FAQ

Frequently asked questions about SoftQCOS and quantum computing.

General Questions

What is SoftQCOS?

SoftQCOS (Quantum Circuit Optimization Service) is a cloud platform for quantum computing that provides:

• Circuit Optimization - Reduce circuit depth and gate count for better hardware performance
• Multi-Backend Execution - Run circuits on simulators and real quantum hardware
• Cost Estimation - Know execution costs before running on QPUs
• Azure Quantum Integration - Access IonQ, Quantinuum, Rigetti, and PASQAL

Do I need a quantum computer to use SoftQCOS?

No! SoftQCOS provides access to:

1. High-performance simulators (free tier available)
2. Real quantum hardware through Azure Quantum (paid)

You can develop and test entirely on simulators, then deploy to real hardware when ready.

What programming languages are supported?

• Python - Full SDK (softqcos-sdk)
• Command Line - CLI tool (softqcos)
• REST API - Any language with HTTP support

What quantum frameworks are compatible?

SoftQCOS accepts circuits from:

• OpenQASM 2.0 - Universal standard
• Qiskit - IBM's framework (auto-converts to QASM)
• Cirq - Google's framework
• PennyLane - Xanadu's framework
• Amazon Braket - AWS's format

# Example: Convert Qiskit circuit
from qiskit import QuantumCircuit
from softqcos_sdk import QCOSClient

qc = QuantumCircuit(2, 2)
qc.h(0)
qc.cx(0, 1)
qc.measure_all()

client = QCOSClient()
result = client.execute(qasm=qc.qasm(), shots=1024)

---

Account & Billing

How do I get an API key?

1. Sign up at portal.softquantus.com
2. Navigate to Settings → API Keys
3. Click Create API Key
4. Copy and store securely (shown only once)

What's included in the free tier?

Feature	Free Tier	Paid Plans
Simulator shots/month	100,000	Unlimited
Circuit optimization	1,000/month	Unlimited
API requests/day	1,000	Unlimited
Azure Quantum	❌	✅
Priority support	❌	✅

How is Azure Quantum usage billed?

Azure Quantum costs are passed through at cost plus a small processing fee:

• IonQ: Per-shot pricing (~€0.01-0.03 per shot)
• Quantinuum: Per-HQC (Hybrid Quantum Credits)
• Rigetti: Per-shot pricing

Always use estimate_cost() before executing:

estimate = client.azure_quantum.estimate_cost(
    circuit=circuit,
    target="ionq.aria-1",
    shots=100
)
print(f"Cost: €{estimate.cost_eur:.2f}")

Can I set spending limits?

Yes! In the portal:

1. Go to Settings → Billing
2. Set Monthly Budget Limit
3. Configure alerts at 50%, 80%, 100%

API-level budget control:

client = QCOSClient(
    api_key="...",
    max_cost_per_job=5.0,  # EUR
    monthly_budget=100.0   # EUR
)

---

Technical Questions

What's the maximum circuit size?

Backend	Max Qubits	Max Depth	Max Gates
Simulator (Aer)	32	Unlimited	Unlimited
IonQ Aria	25	~500	~1000
Quantinuum H1	20	~500	~500
Rigetti Ankaa	84	~100	~500

For larger circuits, consider:

• Circuit cutting
• Tensor network simulation
• Hybrid classical-quantum algorithms

How do I improve circuit fidelity?

1. Optimize the circuit:

optimized = client.optimize(
    circuit=circuit,
    target="ionq",
    objective="max_fidelity"
)

2. Use fewer shots on QPU (reduces queue time and cost)

3. Choose the right backend:

   • IonQ/Quantinuum for high-fidelity needs
   • Rigetti for speed and larger circuits

4. Apply error mitigation:

result = client.execute(
    qasm=circuit,
    shots=1024,
    error_mitigation="zne"  # Zero-noise extrapolation
)

Why are my results different from the simulator?

Real quantum hardware has noise. Expected differences:

Source	Simulator	Real QPU
Gate errors	0%	0.1-1%
Readout errors	0%	0.5-5%
Decoherence	None	Present

This is normal! Use error mitigation or increase shots for better statistics.

Can I run circuits asynchronously?

Yes! For long-running jobs:

from softqcos_sdk import AsyncQCOSClient
import asyncio

async def run_async():
    client = AsyncQCOSClient()

    # Submit without waiting
    job = await client.submit(qasm=circuit, shots=1024)
    print(f"Job ID: {job.job_id}")

    # Do other work...

    # Check later
    result = await client.wait_for_result(job.job_id, timeout=300)
    return result

asyncio.run(run_async())

How do I handle rate limits?

The SDK handles rate limiting automatically with exponential backoff.

