More Bandwidth with IPK®
Choked with Latency
Many of the world’s data networks are “choked with Latency”. Network latency (or lag) measures the time it takes for a client device to send data to the origin server and receive a response. In other words, network latency is the delay between when the user takes action (like clicking on a link) and the moment a reply arrives from the server.
Latency is commonly caused by distance or by “work” that takes multiple compute or CPU cycles to process, thereby creating a lag before the next process can be commenced. In our data security orientated world, the overhead of encrypting data for subsequent transmission across a network is a major contributor to Latency. When latency is present, users demand more bandwidth, like wanting more lanes on your Internet superhighway so the heavy traffic can flow better.
Cause & Effect
In our current networking environments, the need for More Security is paramount. Most organizations are demanding greater levels of data security and compliance with numerous cyber-security procedures and processes. The drive for data security has an unfortunate consequence – the more encryption applied to the data, the greater the latency on these networks.
In an attempt to remedy this challenge, innovative strategies have been created – like creating dedicated encryption processors to un-burden the main system CPU and accelerate the encryption processing. The problem persists, because the CPU cycles required for the current encryption standards are the processing cycles required. AES-256-GCM requires 14 CPU cycles whether those cycles are on the main CPU or on a dedicated Encryption processor, 14 cycles are 14 cycles!
A further emerging challenge that will compound this problem is the threat to existing encryption methods of Quantum Computing. Quantum Computing is very likely to “break” or “crack the code” of the many of the encryption ciphers in use today.
In this case, it's not a “will” or “won’t” question
– it’s a “when” question!
All of these processor intensive functions, further compounded by Artificial Intelligence (AI) and the increasing uptake in usage and demand for more and more complex AI processing – all book-ended by the mandatory encryption of the answer set – means this situation will continue and possibly get worse.
IPK® saw this as an area of opportunity. In our application of the McEliece encryption protocol – designated as a Code-based approach under NIST’s Post Quantum Cipher frameworks - within the Forward Error Correction (FEC) processing, we reduced the encryption overhead to a single CPU cycle (a 98% cycle reduction at the most optimal) or with various other IPK® scenarios, a 20% compute cycle saving while double encrypting the data.
This IPK® approach dramatically reduces the latency on any network. In effect, we remove the need for the dedicated encryption processor and leverage the existing work being undertaken in the FEC processing. Zero overhead data encoding. No compute overhead. Major Energy savings. And Massive bandwidth and throughput advantages. And a Post-Quantum resistant protocol for 100% of the data.
Imagine a network where you have the double benefit of your favourite cipher for your highest level of data security and for 100% of everything else you have the confidence that 100% of data is 100% encrypted. Every bit, byte, block, KB, MB, GB, TB, EB, PB, ZB is secure, and is travelling to it’s intended destination as fast as possible. No delays. No latency.
This and other IPK® innovations will profoundly impact the availability of bandwidth across all transmission environments too! So reach out to IPK® for a discussion on how we can meet your “need for speed” across all your Networks.
