“What do you think will happen when every home is connected to the internet via 100 or 1,000Mbps Ethernet or fiber?”
He goes on to give an answer that is yes, under the assumption that the 100Mbps is symmetrical.
“At some point in the not-so-distant future, then, we’re all going to be connected to the web at LAN-like speeds — 100 megabits per second up and down — and this, just like the advent of the telephone, will change the world as we know it. … ”
“Instead of your entire life being represented by a handful of bytes in amongst Facebook’s faceless sea, symmetric connections will enable the web to becomemetropolitan. Your presence on the web will be your home. ”
“The end result would be a truly decentralized internet that closely mimics human settlement and society. There will still be nodes on the internet where more people congregate — the bars, clubs, and McDonalds of the real world — but for the most part, a symmetric web would let people hang out and connect with the people they care about, and ignore everyone else.”
This is my definition of real cloud computing – something way beyond the standard view which is not much more than a new marketing twist on the old time-sharing data centers.
We see them every day, at work, on the train, at conferences, at coffee shops, and everywhere else that people might gather: hand-held devices that function as telephones, Internet access devices, network access devices, or perhaps merely toys upon which we can play “Angry Birds.”
They are pervasive, inexpensive, and versatile. But they also beg the question of whether they are truly ready for prime time, in the business sense of the term.
The truth is that a lot depends on how these devices are used and what sort of access they are granted. At the end of the day, this question is really about balancing convenience and security. The knee-jerk response is that security will trump convenience every time. However, if convenience enables people to be more connected and do more work than the otherwise might, then that makes the decision somewhat tricky, doesn’t it?
Coldwater Vision Center was plagued by an aging phone system and an unreliable network. When the network went down, staff couldn’t access patient records or file insurance claims. Like many small businesses, Coldwater Vision’s network had grown organically with devices from various vendors added piecemeal over time. The solution? An end-to-end integrated Cisco solution, which improved productivity and staff morale.
The pace of IT and data is moving exponentially. With a rapidly growing networks across all industries, keeping track of all these connection points will give rise to a new business leader: The Chief Information Officer, or CIO.
Why will a CIO be such a crucial part of future business? Read More »
MPI implementations are large, complex beasts. By definition, they span many different layers ranging from the user-level implementation all the way down to sending 1s and 0s across a physical network medium.
However, not all MPI implementations actually “own” the code at every level. Consider: a TCP-based MPI implementation only “owns” the user-level middleware. It cannot see or change anything in the TCP stack (or below). Such an implementation is limited to optimizations at the user space level.
That being said, there certainly are many optimizations possible at the user level. In fact, user space is probably where the largest number of optimizations are typically possible. Indeed, nothing can save your MPI_BCAST performance if you’re using a lousy broadcast algorithm.
However, lower-layer optimizations are just as important, and can deliver many things that simply cannot be effected from user space.