[00:00.00] |
From VOA Learning English, |
[00:02.79] |
this is the Technology Report. |
[00:05.68] |
China's Tianhe-2 supercomputer |
[00:09.53] |
has just been rated No. 1 on the top 500 |
[00:14.74] |
-- a respected list of the world's most powerful computers. |
[00:20.31] |
Experts measured the supercomputer's performance |
[00:24.59] |
at 33.86 petaflop or quadrillion of operations per second. |
[00:34.21] |
China's National University of Defense Technology |
[00:38.58] |
developed the supercomputer, which runs twice |
[00:42.77] |
as fast as the No.2 rated Titan supercomputer. |
[00:48.16] |
It belongs to the United States Government's |
[00:51.80] |
Oak Ridge National Laboratory in Tennessee. |
[00:55.43] |
Both Tianhe-2 and Titan are part of an ongoing race |
[01:01.28] |
to make supercomputers faster and more powerful. |
[01:06.27] |
So what is a supercomputer? |
[01:10.02] |
A basic personal computer has one microchip |
[01:14.79] |
at the center of its operations. |
[01:17.00] |
This Central Processing Unit, or CPU, |
[01:22.18] |
executes a set of commands contained in a predesigned program. |
[01:28.25] |
The first supercomputers had a few more CPUs. |
[01:33.63] |
That number grew as microprocessors |
[01:36.91] |
became cheaper and faster. |
[01:39.74] |
Andrew Grimshaw, a computer science professor |
[01:43.42] |
at the University of Virginia explains: |
[01:47.00] |
"Today, supercomputers are all what we call parallel machines. |
[01:50.55] |
Instead of one CPU - central processing unit |
[01:53.63] |
- they have thousands and thousands. |
[01:55.46] |
And in the case of the Chinese machine, |
[01:57.35] |
depending on how you count, |
[01:58.91] |
millions of the central processing units." |
[02:01.34] |
These parallel machines are made up of |
[02:04.32] |
many individual computers called nodes. |
[02:08.30] |
They are all positioned in one block. |
[02:11.75] |
They used a lot of power, create a lot of heat, |
[02:16.82] |
and require huge cooling systems. |
[02:20.45] |
They also use programs different |
[02:23.69] |
from those used by ordinary computers. |
[02:27.31] |
Professor Grimshaw says anyone with enough resources |
[02:31.75] |
can build a supercomputer to solve problems |
[02:35.97] |
that require millions of mathematical calculations. |
[02:40.61] |
But that's not always necessary. |
[02:44.29] |
A virtual supercomputer can be created |
[02:47.82] |
by networking individual computers |
[02:50.76] |
within a university campus or company. |
[02:54.35] |
These machines then process data during down time, |
[02:59.06] |
when no one is using them. |
[03:01.41] |
"Those are very easy to run on virtual supercomputers |
[03:05.20] |
because each problem is independent of all the others |
[03:08.19] |
and I can scatter these jobs out all around the place. |
[03:10.98] |
We run these all the time at UVA." |
[03:12.81] |
Professor Grimshaw says that entire ten years ago, |
[03:16.79] |
engineers worked on making computers faster. |
[03:20.77] |
Since then, he says, they have worked to create |
[03:25.21] |
more powerful parallel machines. |
[03:27.70] |
"It's transforming science and engineering, |
[03:30.19] |
and it's going to continue to transform it |
[03:32.10] |
in ways I think most people don't fully grasp |
[03:36.52] |
- how well we can model and simulate the world now." |
[03:39.86] |
Professor Grimshaw says |
[03:41.69] |
the increasing computing ability of supercomputers |
[03:45.34] |
makes the future of research very bright. |
[03:49.62] |
And that's the Technology Report from VOA Learning English. |