Hardware-accelerated post-quantum cryptography. Run live operations in your browser, compare against real QUAC-100 hardware benchmarks.
1.2M+ML-KEM-512 ops/sec
≤700nsFull KEM cycle
16Parallel NTT engines
>800 MbpsQRNG conditioned output
Interactive
Cryptographic Operations
ML-KEM Key ExchangeFIPS 203
$ quac kem --await-input
ML-DSA Digital SignaturesFIPS 204
$ quac dsa --await-input
Quantum Random Number GeneratorSP 800-90B
$ quac qrng --await-input
Hybrid / Hash OperationsTLS 1.3
$ quac hybrid --await-input
Benchmark
Software vs. QUAC-100 Hardware
How it works: Runs real cryptographic operations in your browser, then shows what QUAC-100 hardware achieves. The gap demonstrates why hardware acceleration matters for PQC at scale.
$ quac bench --await-input
Throughput
Hardware Performance
SW ML-KEM-768
~15K ops/s
HW ML-KEM-768
800K ops/s
HW ML-KEM-512
1.2M ops/s
HW ML-DSA-44 Sign
400K ops/s
HW ML-DSA-44 Verify
900K ops/s
HW AES-256-GCM
16 Gbps
HW SHA3-256
20 Gbps
Hardware
QUAC-100 Platform Specifications
1.2M+ML-KEM-512 ops/sec
≤700nsFull KEM cycle
16NTT engines
32 GBHBM2e memory
819 GB/sMemory bandwidth
190WPeak TDP
>800 MbpsQRNG conditioned
PCIe x8x8Gen5 interface
Comparison
QUAC-100 vs. CPU (Intel i7-11700 AVX2)
Operation
CPU (AVX2)
QUAC-100
Speedup
ML-KEM-512 KeyGen
~8.7 µs
180 ns
48x
ML-KEM-512 Encaps
~10.8 µs
220 ns
49x
ML-KEM-512 Decaps
~13.4 µs
280 ns
48x
Full KEM Cycle
~32.9 µs
≤700 ns
~47x
ML-DSA-44 Sign
~40 µs
850 ns
47x
ML-DSA-44 Verify
~12.3 µs
320 ns
38x
Benchmarks from AMD Versal HBM production platform. CPU baseline: Intel Core i7-11700 @ 3.9GHz with AVX2.
Capacity
Single QUAC-100 Card
1.2M/secTLS Key Exchanges ML-KEM-512 sustained
400K/secDocument Signatures ML-DSA-44 signing
>800 MbpsQuantum Entropy SP 800-90B conditioned
Architecture
Three Capabilities. One Card.
PQC Acceleration
16 parallel NTT engines with Radix-32 architecture deliver 1.2M+ ML-KEM ops/sec at sub-microsecond latency.
ML-KEM (FIPS 203)
ML-DSA (FIPS 204)
SLH-DSA (FIPS 205)
Hybrid classical/PQC
HSM Functionality
Hardware security module with tamper-resistant key storage and cryptographic boundary protection.
Tamper-evident enclosure
Secure key storage
FIPS 140-3 Level 3 (IUT)
Role-based access control
Quantum RNG
True quantum entropy source eliminates reliance on algorithmic pseudo-random generators.
True quantum entropy
>800 Mbps conditioned output
NIST SP 800-90B compliant
Continuous health monitoring
SDK
Integration Examples
PythonC/C++RustOpenSSL
from quantacore import KEM, DSA, QRNG
# Key Encapsulation (ML-KEM-768)
kem = KEM("ML-KEM-768")
pk, sk = kem.keygen()
ct, shared_secret = kem.encapsulate(pk)
# Digital Signatures (ML-DSA-65)
dsa = DSA("ML-DSA-65")
pk, sk = dsa.keygen()
sig = dsa.sign(b"Contract v2.1", sk)
valid = dsa.verify(b"Contract v2.1", sig, pk)
# Quantum Random Numbers
entropy = QRNG.generate(32)
use quantacore::{Kem, Dsa, SecurityLevel};
let kem = Kem::new(SecurityLevel::L3)?;
let (pk, sk) = kem.keygen()?;
let (ct, ss) = kem.encapsulate(&pk)?;
let dsa = Dsa::new(SecurityLevel::L3)?;
let (pk, sk) = dsa.keygen()?;
let sig = dsa.sign(b"message", &sk)?;
# openssl.cnf provider configuration
[provider_sect]
quac100 = quac100_sect
[quac100_sect]
module = /usr/lib/ossl-modules/quac100.so
activate = 1
# Start PQ-TLS server
$ openssl s_server -cert server.pem -key server.key \
-groups kyber768 -port 443
# PQC acceleration is automatic when QUAC-100 is present