gszczesz
Full Member level 2
With the new Cadence token system, an individual now can use almost any of Cadence's advanced features as long as there are free tokens. It also means a user can easily suck up many tokens and leave everyone else waiting...
Is there a way add priority levels to the tokens, such that a token-hogging feature such as APS simulation (8 tokens) gets stalled if a Spectre simulation using 1 token is requested? Maybe Cadence can't do real-time suspention of running simulations, but maybe restrict big simulations when running out of tokens?
So optimal solution would halt low priority simulations and allow higher priority ones to go through.
Non-optimally would be to dis-allow low-priority simulations to run when a certain threshold of free tokens is exceeded.
If Cadence doesn't support this, is there a 3rd part software that does? Since they use Flexlm, you'd think there would be some generic Flexlm priority-control module...
Greg
Is there a way add priority levels to the tokens, such that a token-hogging feature such as APS simulation (8 tokens) gets stalled if a Spectre simulation using 1 token is requested? Maybe Cadence can't do real-time suspention of running simulations, but maybe restrict big simulations when running out of tokens?
So optimal solution would halt low priority simulations and allow higher priority ones to go through.
Non-optimally would be to dis-allow low-priority simulations to run when a certain threshold of free tokens is exceeded.
If Cadence doesn't support this, is there a 3rd part software that does? Since they use Flexlm, you'd think there would be some generic Flexlm priority-control module...
Greg