The reality is that much of what is called cloud is private cloud, not public cloud. I have seen recent reports that show 70% of cloud spending is on private clouds. Now this is just the same as the old IBM mainframe computers. There consist of a virtualized, centralized resource accessed by a network.
Some of the "benefits" of the cloud are just the network. These include collaboration tools. You don't need the :on-demand" cloud for that.
As for computationally intensive workloads, it all depends. There are some that are loosely coupled and do not require a lot of communication between parallel nodes. These are great for the cloud. On the other hand, a tightly coupled computation using specialized resources is not a great candidate for the cloud.
While the cloud is great for start-ups (and I work on some of those), it does cost more in the end. If your load is not steady, then the cloud is also a useful tool. In this case, though, you will probably be well served by having a provate cloud and using a public cloud for peaks in processing. If your load is steady, and you know what it is, a private cloud is much more appropriate.
I worked, many years ago, with and on some of the largest service bureaus in the world. These were put out of business when it was found that buying mini-computers was cheaper and more responsive. I wonder how long this will take with the cloud.
You raise some good points, Naperlou, but I do think the cloud offers benefits not realized in the older centralized computing models. For one thing, access to cloud based apps means you can tap them any time, any where, from a browser, and of course the next logical step is tapping them on mobile devices like tablets. With the old centralized mainframe model, you were tied to your dumb terminal and there was no where near enough processing power on it for graphics-intensive applications. Also, part of the promise of the cloud is ease of administration and allowing users to spin up and take down compute resources whenever needed, without requiring an assist from IT. Your points about having peak workloads being a candidate for the cloud would fit with this benefit. And I do agree with you that most efforts, especially for engineering-driven applications, will be private or hybrid clouds, not happening in the public cloud for all the usual security/IP issues.
Naperlous, you are right. Most of the companies prefer either private or hybrid cloud, and public cloud is their least preferred one. This is a part of security concern and connectivity issues. But public clouds are the best business model for startup companies, where they are renting the resources for business establishments on a pay as you go mode.
@Mydesign: What kind of startup companies are you referring to in terms of public clouds being best for business? If you're talking about being a design tool vendor startup and using the public cloud to deliver your offering as Software-as-a-Service, then the startup is responsible for tuning the security and resiliency capabilities as part of their service in order to attract engineers and keep them comfortable with the cloud-based software delivery model.
realted to this post, I'd like to introduce this cloud computing software for analysis of composite materials,
Composites Materials Software, New Release 12, now implementing Cloud Computing
New Cloud Based Engineering Software has been released. Inspired on the installable application CADEC-R10 that has been evolving since 1998, this new release R12 is a completely new version with expanded functionality and now implemented on the Cloud.
CADEC-ONLINE.COM (http://www.cadec-online.com ) has been released. Inspired on the trustworthy CADEC that has been evolving since 1998, this new release is completely new version with expanded functionality and now implemented on the Cloud. CADEC R12 is an online application that performs composite materials analysis. It is structured to provide a user friendly computational engine for the analysis equations and methods described in the textbook Introduction to Composite Materials Design--Second Edition, CRC, 2010 (http://barbero.cadec-online.com/icmd/index.html ). CADEC-ONLINE.COM performs micromechanics for composites reinforced with unidirectional fibers, and random fibers, as well as plain weave, twill, and satin textile fabrics. Predicted properties include lamina moduli, strength values, CTE, moisture expansion, and more. Its laminate analysis capabilities are extensive, including stress, strain, and failure. Failure predictions including first ply failure and last ply failure under mechanical, thermal, and moisture loads. The most advanced analysis formulas are implemented. For example, failure predictions include in situ effects and model failure criteria such as Hashin, Puck, and so on. The Cloud implementation means that updates are transparent; no new releases have to be installed by the user. Instead, new functionalities appear without disruptions. One of its many unique features is the comprehensive analysis capabilities for laminated composite thin walled beams with general cross sections. Beams can be asymmetric and loaded by general combinations of forces in three planes (axial, vertical and horizontal) as well as by three moments (torque and two bending moments). The software computes all the section properties including the shear center and so on. CADEC-ONLINE.COM features the latest IT technology, including object-oriented programing, database management, an advanced web interface, and more. For example, data integrity allows users to refine the properties of any object, such as fiber properties, and the software will automatically and transparently update all the dependent objects, such as laminas, laminates, and beam sections. An encrypted database maintains all of the user's objects, which are on the cloud, available everywhere. A 30 day demo subscription allows interested users to drive test the software. Faculty and students get a lifetime free subscription while commercial users can purchase a premium subscription.
Intuitive navigation includes the "left tree" shown in the attachment.
For a fully functional DEMO subscription, just visit cadec-online.com
Beth, security is a major concern with public clouds, but for minimizing the investment cost startup companies prefers it. Later they can move to either private or hybrid cloud. Hybrid clouds have the advantage of both public and private cloud, so they can decide which portion has to move to the public/private cloud base on various parameters like security, connectivity etc.
Absolutely. From a startup vendor's perspective, embracing public cloud to offer your 3D design or simulation tool as a service is much more economical in terms of getting up and running quickly. But of course then the onus is on you as the cloud software provider to ensure the proper security protocols and controls are in place and to convince engineering organizations that you have them covered. From an internal organization's perspective, hybrid or private clouds are the more likely scenario unless of course, they are comfortable with a particular SaaS design tool offering.
A new service lets engineers and orthopedic surgeons design and 3D print highly accurate, patient-specific, orthopedic medical implants made of metal -- without owning a 3D printer. Using free, downloadable software, users can import ASCII and binary .STL files, design the implant, and send an encrypted design file to a third-party manufacturer.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.