Signal integrity at the speed of design

Design high-speed PCB channels in your browser: pick the laminate, budget the loss, size the traces and check against PCIe and Ethernet standards. Results update as you type.

Join the WaitlistRequest Demo
Runs in your browserNo install or license serverEvery formula cited
  pcie_channel.sim — DSS · layout
Zdiff 85 ΩTXRXS-parameters to 40 GHz · differential microstrip · solved in 240 ms
Software Lineup

Signal integrity, from laminate to standard

Browser-based tools for high-speed PCB channels: choose the material, budget the loss, size the traces and check the result against the interface standard.

〰️

DSS — PCB Channel Workbench

Step-by-step channel setup: laminate selection, Tx-to-Rx loss budget, trace impedance, net groups, stackup and bump map. Closed-form models (Hammerstad–Jensen, IPC-2141) update as you type.

Join the waitlist →
🧩

Component Library

Vias, microstrip, grounded CPW, stripline, wire bonds and SMA, 2.92, 2.4 and 1.85 mm connectors. Sweep geometry and filter datasets by impedance, insertion loss and return loss.

Browse the library →
✅

Standards Check

Compare every channel with PCIe Gen3–Gen6, 10G and 25G-KR and 100G PAM4 budgets: insertion loss, differential impedance, laminate loss and return-path checks in one table.

Join the waitlist →
How It Works

From requirements to sign-off, step by step

1

Material

Set data rate and loss budget; compare laminates by loss, thermal limits and cost.

2

Channel

Lay out each link from Tx to Rx (packages, routes, vias, connectors) against its loss budget.

3

Trace

Solve trace width for a target impedance, with S-parameters and eye diagram.

4

Group

Group nets into buses and clocks with length-match and via limits.

5

Layers

Build the stackup and assign each group to a signal layer next to a reference plane.

6

Bumps

Lay out the bump map and assign nets, ground and power.

7

Standards

Check every channel against PCIe and Ethernet budgets and impedance limits.

S21 · model vs measuredModelMeasuredFrequency, 1–40 GHz
Instant Models

Answers as you type, no solve queue

Impedance, loss, delay and S-parameters come from closed-form models that run in your browser, so every change updates at once. Measured data sits on the same plots for comparison.

  • Hammerstad–Jensen and IPC-2141 impedance, with conductor and dielectric loss
  • S21, S11 and a first-order eye diagram for every trace type
  • Upload measured S-parameters and overlay them on the model
Connected Steps

Every step feeds the next

The laminate you pick sets trace impedance and stackup, channel routes name the trace types, and the standards check reads every channel, trace and layer. Change one input and the rest follow.

  • Separate projects, each with its own data and units
  • Channels, traces, groups, stackup and bump map linked together
  • Standards check across all channels in one table
1MaterialIsola FR408HR · Er 3.662Channel2 channels · 18.4 / 28 dB3TraceSE 50 Ω · w 9.4 mil5Layers8 layers · all refs OK7StandardsPCIe Gen4 · 1 checkProject 1 · setup steps · standards check flags one item
Applications

Built for high-speed and RF hardware

Wherever a channel has to meet a loss budget or an impedance limit.

🖥️

Data center and networking

PCIe Gen3–Gen6 and 10G to 100G Ethernet channels, from laminate choice to loss budget.

🔲

Chip packaging

Bump maps, wire bonds and package-to-board transitions.

📡

RF and mm-wave

Grounded CPW, stripline and precision connectors up to 67 GHz.

📏

Test and measurement

Overlay measured S-parameters on the model to see where they disagree.

7setup steps, material to standards
6component families in the library
7interface standards checked
21sources cited in our formula guide
From the Blog

The engineering behind the tools

Using Analytical Formulas for High-Speed Interconnects

The formulas behind the models, their sources and valid ranges, and six considerations.

Read the article →

How Much Data Is Enough?

How many solver runs a sensitivity study needs, and how to predict where the response curves.

Read Part A →

Simulation–Measurement Correlation

Why simulated and measured responses disagree, and five ways engineers close the gap.

Read the article →

Ready to check your next channel?

Join the waitlist and we'll email you when your access is ready.

Join the WaitlistTalk to Us