Why Japan Continues to Use In-House Capabilities For 3D Rendering and Research
While Japan continues to see major advances in 3D computer graphic rendering, it has surprised many that this user base in not moving aggressively to the cloud. Research carried out by industry experts and by IT companies themselves has delivered important insights as to why this is so. It is hoped that when policy makers understand the nature of the problem fully, solutions will be found.
To begin with, it goes without saying that all the usual advantages of cloud computing also apply to the 3D market if it were to use general purpose graphics processing units (GPGPU) in the cloud. Industries with a proven demand for high powered 3D processing include the construction and design sector, educational institutions, the animation industry, commercial message production, and the gaming industry to name just a few.
Surprisingly, many of these still rely on in-house servers and IT staff to manage their graphics programming needs. And over the last two years, industries of this kind have made annual capital investments for on-premise GPGPU equipment in excess of ￥50,000,000. Much of these capital expenditures could have easily been shifted to the GPGPU cloud resulting in substantial savings. A large number of universities and major scientific institutions in Japan have made very major investments in on-premise GPGPU equipment as well.
The key question here is why are these companies not using the GPGPU cloud?
After some research, the reasons become clear. There are both technical and cultural reasons for the reluctance to shift.
Technical reasons
3D data tends to be huge in size. 150 TB is not an unusual volume. Many would-be users are afraid that the available bandwidth would not support this kind of data, and even the first time upload of 150 TB will take a very long time. Working with this kind of data on their network, they fear, will be very slow. There is also worry that crucial company secrets could be lost. If you are designing a new jet engine, you will naturally be very reluctant to place a design for the latest turbine blade or combustion chamber in the cloud.
Since companies have already invested in GPGPU class machines, they would like to continue to exploit these fully as well.
Cultural Reasons
The cultural reasons for continuing to use on-premise systems are also very clear.
Many research institutions are able to get very large grants for academic purposes by claiming very special needs that cannot be met by a general purpose cloud. Even something such as a monthly billing of cloud services is being resisted by the clerical staff in the accounts departments that resent the additional work.
Actual users are really not concerned about where their computing power comes from. They do not care if it comes from the cloud or from their own servers so long as there are no performance issues.
The solution
It is well known and accepted that the cloud gives you far greater power than in-house servers could. A cloud-based Tesla M2075 processor has 56 times the cores of a Xeon processor and is three times as fast yet it consumes only twice the power. Perhaps the decision makers in the companies are not being educated adequately about facts such as these. That there are major advantages of the move to the cloud which cannot be denied, but obviously, the huge capital expenditure that has already been made will have to be used first.
To give you a better idea of how the computing power of the cloud can be maximized – especially in 3D rendering – GMO Cloud presents an illustration that can sum up the usage of the technology through high availability systems.
Government entities can also realize the advantages and alter the grants and research funding to gently nudge institutions to greater use of the cloud. If security can be ensured and proven to be as strong as in-house systems can provide, this will make a big difference in the adoption.
About the Guest Author:
Sanjay Srivastava has been active in computing infrastructure and has participated in major projects on cloud computing, networking, VoIP and in creation of applications running over distributed databases. Due to a military background, his focus has always been on stability and availability of infrastructure. Sanjay was the Director of Information Technology in a major enterprise and managed the transition from legacy software to fully networked operations using private cloud infrastructure. He now writes extensively on cloud computing and networking and is about to move to his farm in Central India where he plans to use cloud computing and modern technology to improve the lives of rural folk in India.
