About Intrinsity
Intrinsity provides the tools, technologies, and expertise so that our customers can efficiently and predictably produce the fastest digital logic. The key to achieving this is our proprietary Fast14® technology , which provides not only circuit speeds of up to and beyond 3 GHz, but also the means to trade off speed, power and area to achieve the optimal solution to customer design targets.
Intrinsity FastCores™
Intrinsity FastCores are versions of popular embedded IP processor cores built using Intrinsity’s proprietary Fast14 1-of-N Domino Logic (NDL) circuit technology. Cycle-accurate RTL FastCores built with Fast14 NDL can typically run 2x or more as fast as ordinary synthesized static logic versions, and use little to no more silicon area or power per MHz, once Intrinsity's high-speed memories are incorporated. Additionally, FastCores consume significantly less power than synthesized static versions at speeds that "push" conventional logic. FastCores are built using the same silicon processes at the same foundries as slower versions built using ordinary standard cell methodologies.
Here are some typical examples of what cycle-accurate RTL FastCores can do:
| CORE/METHOD | CLOCK SPEED | POWER | AREA (core only) | DMIPS |
| Cortex-R4 (SVt)/synth. | 312 MHz | 145 mW | 0.84 mm2 | 500 |
| Cortex-R4 (LVt)/synth. | 400 MHz | 279 mW | 640 | |
| Cortex-R4X (SVt)/Fast14 | 400 MHz | 181 mW | 0.95 mm2 | 640 |
| Cortex-R4X (SVt)/Fast14 | 600 MHz | 340 mW | 960 |
Note that these results are for the core only - when ordinary cores are coupled with compiled RAMs, and FastCores with Intrinsity FastRAMs™, the combined high-speed FastCore power and area characteristics improve even further.
Intrinsity FastCores may be constructed in one of two ways - the first, called RTL FastCores, refer to FastCores built from the synthesizable RTL code embedded IP vendors typically supply their customers (to build the cores). RTL FastCores are cycle-accurate, meaning that they behave in exactly the same manner, clock cycle-for-cycle, as slower synthesized versions, and can run any code written for the original version unmodified. A good example of an RTL FastCore is the ARM/Intrinsity Cortex-R4X embedded IP core.
The speed of Fast14 circuit technology can allow even greater speedup ratios if the design is not constrained by cycle accuracy. When Intrinsity is licensed to build a core from an instruction-set architecture (ISA) description of a processor, instruction-set compatible (but not necessarily cycle accurate) ISA FastCores can be built that often provide 3x or more the performance of common synthesized standard cell implementations. A good example of an ISA FastCore is the AMCC/Intrinsity Titan core.