New core combines KiviCore research and development expertise with CAST quality and usability standards for a better PQC IP experience
Jan 16, 2025
Silicon IP Cores
SNOW-V
SNOW-V Stream Cipher Engine
The SNOW-V IP core implements the SNOW-V stream cipher mechanism, aiming to meet the security demands of modern high-speed communication systems. It conforms to the official SNOW-V mechanism, published in 2019 by the IACR Transactions on Symmetric Cryptology, as an extensive revision of SNOW 3G stream cipher.
Receiving a 256-bit Key and a 128-bit Initialization Vector (IV), the core processes 128 bits of information in one cycle and it produces a stream of 128-bit keys. It employs two main building blocks, a Linear Feedback Shift Register (LFSR) and a Finite State Machine (FSM) that applies an Advanced Encryption Standard (AES) round function.
The core can be easily incorporated in a Galois/Counter mode (GCM) topology and by interoperating with a Galois Message Authentication Code (GMAC) realize an Authenticated Encryption with Associated Data (AEAD) mechanism. What is more, the core is a drop-in replacement for SNOW 3G in EPS Encryption/Integrity Algorithm (EEA/EIA) architectures and New Radio Encryption/Integrity Algorithm (NEA/NIA) architectures for 4G and 5G communications, while also targeting future mobile network generations (e.g. 6G).
The SNOW-V IP core is a microcode-free and fully synchronous design developed for reuse in ASIC and FPGA implementations, aiming at throughput-demanding environments. The efficient and compact hardware design enables notably high throughput, achieving over 100 Gbps in mainstream ASIC technologies and over 40 Gbps in high-end AMD and Altera FPGAs. Moreover, being a scan-ready, LINT-clean, and single-clock design with a simple handshake interface, facilitates straightforward integration
The SNOW-V IP core can help provide fast and secure communication in high-speed 4G, 5G, and future 6G networks. Furthermore, it can safeguard device interactions and prevent cyber threats in IoT ecosystems, while for multimedia streaming, it will enable real-time encryption to protect content from unauthorized access. Additionally, the SNOW-V IP core may be used for data storage encryption and high-quality pseudo-random number generation.
The SNOW-V core can be mapped to any ASIC technology. The following are sample ASIC pre-layout results reported from synthesis with a silicon vendor design kit under typical conditions, with all core I/Os assumed to be routed on-chip.
|
Technology |
Logic |
Memory |
Frequency |
Throughput |
|---|---|---|---|---|
|
Samsung 5nm |
36.8k eq. gates |
0 bits |
1,000 MHz |
128 Gbps |
|
TSMC 7nm |
32.2k eq. gates |
0 bits |
1,100 MHz |
140.8 Gbps |
|
TSMC 16nm |
34.1k eq. gates |
0 bits |
1,000 MHz |
128 Gbps |
|
TSMC 28nm HPC |
31.4k eq. gates |
0 bits |
650 MHz |
83.2 Gbps |
The provided figures do not represent the highest speed or smallest area for the core. Please contact CAST to get characterization data for your target configuration and technology.
The SNOW-V core can be mapped to any Altera FPGA device. The following are sample Altera FPGA pre-layout results reported from synthesis with a silicon vendor design kit under typical conditions, with all core I/Os assumed to be routed on-chip.
|
Technology |
Logic |
Memory |
Frequency |
Throughput |
|---|---|---|---|---|
|
Agilex 7 (A2) |
1,190 ALMs |
16 RAMBs |
350 MHz |
44.8 Gbps |
|
Arria 10 GX (1) |
963 ALMs |
16 RAMBs |
325 MHz |
41.6 Gbps |
|
Cyclone V GX (C6) |
1,132 ALMs |
16 RAMBs |
150 MHz |
19.2 Gbps |
|
Max10 (7) |
2,239 LEs |
16 RAMBs |
125 MHz |
16 Gbps |
|
Stratix V (R1) |
1,050 ALMs |
16 RAMBs |
350 MHz |
44.8 Gbps |
The provided figures do not represent the highest speed or smallest area for the core. Please contact CAST to get characterization data for your target configuration and technology.
The SNOW-V core can be mapped to any AMD FPGA device. The following are sample AMD FPGA pre-layout results reported from synthesis with a silicon vendor design kit under typical conditions, with all core I/Os assumed to be routed on-chip.
|
Technology |
Logic |
Memory |
Frequency |
Throughput |
|---|---|---|---|---|
|
Kintex 7 (-3) |
1,242 LUTs |
16 RAMB18 |
300 MHz |
38.4 Gbps |
|
Virtex 7 (-3) |
1,246 LUTs |
16 RAMB18 |
250 MHz |
32 Gbps |
|
Kintex US (-3) |
1,236 LUTs |
16 RAMB18 |
400 MHz |
51.2 Gbps |
|
Kintex US+ (-3) |
1,253 LUTs |
16 RAMB18 |
540 MHz |
69 Gbps |
|
Versal AI (-2) |
1,253 LUTs |
16 RAMB18 |
350 MHz |
44.8 Gbps |
|
Versal Prem (-2) |
1,253 LUTs |
16 RAMB18 |
450 MHz |
57.6 Gbps |
|
Zynq US+ (-1) |
1,232 LUTs |
16 RAMB18 |
400 MHz |
51.2 Gbps |
The provided figures do not represent the highest speed or smallest area for the core. Please contact CAST to get characterization data for your target configuration and technology.
The core as delivered is warranted against defects for ninety days from purchase. Thirty days of phone and email technical support are included, starting with the first interaction. Additional maintenance and support options are available.
The core is available in RTL (VHDL or Verilog) source code or as a targeted FPGA netlist. Its deliverable package includes the following:
- Self-checking HDL testbench
- C Model & test vector generator
- Sample simulation & synthesis scripts
- User documentation
This core implements encryption functions and as such it is subject to export control regulations. Export to your country may or may not require a special export license. Please contact CAST to determine what applies in your specific case.
Engineered by Ocean Logic.
Features List
Security Mechanism Support
- SNOW-V stream cipher
- IACR 2019 publication
- AEAD-mode ready
- Easy adoption to a GCM core
- Seamless interoperation with a GMAC block
- Drop-in replacement of SNOW 3G in 4G/5G security architectures
Target Applications
- High-speed 4G/5G and future (6G) mobile communication
- IoT device communication
- Secure multimedia streaming
- High-speed encrypted data storage
- Pseudo-random number generation
- General data encryption
High-throughput and Compact Design
- Processing 128 bits/cycle
- Over 100 Gbps and less than 37k eq. gates in mainstream ASIC technologies
Easy Integration and Technology Mapping
- Simple handshake interface
- Fully synchronous, single-clock domain, re-usable design
- No false or multicycle timing paths, scan-ready, LINT-clean
Deliverables
- RTL source code (VHDL or Verilog) or targeted FPGA netlist
- Complete testbenches
- C model and test-vector generator
- Simulation and synthesis scripts
- Documentation