PCB & Schematic · SI / PI · FMEDA & FMEA · IP Catalogs · RISC-V / SoC · Verification · Due Diligence · Packages
Why circuit-design.space
From Fortune 500s to startups, our team blends deep industry knowledge with a genuine appetite for what’s next — and the delivery discipline to ship it.
Delivery
Exceptional management and delivery — every project owned end-to-end.
Technology
Cutting-edge tools, chosen for the problem rather than the trend.
Quality
Driven by quality — zero-failure delivery is the standard, not the stretch goal.
Speed
One week from first conversation to project kickoff.
Talent
Top-tier industry talent across design, engineering, and production.
Partnership
We work as an extension of your team, not a vendor at arm’s length.
1. PCB Design, Schematic Capture & Layout
Our core work. Most of it is mixed-signal: an FPGA or microcontroller next to sensitive analog front-ends, clean power rails, and a fabric of signal-integrity constraints that have to be settled before routing starts.
- Cadence Allegro
- Altium
- 4 to 20+ layers
- Blind/buried vias, HDI
- Schematic capture: Cadence Allegro, Altium, OrCAD. Library management, BOM discipline
- Circuit analysis and design: analog front-ends, power supplies (switchers and LDOs), clocking trees, reset/brownout, level shifting, isolation
- PCB layout: multi-layer stackups including buried capacitance, controlled impedance, blind/buried vias, HDI. 4-layer to 20+ layers
- Simulation: LTspice for analog, thermal hand-analysis with FEA where the power density warrants it
- Design for manufacture: fabrication and assembly constraints applied while the layout is still movable, not after the quote comes back
- Fabrication: PCB manufacturing resource available — we can take a design from schematic to populated board without you chasing vendors
- Circuit design only (customer does layout)
- Full schematic + PCB layout
- Design review of an existing board (1 week, fixed scope)
- Failure analysis and debug on an existing product
2. Signal & Power Integrity, EMI/EMC
Boards fail in the chamber and in the field for reasons that were settled at stackup. We model them before that happens — and when a product is already built, we find out why it is misbehaving rather than guessing.
- Ansys SIwave
- Siemens HyperLynx
- IBIS / IBIS-AMI
- Pre- and post-layout
- Pre-layout constraint development: topology, termination, stackup and impedance targets fixed before routing starts
- Post-layout verification: SIwave and HyperLynx run on the real board, not on the intent
- Power integrity: PDN impedance targets, decoupling strategy, plane resonance and IR-drop analysis
- Signal integrity: reflection, crosstalk, timing and eye closure using IBIS and IBIS-AMI models
- 3D field solver analysis where the geometry stops behaving like a transmission line — connectors, via transitions, breakout regions
- EMI/EMC: design for pre-compliance before you step into the chamber
- Failure analysis: measurement on the bench, correlation back to the model, and the fix
- Pre-layout constraint study (1–2 weeks)
- Post-layout SI/PI verification of an existing design (1–2 weeks)
- PDN and decoupling review (1 week, fixed scope)
- Failure analysis and debug on a product already in the field
3. Functional Safety — FMEDA & FMEA Review (ISO 26262)
Safety-critical board and silicon designs live or die on diagnostic coverage. FMEDA — Failure Modes, Effects, and Diagnostic Analysis — is where that coverage gets proven to auditors.
- ISO 26262
- IEC 61508 / 62304 / DO-254
- Board and silicon level
- Pre-audit gap analysis
- FMEDA spreadsheets and reports for ASIL-A through ASIL-D parts
- Design FMEA and Process FMEA review — structure, ranking discipline, and whether the listed actions actually close the risk they claim to
- Failure rate (FIT) budgets per subsystem
- Diagnostic coverage modeling, including latent-fault detection
- Safety mechanisms: lockstep cores, ECC strategies, watchdog architectures, BIST
- Safety case artifacts: assumptions of use, safety goals, safety requirements traceability
- Pre-audit gap analysis — what your existing documentation is missing before the certification body arrives
- Review an existing FMEDA or FMEA and identify gaps (1–2 weeks)
- Build an FMEDA from scratch for a new board, IP or module (3–8 weeks)
- On-call safety engineer during an active certification cycle
Free tools:ASIL determination, the FIT and PMHF budget calculator, the FMEDA rollup, board FIT from a BOM and FMEA action priority — all free, all in your browser. See all nine tools.
4. Safety Soft-IP & Verification IP
Alongside the design services we license two catalogs: safety-instrumented soft-IP with the safety analysis already done, and independent verification IP written clean-room from the public protocol specifications.
