Thank you for putting all this information together in one place.
Must have been a long day / night.For me this was the information which answered my questions from yesterday:
Full RTX optimizations and NV Link support are experimental today, and planned for Enterprise annual subscriptions which $20 /Month ( $15/ Month through Black Friday) but require that you have first made a one time purchase of OctaneRender Standalone+DCC Enterprise (either Combo or All-access bundle) from the purchase page. The All Access Bundle comes with 1-3 years included.
and
You can still use Octane 4 the same way you do now without an annual pass, and any permanent licences you have purchased will work and are yours forever. Any USB Dongles activated for annual pass can be reverted to V4.
After the term of your annual all-access pass, you won’t have access to the extra plugins, render slaves and version updates that came with your annual pass.
However you can re-activate an annual pass later at any time.
I was worried that the all-access $20 /Month pass would be a one time limited offer for existing permanent license owners. It was not clear to me if after the offer expired users would have to "rent" an OctaneRender® Studio Enterprise Subscription for $539 /Yr. (BF discounted price) every year.
For 2018.1 and later, we may need to rethink the 2 GPU limit as a differentiation from the paid tiers . it is meaningless when two GPUs give you 800-1000 OB and 22 GB to 50 GB of combined VRAM:
What kind of new users should be able to use the free tier? (students, hobbyists?)
What kind of new users should be able to use the Studio license? (free lancers, small studios?)
What kind of new users would need the advanced features of an Enterprise license? ( Large studios on a Hollywood production level?)
Is it intended that students or freelancers who spent a large sum of their available allowance on two RTX 2080 Ti need to purchase an Enterprise tier license first to be able to use Tensor cores or NV Link?
In a few years Tensor Cores and NV Link may be just a normal component of a GPU as Cuda is now.
- - -
Please delete this feedback if it disrupts the intended purpose of this thread.