[转载] look at this computer - 未名空间MITBBS历史存档

国际科技财经博客移民网络热点娱乐民生时事公众号

Redian新闻

>未名空间

>Computation - 科学计算

[转载] look at this computer

[转载] look at this computer# Computation - 科学计算

j*g2004-03-04 08:03

1 楼

【以下文字转载自 ME 讨论区,原文如下】
发信人: jbjiang (temp), 信区: ME
标题: look at this computer
发信站: Unknown Space - 未名空间 (Thu Mar 4 17:34:39 2004) WWW-POST
Interl P 4 cpu 2.80Ghz
At/At compatible
1047556kb Ram.
It is enought to do scientific calculation?
thanks

g*e2004-03-04 08:03

2 楼

ft...得看多大的计算量啊.

【在 j*****g 的大作中提到】

: 【以下文字转载自 ME 讨论区,原文如下】
: 发信人: jbjiang (temp), 信区: ME
: 标题: look at this computer
: 发信站: Unknown Space - 未名空间 (Thu Mar 4 17:34:39 2004) WWW-POST
: Interl P 4 cpu 2.80Ghz
: At/At compatible
: 1047556kb Ram.
: It is enought to do scientific calculation?
: thanks

j*g2004-03-04 08:03

3 楼

Thanks for the reply.
If I do Direct numeric simulation, is PC possible?
What's the maximam memory for the pc up to now?
thansk

【在 g********e 的大作中提到】

: ft...得看多大的计算量啊.

a*s2004-03-04 08:03

4 楼

不够，你原帖的数字都乘以16就够了。

【在 j*****g 的大作中提到】

: Thanks for the reply.
: If I do Direct numeric simulation, is PC possible?
: What's the maximam memory for the pc up to now?
: thansk

f*r2004-03-04 08:03

5 楼

DNS? You may need Earth Simulator. :) OK. Just kidding.
The requirement of computational facilities really depends on what
you are going to solve. As for the case of DNS, a rough estimation
of grid nodes is propotional to Re^{9/4}. Meanwhile, you time
step should also be small enough to resolve the smalles time
scale in the turbulence flow field. Therefore, if the Re of target
flow field is not very high, such a machine might be OK. However,
the solution may hardly be meaningful, 'coz the work's

【在 a******s 的大作中提到】

: 不够，你原帖的数字都乘以16就够了。

a*s2004-03-04 08:03

6 楼

这个问题最有发言权的也许就是搞湍流的，嘿嘿

【在 f****r 的大作中提到】

: DNS? You may need Earth Simulator. :) OK. Just kidding.
: The requirement of computational facilities really depends on what
: you are going to solve. As for the case of DNS, a rough estimation
: of grid nodes is propotional to Re^{9/4}. Meanwhile, you time
: step should also be small enough to resolve the smalles time
: scale in the turbulence flow field. Therefore, if the Re of target
: flow field is not very high, such a machine might be OK. However,
: the solution may hardly be meaningful, 'coz the work's

a*s2004-03-04 08:03

7 楼

然后搞湍流的崩出来说，就你这机器都乘以32，10台都不够，哈哈

【在 a******s 的大作中提到】

: 这个问题最有发言权的也许就是搞湍流的，嘿嘿

l*g2004-03-04 08:03

8 楼

For memory limit, it depends your operating system. The 32-bit windows system
can only support 2G memory.

【在 j*****g 的大作中提到】

: Thanks for the reply.
: If I do Direct numeric simulation, is PC possible?
: What's the maximam memory for the pc up to now?
: thansk