Silex Insight achieves 100Gbps in FPGA for ChaCha20-Poly1305 IP core - Embedded.com

Silex Insight achieves 100Gbps in FPGA for ChaCha20-Poly1305 IP core

Crypto coprocessor IP core provides authenticated encryption between highly networked applications, for example in IoT or cloud computing, using Chacha20 stream cipher combined with the Poly1305 message authentication code (MAC).

Silex Insight has said it has achieved record-breaking speed for its ChaCha20-Poly1305 hardware crypto engine, managing 800Gbps in ASIC and 100Gbps in an FPGA. Its’ RFC7539 compliant intellectual property (IP) core, available for ASIC and FPGA designs, offers customers future-proof authenticated encryption between highly networked applications, for example in IoT or cloud computing.

The crypto coprocessor IP core provides authenticated encryption with associated data (AEAD) using the Chacha20 stream cipher combined with the Poly1305 message authentication code (MAC). Its maximum throughput is 800Gbps with ASICs and 100Gbps with FPGAs, and it can be fully configured to match the specific throughput requirements of the application and hardware platform. For system-on-chip (SoC) solutions, the core can be combined with optional scatter/gather DMA for easy and efficient integration. Alternatively, this IP can also be deployed for encryption or authentication separately.

Silix Insight ChaCha20-Booster

The ChaCha20-Poly1305 block may be combined with the AES-GCM multi-booster, which implements a high-throughput AES-GCM crypto engine. This combination offers future-proof security, as Chacha20-Poly1305 is a mandatory AEAD, as is the AES-GCM, in TLS 1.3. It is also a more secure alternative to older algorithms in TLS/SSL, the underlying protocol that secures most internet communication in, for example, browsers (https), voice-over-IP, mobile devices and IoT applications.

According to Sebastien Rabou, security division director at Silex Insight, the company’s ChaCha20-Poly1305 IP core allows its customers to both create dedicated ASIC solutions as well as create extreme high performance FPGA accelerators on existing cloud platforms like AWS or others. He said it would benefit customers not only when it comes to speed, but also in bringing down the cost of ownership of their datacenter as more and more of the network and security processing can be offloaded to fast hardware.


Related Contents:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.