a*****e
发帖数: 1700 | 1 automata 是初级 theory 课上教的吧，算不得 PL。当然也有 PL
领域是专门研究 automata 的，但这不会是课上的内容。
初级 PL 课主要教形式语义，类型系统，lambda，函数式或者命令
式语言的实现，等等。也有一定的逻辑推导和程序证明，要看是谁
教，用哪本教材了。
我遇到的常见论调就是 wdong 说的这种，认为没用。但实际上，
我迄今还没有遇到一个认真学习了 lambda 和 type theory 的人
之后告诉我这两样没用的。相反，大多数都会认为打开了一扇门
看到了别处的风景，然后能够站在新的高度上回头审视自己的编程
技巧和理解能力。
像 parallel programming 和 system architecture 这种课，我觉
得自学也可以取得不错的效果，但是 PL 绝对是要认真上课认真完
成作业才可能有收获，对思维是很好的挑战和训练。靠自学难度相
当大。 |
|
w*r
发帖数: 2421 | 2 做parallel database的？Teradata?这个东西会有教程吗？？？
还是回去把上的课里面的paralle computing的书温温吧，这种怪物是没有教程的，每
个公司都有自己的implementation, so far 在DB里面做得最成功的就是Teradata的PDE
了 |
|
d*****n
发帖数: 2132 | 3 随便尝试了几个例子,parallel computing 一跑起来所有cpu core 全是100% usage,
这样一来其它程序不用跑了,谁敢把这样的程序放到production的服务器上去?
微软那帮人到底是怎么想的? |
|
k**n
发帖数: 3989 | 4 用p不就是为了尽量用cpu吗？
那些例子估计如果不用parallel computing,
cpu useage是多少？
在单cpu与多核cpu上有啥不同？ |
|
S***k
发帖数: 370 | 5 One of my project uses parallel in production. I have not noticed such issue
yet.
Could you give a specific sample? |
|
x**n
发帖数: 461 | 6 Our projects (healthcare industry) use parallel computer of .net 4 without
any issue. Your problem is that you need new hardware/scale out or revise
your algorithm. |
|
d*****n
发帖数: 2132 | 7 thanks for reply.
If it works for you, then there must be some disconnection somewhere between
you and me.
I attached the sample below and did the test just now on a 6 core, 8 gb
machine, it just reconfirmed my previous observation: all cpu runs at 100%
while the program is running.
Can you try on your server and see if you run into the same thing?
Thanks!
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Threading;
... 阅读全帖 |
|
d*****n
发帖数: 2132 | 8 thanks for your reply, the hardware I used to test is already decent 6 cores
with 8 gb of memory. I also attached a example I used to test .net 4.0
parallel computing in a previous post under this thread, can u check if it
happens to you too?
thx. |
|
|
S***k
发帖数: 370 | 9 Try to use parallelOption
var options = new ParallelOptions {
MaxDegreeOfParallelism = 2 // set it to half of num of cores.
};
Parallel.For(2, 1000, options, (i) => {
var result = SumRootN(i);
Console.WriteLine("root {0} : {1} ", i, result);
});
between |
|
|
w*l
发帖数: 43 | 11 Hi,
Who are working on java parallel database ? Can you give me some ideas about
it, such as how to create the database and how to query the database ? In fact
I need some marterials about it, can you give me some websites about it?
