signal routing logic

Timing behavior

Example instantiation

Related modules

Testing