New NVIDIA Open-Source Linux Kernel Graphics Driver Appears

[Jabberwocky]

He's not going to be doing with any of the real design work. When you're that far up the ladder, you're far to busy to take any of it on, and you don't have time to keep up to date either. There's virtually no risk from someone that far removed from the front lines A) learning about proprietary information that's still relevant and then B) somehow communicating it all the way back down the ladder to the guy doing the circuit layout.

This is my perspective.

Doing real design work is not that relevant especially considering the pace of evolution in the field. Media encoding/decoding is maybe a single exception. I'll repeat myself: So you have to an engineer to know what the engineers are doing?

The IC industry has changed a lot over the past ~decade. Highlevel CPU designs can change and adapt quickly, thanks to the investment into programmatic designs and sophisticated tooling. Take a look at the difference between Zen(+) and Zen2... The designs are worlds apart. Again, yes a manager typically won't design low level components, however this is where I disagree: A good manager would know and understand the advantages and disadvantages of many different designs and design configurations.

You don't even need an engineering background to obtain valuable trade secrets, high-school math/science and some experience is enough. A low level trade secret like branch prediction training is probably not something a manager (without an engineering background) would completely understand. Additionally you can't just copy the results directly from one design to another... On the other hand high-level data-centrality (e.g. Infinity Cache) practical R&D costs an insane amount of resources. I won't not be surprised if it costs more than something as challenging as branch predictors (which is a lot cheaper to validate). Data-centrality is a beast that spans over multiple fabs and packaging tech. Almost anyone can understand the results once you have found what works well.

[Developer12]

This is my perspective.

Doing real design work is not that relevant especially considering the pace of evolution in the field. Media encoding/decoding is maybe a single exception. I'll repeat myself: So you have to an engineer to know what the engineers are doing?

The IC industry has changed a lot over the past ~decade. Highlevel CPU designs can change and adapt quickly, thanks to the investment into programmatic designs and sophisticated tooling. Take a look at the difference between Zen(+) and Zen2... The designs are worlds apart. Again, yes a manager typically won't design low level components, however this is where I disagree: A good manager would know and understand the advantages and disadvantages of many different designs and design configurations.

You don't even need an engineering background to obtain valuable trade secrets, high-school math/science and some experience is enough. A low level trade secret like branch prediction training is probably not something a manager (without an engineering background) would completely understand. Additionally you can't just copy the results directly from one design to another... On the other hand high-level data-centrality (e.g. Infinity Cache) practical R&D costs an insane amount of resources. I won't not be surprised if it costs more than something as challenging as branch predictors (which is a lot cheaper to validate). Data-centrality is a beast that spans over multiple fabs and packaging tech. Almost anyone can understand the results once you have found what works well.

You're arguing hypotheticals and imagination to someone who actually worked in one of the two tier 1 silicon firms. I had lunch with their VP of engineering (guy was quite a fan of cars), and the senior management below that, and the multiple senior engineers below that, and then my actual colleagues.

Anyone above the senior engineers was firmly in management and was waaaaay too busy managing teams and allocating resources to know more than glancing information about what anyone was actually building. They had last touched circuitry 20 years ago, which is eons in semiconductor design. Anyone above that? Add another 20 years for each level of management.

Not only did they have no idea what was actually going on on the design side, they *never* suggested design ideas or otherwise influenced how the engineers below them solved problems. They were simply A) not knowledgeable enough anymore, and B) it wasn't their job. They kept the money flowing and the project on schedule, not electrons and clock signals.

[EmmaJaneBonestell]

This. Right here. This shows you didn't read the comment you were quoting at all.

You may have missed the word "tainted." If you read the leaked source code you are tainted. You have ideas in your head that came from leaked source and do not belong to you. Any code you write for nouveau is arguably a derivative work, and argue nvidia's lawyers will. Any contribution you give without their approval is flatly and completely illegal.

One very likely cannot write the code itself (and most of such-projects don't allow thd following), but one may use the Chinese-wall style (two teams, one disassembles & documents, the other implements from the documentation) of clean-room design to make contributions, rather than black-box testing; so, your claim that "any contribution... is completely illegal," is objectively false. And in an attempt to dissuade claims of licensing issues making the disassembly/RE illegal, not all courts/states/countries accept idiotic click-wrap agreements as legally binding.

So maybe try not to spread misinformation with such condescending authority in the future.