davide445 Posted January 23, 2020 Author Share Posted January 23, 2020 The goal on my side is being competitive trough innovation, and use local available abundant funding on innovation. Since in theory its not impossible I'm currently - discussing with fiber optic connectivity provider able to provide from 7 to 63Gbps connection at 2ms latency - discussing with HPC computing provider for a 6xGPU server - discussing with multiview tech provider for the simultaneous rendering - discussing with intelligent caching and 5g provider to integrate all the content - evaluating edge fiber optic connectivity cards - evaluating wireless HMD solutions and their ultra low latency encoding / decoding solutions - discussing with power plant developer for a test The needed and available budget will decide if a test will be possible. Link to comment
demostenes Posted January 23, 2020 Share Posted January 23, 2020 3 minutes ago, davide445 said: The goal on my side is being competitive trough innovation, and use local available abundant funding on innovation. Since in theory its not impossible I'm currently - discussing with fiber optic connectivity provider able to provide from 7 to 63Gbps connection at 2ms latency - discussing with HPC computing provider for a 6xGPU server - discussing with multiview tech provider for the simultaneous rendering - discussing with intelligent caching and 5g provider to integrate all the content - evaluating edge fiber optic connectivity cards - evaluating wireless HMD solutions and their ultra low latency encoding / decoding solutions - discussing with power plant developer for a test The needed and available budget will decide if a test will be possible. You would need that 2ms latency E2E. So price will be insane, use case very limited and result everything but guaranteed. So questions is, if such solution is competitive... Link to comment
davide445 Posted January 23, 2020 Author Share Posted January 23, 2020 (edited) Needed latency for VR is 20ms E2E, and we can reach that. Also is focused on industrial applications and not gaming, so didn't need to reach 144fps for shooting first :-). For a single project my goal is have costs below a specific amount, being a first on the topic will be much than offset from the marketing potential, having also part of the costs absorbed from R&D incentives. Future projects can greatly benefit from upcoming 5G infrastructure and maturing of the technology (GPU, caching, h265 encoding, ML). Stadia, Google Now, Shadow, Paperspace are all providing game streaming solutions, VR streaming it's just the next logical evolution (apart autonomous headset such as the Quest, that can't be used in autonomous mode for industrial complex projects). Providing a service will be beneficial for the provider (centralized maintenance, scalability, cross-selling of services, etc) and the user (less capital invested, less maintenance headcount, flexibility in term of office positioning, faster support, etc). @ulf.schroeterI suppose you are expert on IG from your previous posts. CloudVR it's just a way to provide a service, such as VR is for certain applications such as simulation and maintenance. In my understanding CloudVR to become successful in the industrial use case require two elements: concurrent multiple rendering views, and integration with industrial systems to create a digital twin. The first can be provided from Unigine I suppose, but what about the second? Edited January 23, 2020 by davide445 Link to comment
Recommended Posts