For custom handling:

from softqcos_sdk import QCOSClient, RateLimitError
import time

client = QCOSClient()

for circuit in circuits:
    try:
        result = client.execute(qasm=circuit, shots=1024)
    except RateLimitError as e:
        print(f"Rate limited. Waiting {e.retry_after}s...")
        time.sleep(e.retry_after)
        result = client.execute(qasm=circuit, shots=1024)

Rate limits by tier:

Tier	Requests/second	Requests/day
Free	10	1,000
Starter	50	10,000
Professional	100	Unlimited
Enterprise	Custom	Unlimited

---

Optimization Questions

What optimization objectives are available?

Objective	Description	Best For
max_fidelity	Maximize circuit fidelity	Noisy hardware
min_depth	Minimize circuit depth	Decoherence-limited
min_gates	Minimize gate count	Cost optimization
min_cnots	Minimize two-qubit gates	Hardware-limited

optimized = client.optimize(
    circuit=circuit,
    objective="max_fidelity",
    target="ionq"
)

How much does optimization help?

Typical improvements:

Circuit Type	Depth Reduction	Gate Reduction	Fidelity Gain
Random	20-40%	15-30%	5-15%
VQE Ansatz	30-50%	20-40%	10-25%
QFT	40-60%	30-50%	15-30%
Grover	25-35%	20-30%	8-15%

Does optimization change circuit behavior?

No! Optimization preserves the mathematical transformation of the circuit. The optimized circuit produces identical results (on a perfect quantum computer) to the original.

---

Security Questions

Is my code/data secure?

Yes. SoftQCOS security includes:

• TLS 1.3 encryption in transit
• AES-256 encryption at rest
• SOC 2 Type II certified
• GDPR compliant
• Circuits are never stored permanently (configurable retention)

Can I use SoftQCOS for proprietary algorithms?

Absolutely. Your circuits are:

• Encrypted during transmission
• Processed in isolated environments
• Deleted after execution (by default)

Enterprise customers can configure:

• Private deployment
• Custom data retention
• Audit logging

Where is my data processed?

Default: EU (West Europe - Amsterdam)

Available regions:

• EU West (Amsterdam)
• EU North (Dublin)
• US East (Virginia)
• US West (California)

Enterprise: On-premises deployment available

---

Azure Quantum Questions

Which Azure Quantum providers are available?

Provider	Targets	Technology
IonQ	ionq.simulator, ionq.aria-1, ionq.forte-1	Trapped Ion
Quantinuum	quantinuum.sim.h1-1e, quantinuum.qpu.h1-1	Trapped Ion
Rigetti	rigetti.sim.qvm, rigetti.qpu.ankaa-2	Superconducting

Do I need an Azure subscription?

No! SoftQCOS manages Azure Quantum access for you. Your SoftQCOS credits work directly.

Enterprise customers can connect their own Azure subscription for:

• Dedicated quotas
• Reserved capacity
• Custom billing

How long do QPU jobs take?

Typical queue times:

Provider	Queue Time	Execution Time
IonQ Aria	5-30 min	1-5 min
Quantinuum H1	30-120 min	5-15 min
Rigetti Ankaa	5-15 min	< 1 min

During high-demand periods, queues may be longer. Simulators are instant.

---

Troubleshooting

"Authentication failed" error

# Check your API key
import os
print(os.environ.get("SOFTQCOS_API_KEY", "Not set"))

# Or explicitly set it
client = QCOSClient(api_key="your-api-key")

Common causes:

• API key not set or typo
• Key revoked or expired
• Wrong environment (test vs production)

"Circuit too large" error

Your circuit exceeds backend limits. Options:

1. Use a different backend with more qubits
2. Optimize the circuit first
3. Use circuit cutting for very large circuits

# Check circuit properties
analysis = client.analyze(circuit)
print(f"Qubits: {analysis.num_qubits}")
print(f"Depth: {analysis.depth}")

"Quota exceeded" error

You've reached your plan limits. Options:

1. Wait for quota reset (monthly)
2. Upgrade your plan
3. Contact support for temporary increase

# Check current usage
usage = client.usage.get_summary()
print(f"Shots used: {usage.shots_used}/{usage.shots_limit}")

Jobs stuck in "queued" status

For Azure Quantum jobs:

1. Check Azure Quantum status: status.azure.com
2. Some providers have maintenance windows
3. High demand causes longer queues

You can cancel and retry:

client.azure_quantum.cancel_job(job_id)

---

Getting Help

Documentation

• Getting Started
• SDK Reference
• API Reference
• Examples

Support Channels

Channel	Response Time	Best For
Documentation	Instant	Self-service
Community Discord	Hours	General questions
Email: support@softquantus.com	24h	Account issues
Enterprise Slack	< 1h	Urgent issues

Reporting Bugs

Please include:

1. SDK version (pip show softqcos-sdk)
2. Python version
3. Full error message
4. Minimal reproduction code

Submit via:

• GitHub Issues: github.com/softquantus/softqcos-sdk
• Email: bugs@softquantus.com