- 182 safety blocks
- ASIL-B
- FMEDA + IP-XACT included
- 139 VIP · 55 shipping
- See the catalog →
- Safety Soft-IP: 182 blocks with published SPFM/LFM figures, FMEDA, IP-XACT descriptors and formal sign-off — the analysis ships with the block, not as a later invoice
- Subsystems: safety island, NPU island, coherent fabric (CHI), post-quantum cryptography
- Verification IP: 139 protocol VIP catalogued, 55 shipping today — BFMs, monitors and passive protocol checkers
- Independent second source: our VIP is written from the public specification, so it can genuinely disagree with your primary vendor’s VIP rather than share its assumptions
- Integration support: bringing a catalog block into your design, with its safety case intact
- License a block or a VIP (deliverables and terms under mutual NDA)
- Second-source verification of a protocol you already have a VIP for
- Custom safety instrumentation of an existing block
5. RISC-V / SoC Architecture
Our principal has been defining silicon for over two decades — from custom cores at IBM and Apple to full SoC architecture at APCON and Intensivate-era RISC-V work.
- Safety island
- NPU island
- Coherent fabric (CHI)
- Post-quantum crypto
- See the catalog →
- Core specifications: ISA profile selection, pipeline definition, microarchitecture trade-offs
- Cache and memory hierarchy design (L1/L2/L3, coherence protocol choice, MMU/PMP configuration)
- Custom ISA extensions for domain-specific workloads
- Full SoC integration: bus fabric (AXI, CHI, TileLink), NoC topology, power domains, clock domains
- Architecture reviews — does the proposed design actually meet the performance, power, and area targets, and if not, where’s the leak?
- Pre-RTL modeling (cycle-approximate SystemC or pure-Python)
- Design review (1 week, fixed scope)
- Architecture definition for a new core or SoC (1–3 months)
- Advisory retainer — monthly check-in + on-call for a team doing their own implementation
6. Verification & Emulation
Verification eats schedules. We’ve spent years pushing testbenches, formal, and emulation platforms to close coverage faster.
- SystemVerilog + UVM
- JasperGold / Synopsys Formal
- ZeBu & HAPS
- TLA+
- See the catalog →
- Simulation-based verification: SystemVerilog + UVM testbenches, constrained-random generation, functional coverage closure
- Formal verification: JasperGold, Synopsys Formal. Property specification, assume/assert strategy, bounded and unbounded proofs
- Specification-level verification: TLA+ models for protocol correctness before RTL exists
- Emulation bring-up: Synopsys ZeBu and HAPS platforms. Design partitioning, clock domain handling, DUT instrumentation, regression flows
- FPGA prototyping: Xilinx-based prototypes for software bring-up ahead of tape-out
- Tool ownership: Synopsys VCS and Verdi, Cadence Xcelium, waveform-first debug methodology
- Testbench audit — find coverage gaps in an existing environment (1–2 weeks)
- Formal property set for a new block (2–6 weeks)
- ZeBu/HAPS bring-up for a new SoC (4–12 weeks)
- TLA+ model for a contested protocol spec (1–3 weeks)
7. Technical Due Diligence
For VCs evaluating a hardware investment, acquirers inspecting a target, or boards that want a second opinion on their own team’s plans.
- Light — 1 week
- Standard — 2 weeks
- Deep — 3–4 weeks
- Written risk-ranked report
- Architecture soundness: does the proposed silicon or system actually do what the pitch claims?
- Execution risk: are the schedules realistic? Is the team sized for the work? What’s the critical path?
- IP review: freedom-to-operate check, open-source license exposure, third-party IP dependencies
- Team capability assessment: interviews with key engineers, code/RTL samples, tool proficiency
- Comparable analysis: how does this compare to what competitors are doing or have shipped?
- Written report: executive summary + detailed findings + risk-ranked recommendations
- Light DD — 1 week, documentation and interviews only
- Standard DD — 2 weeks, includes one-day site visit and deeper code/RTL inspection
- Deep DD — 3–4 weeks, includes running hands-on evaluation of the tech
Fixed-scope engagements
Short, bounded pieces of work with a defined deliverable and a known duration — the usual way a longer engagement starts. Each one ends in a written artifact you keep, whether or not the work continues.
Bring a schematic, a stackup, a failing board or a specification that is not yet a design. Half an hour on your actual problem and a straight answer about whether it needs a review, a redesign, or nothing at all. Books straight into the calendar.
Fixed scope. Schematic and layout against manufacturability, signal integrity and safety intent — findings ranked by what they actually let through.
We read your existing analysis and tell you what an auditor will find missing, before the certification body arrives.
Stackup, impedance targets, topology and termination settled before routing — the cheapest point at which any of it is still free to change.
SIwave and HyperLynx on the board you actually have, with the PDN and eye closure written up.
Fixed scope. Does the proposed silicon or system meet its performance, power and area targets — and if not, where is the leak?
Documentation and interviews. Architecture soundness and execution risk for an investment decision.
All engagements by negotiation
Scope, rate, and terms tailored to the project. Short reviews, multi-month builds, fractional engineering — all on the table.