Public cloud computing has allowed enterprise users to scale up quickly and to move short-term workloads out of their datacenters, reducing costs. At the same time, consumers have been trading their clunky laptops for smaller, less-powerful netbooks, tablets, and smartphones. With Kindle Fire, Amazon has just taken a big step toward making smaller, less-powerful devices more practical for enterprise use by marrying its new Silk browser to its powerful AWS EC2 cloud service. The new tablet will be able to handle simple tasks, while heavy processing will be performed in Amazon’s EC2 cloud.
As I said in my previous post on this year’s Intel Developer Forum (IDF), the conference was full of innovative ideas from Intel and its partners. Because of space constraints, my first post didn’t describe every innovation I wanted to talk about, so here they are, in no particular order:
Intel and IBM recently outlined development efforts that stack semiconductor chips to achieve greater density and higher performance. Chip stacking is a “hot topic,” but its success will depend on packaging stacks in such a way that keeps them from becoming too thermally “hot.”
Three-dimensional (3D) stacking of chips not only offers a smaller physical package but also shortens wires, which can allow higher performance. Early use of stacked chips focused on dense-packaging advantages, typically by stacking relatively low-power memory chips. Now, Intel and partner Micron Technology are looking to increase memory bandwidth with a Hybrid Memory Cube (HMC) using through-silicon via (TSV) connections. IBM and partner 3M are developing a thermal adhesive to address the problem of removing heat from stacked chips.
Wouldn’t it be nice if buying a server was as simple as buying a plane ticket? When I took a recent trip to New York City, the travel website Orbitz found a flight that saved me time and money by taking me to the small Westchester County Airport instead of one of the large New York hubs. Orbitz and travel sites like it have revolutionized the way we buy airline tickets. We no longer need to depend on a travel agent; instead, we just plug our data into the computer and the automated, vendor-neutral tool runs thousands of “what-if” scenarios, customizing these searches to our needs.
If you’re wondering where the “Orbitz” model for server selection is, wonder no more. IDEAS is bringing server selection into the modern era with the Target Server Wizard, just added to the ServerCAR server consolidation tool. Instead of spending hours scouring spec sheets or relying on reseller recommendations, the Target Server Wizard automatically searches a constantly updated database of over 20,000 server configurations to find the one that best suits your needs.
IDEAS recently attended the IBM Smarter Computing Summit in Beijing China. Our previous story looked at the importance of the Growth Markets to IBM’s business and what IBM is doing around cloud. In this second part of the story, we will take a closer look at the other two main pillars of IBM’s Smarter Computing message: Optimizing Systems, and Big Data and Analytics.
It’s not news that businesses and industry analysts often idealize new technology innovations, and the cloud is certainly no exception. One of the main selling points of cloud computing is that it eliminates inventory risk. But end users may not realize that with the cloud they take on a new risk – inflation.
Money being used as wallpaper during hyperinflation in the Weimar Republic. SOURCE Credit: Pahl, George
Growth Markets, Big Data, Cloud, and Optimizing Systems for workloads were the three key themes of the recent IBM Smarter Computing Summit in Beijing China for press and analysts, which IDEAS attended. This event was run by the Systems & Technology Group (STG) from the Growth Markets business unit of IBM and therefore had a lot more technology content than IDEAS is used to seeing at such events of late. The following provides an overview of the key topic areas covered over the two days of sessions.
When the first Transaction Processing Performance Council (TPC) benchmark appeared in 1989, a customer database that contained records through the prior month was sufficiently current to do business analysis. Today, a database snapshot that is 24 hours old is too stale to be relevant when analyzing the impact of a new Twitter campaign. Papers presented at a TPC conference this week suggest changes are needed in database benchmarks to keep up with these changes in database usage.
