(1) eWave adopts the standard formats commonly used in the industry for input and output, which can automatically read the structure to be analyzed from a layout file and output the analysis results in various forms such as the S-parameter. eWave uses the MoM analysis method and the multi-layer Green's function used to analyze the substrate loss to solve the current or electric field distribution of on-chip structures of any shapes.
(2) ePCD can be used to optimize the design of on-chip spiral inductors, transformers, capacitors, etc. The inputs are the circuit parameters that the user wants to reach, such as the inductance value L and the quality factor Q, as well as the parameters that can be optimized and their search space
(1)The simulation speed of eWave/ePCD is close to the NlogN algorithm, which is really fast and theoretically optimal.
(2) Excellent multi-threading technology achieves a speed faster by nearly N times with N cores, much faster than similar products in the industry.
(3) Distributed computing technology can conveniently distributes frequency point calculations to multiple computing clusters, and finally synthesizes the results, with which users can seamlessly experience distributed speed improvement, without even noticing the redundancy.
(1)eWave’s world-first rectangle + triangle subdivision method takes into account both the precision and speed, with the least unknown computing amount, which not only saves memory and improves the speed, but also has higher precision than pure triangulation.
(2) A variety of usage modes are available. For example, the Port mode can perform finer subdivision of metal connected with the port, while the part not connected with the port is subdivided in a more rough way; the Cap mode can finely subdivide the upper and lower surfaces of the same layer of metal, and thus accurately simulate the capacitance between metals with very close intervals, while the same inductor does not need to be over-subdivided.
(3) Experienced users can slide the subdivision density slider and choose their own balance point in terms of precision and speed.
(1) Improve the design efficiency: Depending on the use of passive devices by the design company, the efficiency improved by using ePCD/eWave is equivalent to reduce three months to two years of workload for an RF engineer, compared with manual design.
(2) Improve the success rate of tape-out. For example, eWave/ePCD support more complete DRC/LVS checking to avoid the risks of manual design. Once there was a customer whose manually designed Tap port was not truly grounded, resulting in a tape-out failure and a loss of hundreds of thousands of US dollars.
(3) Improve the design yield. Using eWave/ePCD to perform multiple times of simulation conformation for coupling effect, component orientation and even PDK components has been proven to improve the design yield. Even if the design yield is improved by 1%, the benefits will be considerable.
(4) Reduce the cost. ePCD’s powerful synthesis (finding the optimal solution) function can reduce the design area by 10%-20% on average compared with manual design. The total area saved on chip by multiple passive devices is considerable.
(5) NineCube has a potential cooperation plan for start-ups. For details, please contact a sales representative by email to sales@ic9cube.com.
