editor's blog
Subscribe Now

Dude, That’s So Random

Many of us grew up knowing that you can’t have truly random numbers in your algorithms or circuits. The best you can do is “pseudo-random,” which means take a non-random number and shake it up really hard so it looks random. But… if you start with the same number every time, then you’ll end up with the same sequence every time.

Oh yeah, that’s the other thing: pseudo-random number generators (PNRGs) are all about sequences. You seed the thing with a value, and, from then on, it supplies a stream of numbers in an unpredictable (for practical purposes) fashion like a Pez dispenser where you don’t know what color is going to come up next.

So it’s all about the seed, and it’s been very hard to produce seeds that are truly random.

Actually, there’s even a problem with the concept of “truly random”: it’s not that simple. If you have an 8-bit field that unpredictably generates one of 5 values, well, that’s kind of random. But you could be generating one of 256 values to make full use of the 8 bits. Such a scheme would be said to have low entropy; most of the 8 bits aren’t, in fact, random at all.

We looked at some aspects of hardware randomness in our article on physically unclonable functions (PUFs) a few months back. One of the companies we covered, Intrinsic ID, has just announced a new random number generator, iRNG, based on the technology they use for their PUF. It makes use of the noise inherent in the power-up of SRAM bits.

If you look at their basic drawing, they have a conditioned “true random seed” coming from their “entropy source” that feeds a DRBG, which stands for “deterministic random bitstream generator.” And if you go to Wikipedia and look that up, it takes you to the page for PRNGs. In other words, a DRBG is another name for a PRNG. And I can totally understand how they would not want the phrase “pseudo-random” showing up anywhere in marketing materials that are trying to convey a message of “true” randomness.

However, this is simply a case of the PRNG we’ve been talking about, only now fed by a seed that they say is truly random. Of course, since there’s no such thing as “truly” random, then how random is it? They claim “high entropy” (and they can probably quantify that for a given technology). But we can get a sense from thinking about the underlying mechanism.

An SRAM bit is more or less a bistable element that, in the extreme, is very well balanced so that, as it’s powering up in a perfectly noiseless environment, it would have an equal chance of toppling over into a 1 state or a 0 state. The opposite extreme can be seen in flip-flops that have been explicitly biased to come up in a predictable state by intentionally unbalancing things.

Real SRAM cells, even if designed perfectly, won’t be perfect due to process variations. Intrisic ID uses a 2-kbit memory; each of those bits will be ever so slightly different. Some will be subtly biased to come up 0, some to come up 1. Each chip will be different, so you certainly have a situation where no two systems will behave alike.

If the power-up conditions were perfect and noiseless, then each of these systems would come up the same way each time except for those bits that happened to be perfectly poised – giving you some level of non-zero entropy.

But noiseless is impossible. (At some point I assume even quantum fluctuations would play a part.) So high entropy comes in if the noise available is high enough – and unpredictable enough – that even those bits subtly biased in one direction or another can still be pushed over to the other side. Instead of a 50/50 chance of a 1 or 0, perhaps you’ve got a 70/30 chance.

So, just as the 5-value randomness didn’t make complete use of an 8-bit field above, here also, the entropy you get won’t make use of the full 2-kbit field. But if you’re getting a good percentage of those bits to be random, well, that’s still a pretty wide field.

So the key to the Intrinsic ID is this SRAM plus the conditioning; the DRBG/PRNG is designed to meet the requirements of various security standards. SRAM aside, this is available as hardware or software IP.

More info in the release

Leave a Reply

featured blogs
May 25, 2023
Register only once to get access to all Cadence on-demand webinars. Unstructured meshing can be automated for much of the mesh generation process, saving significant engineering time and cost. However, controlling numerical errors resulting from the discrete mesh requires ada...
May 24, 2023
Accelerate vision transformer models and convolutional neural networks for AI vision systems with the ARC NPX6 NPU IP, the best processor for edge AI devices. The post Designing Smarter Edge AI Devices with the Award-Winning Synopsys ARC NPX6 NPU IP appeared first on New Hor...
May 8, 2023
If you are planning on traveling to Turkey in the not-so-distant future, then I have a favor to ask....

featured video

Automatically Generate, Budget and Optimize UPF with Synopsys Verdi UPF Architect

Sponsored by Synopsys

Learn to translate a high-level power intent from CSV to a consumable UPF across a typical ASIC design flow using Verdi UPF Architect. Power Architect can focus on the efficiency of the Power Intent instead of worrying about Syntax & UPF Semantics.

Learn more about Synopsys’ Energy-Efficient SoCs Solutions

featured contest

Join the AI Generated Open-Source Silicon Design Challenge

Sponsored by Efabless

Get your AI-generated design manufactured ($9,750 value)! Enter the E-fabless open-source silicon design challenge. Use generative AI to create Verilog from natural language prompts, then implement your design using the Efabless chipIgnite platform - including an SoC template (Caravel) providing rapid chip-level integration, and an open-source RTL-to-GDS digital design flow (OpenLane). The winner gets their design manufactured by eFabless. Hurry, though - deadline is June 2!

Click here to enter!

featured chalk talk

Johnson RF Connectivity Solutions
The growing need for remote patient monitoring and wireless connectivity has made RF in medicine applications more important than ever before. In this episode of Chalk Talk, Amelia Dalton chats with Ketan Thakkar from Cinch Connectivity Solutions about the growing trends in medicine today that are encouraging the use of RF, why higher frequency, smaller form factor, cable assembly expansion and adapter expansion are vital components in today’s medical applications and why Johnson medical solutions could be a great fit for your next medical design.
Nov 28, 2022
23,177 views