Cost/When am I charged

[ronjart]

So I understand the basic billing principles.

My question is, when does the charging-timer start and stop? More precisely, is there a way to only get charged when there is an active connection (RDP or ORBX)? It seems as if I have to manually stop or terminate the ec2 instance through the AWS console, which is really cumbersome.

Thank you for your help.

[kevinshane]

same question here also i would like to know that if there is an option to use more gpu to compute(say 4-8 Titans per instance??or at least 2 Titan ??? )

[ronjart]

I think for now its only one gpu per instance, bit future updates will allow for multiple gpus (i think up to 128) to run in parrallell.

[Goldorak]

The costs and GPU config for these AMIs are set by what Amazon EC2 charges for using the G2 instance type in your region. Amazon charges per hour, rounded up to the nearest hour AFAIK.

If you are an end user looking to use Octane on the AWS cloud, best to wait for the Octane Cloud user service (still in testing), which will help reduce costs by buying GPU time in volume for our entire user base. Right now, we are only supporting one GPU per VM , but the plan is to enable rendering across any number of GPUs as a next step.

[ronjart]

Thanks for your reply. My question was if there is a way to manage usage time independantly from the aws console i.e.using orbx to start and stop usage.

My concern is, that my instance is going to be idle most of the time and ill have to pay for 24/7 use anyways unless i manually start and stop the inance through the aws console. So approx. 500$ per month seems unreasonably high cost.

I understand if this is maybe the wrong place for this question. I can always contact the aws support since they rund the ec2 billing anyways.

[Goldorak]

Thanks for your reply. My question was if there is a way to manage usage time independantly from the aws console i.e.using orbx to start and stop usage.

My concern is, that my instance is going to be idle most of the time and ill have to pay for 24/7 use anyways unless i manually start and stop the inance through the aws console. So approx. 500$ per month seems unreasonably high cost.

I understand if this is maybe the wrong place for this question. I can always contact the aws support since they rund the ec2 billing anyways.

Correct - this is not so much an OTOY issue as an AWS one, as they handle EC2 billing and instance management for our AMIs. You can definitely start/stop an EC2 instance through AWS not get charged beyond that one hour.

If you are not an AWS developer, and are primarily looking to get a Cloud PC for use with Octane Render or other apps, our recommendation is to wait for the Octane Cloud service. This will handle all the billing and instance load balancing for you, and expose render/compute credits for efficiently leveraging multiple cloud GPUs beyond just the ones you need for the baseline Windows/Linux applications.

[ronjart]

Thanks, that makes sense.

I use octane, but for now my workstation seems to offer a better value.

I'm just experimenting with some other real time applications running over orbx and found that the otoy autodesk ami seems to work great for that.

[Goldorak]

Thanks, that makes sense.

I use octane, but for now my workstation seems to offer a better value.

I'm just experimenting with some other real time applications running over orbx and found that the otoy autodesk ami seems to work great for that.

Very glad to hear that you have had a good experience using the AMI.

Cloud rendering with Octane is ultimately designed to expose the power of hundreds of concurrent GPUs, for offline or even real time rendering (depending on the user's needs and budget). Getting the host apps in the cloud, as we're doing now, is a requirement for the latter service. But it also opens up new possibilities by removing hardware and OS requirements on the user's end.

Regarding ORBX, beyond live streaming, future iterations are planned to work as an offline file format that Octane can render into (likely exposed first via lua api after 1.5). This would be an alternative to lossless image/video formats . ORBX video will efficiently support encoding info channels (depth/normals/UV,...), multiview, HDR multi-light passes, floating point precision per channel, etc. - which H.264 and HEVC are not good at handling. And it will still decode in JS with no plug-ins

[ronjart]

Really exciting stuff.

Couple of notes:

My Laptops screen resolution is 1600x900 (Razor Blade 14 inch) - I can change the resolution of my virtual desktop over ORBX, but when I set it to 1600x900, the connection freezes up. Any idea why that could be? I would really like to run in full screen without borders. But without being able to match resolution and the FULLSCREEN= command not working I seem to be out of options.

After streaming for about 15 minutes, my laptop started to really overheat. I never had that happen, even on heavy load (Running real time GPU intensive applications) I'm assuming it's my CPU overheating since that is where the decoding happens? CPU on my laptop is pretty decent (i7 4702HQ @ 2.2ghz). Any idea why that could be? Here is my launch URL:

http://aws.otoy.com/?connect=//xxx:8090 ... HEIGHT=720

Any idea when mobile browsers will be supported (Right now I just get a greenish looking screen,that works half way)? Will the ORBX API expand on user input (touch, rawinput etc.)? What is the roadmap for this?

[andrear]

hi,
just to say that I've receved a notification of a 390$ charge on my bill for a over 400 hrs of use of the service.
it was stated it was free until the end of the month and the whole procedure is so weird that is impossible to understand how to stop the service, nether if it is required to stop it or not.
I've stopped the payment from my bank and sent a feedback to amazon announcing it.

best regards,
andrea