Thanks |
|
L*******r
发帖数: 310 | 12 今天在ubuntu上装了一个intel@Parallel Studio XE 2011
本来指望能用它来学习fortran的。
装成功后，我怎么也找不到这个程序，这是什么情况？它躲在哪里？ |
|
L*******r
发帖数: 310 | 13 谢谢。
在bin里面找到一个叫“ip”从属性看估计就是昨天装的parallel studio。
这个要怎么运行，双击打不开? |
|
w*s
发帖数: 7227 | 14 intel, MS, ... all supports parallel programming ... |
|
c******o
发帖数: 1277 | 15 今天卡看了一个slides，明白为啥我觉得akka没啥用。。。
http://www.slideshare.net/jboner/going-reactive-eventdriven-sca
Jonas Bonér TypeSafe CTO 做的， 里面讲了编写parallel时候可以用的工具（都很
流行啊）
1.先用immutability, future/data-flow 等语言直接支持的的工具,能并行够用就好
2.不够用，用Actor/Rx/Agent, 这类Async的分散式计算工具，这样可以有async的通讯。
3.不够用，必须有transaction,和share states, 就再加上STM
4.还不够用，那就只能上old Locks/显性thread 了。
我么只用到了 1/2之间，可能以后有3 |
|
c******o
发帖数: 1277 | 16 那是单机吧？
MMO怎么保证？MMO你没有办法集中资源在一个视角上。
你基本上是保证数据量不到一个上限，用那个来parttition,这个也就是说你能handle
本来async就是牺牲consistency 保证 availability . 你都没有availability问题，
用什么async
Async/big data/nosql都是应为单个node/thread不行，必须parallel
所以consistency/availability不能都达到，所以才出现的，也不是什么都适合。 |
|
c******o
发帖数: 1277 | 17 你这儿的假设就是场景之间的联系很少，可以近似看成独立的， 而且你也没法控制人
群（wow就碰到过很多次）。
独立的node如果能力够是可以availability/consistency(场景内一致）都有
所以不需要async/parallel
只有必须多node/thread/process才是要async
FB/twitter以前肯定也不需要考虑这个，但是数据量上去了不就必须考虑了么？ |
|
p*****2
发帖数: 21240 | 18 人家自己都说了go不是用来搞parallel的吧？ |
|
m******t
发帖数: 635 | 19 大牛折腾过clojure的core.async么? 貌似学习的Go在parallel方面的思路 |
|
c*********e
发帖数: 16335 | 20 parallelism 到底是个什么？这个你知道吧？ |
|
s***o
发帖数: 2191 | 21 我觉得Spark的问题很有道理阿，parallelism != concurrency。你们几个能不能好好
解释一下，大家也都学习学习。 |
|
s********k
发帖数: 6180 | 22 看看rob pike的经典slides吧，parallelism可以被concurrency执行，也可以不， |
|
s********k
发帖数: 6180 | 23 举个例子：你接娃回家和老婆做饭就是parallelism两件事，因为不share资源，这两件
事可以concurrent执行，你接回家老婆饭也做好，但是也可以不concurrent执行，你接
回家老婆再做饭 |
|
s********k
发帖数: 6180 | 24 你的例子跟我其实本质一样，不过parallelism强调的不是同时进行，而是互不冲突可
以同时进行 |
|
v****m
发帖数: 26 | 25 请教各位，在unix下用fortran90怎么实现parallel计算，可利用多cpu。据说只需编译时
指定某一命令
有没有那位给说一说。多谢。 |
|
t****n
发帖数: 39 | 26 I don't know what's your server. If you are using SGI serious, such as
O2, Origin 2000, etc. the compiler come with the system can do the
parallization for you. The command is "f90 -apo source.f90 -o des"
There is another choice name OpenMP. You can go to www.openmp.org for detail.
What you need to do is just add some Macro to your original code and
compile with -mp option, you can get parallelized executable code.
The other choice is to rewrite the code thoroughly with MPI or PVM. Try to
search |
|
s**i
发帖数: 381 | 27 Is there any good library to help us program in parallel(MPI) using
unstructured grid?
Thanks a lot |
|
s*****n
发帖数: 62 | 28 Yes, inorder to accept an exchange.
Another question,
Suppose you are doing Parallel tempering in canonical emsembles, if you
replace one canonical
simulation with microcanonical simulation for one replica, does this affect
the canonical disbributions of
other replicas?
te |
|
M*****t
发帖数: 120 | 29 It is called Parallel Computing Toolbox? |
|
c*******h
发帖数: 1096 | 30 Let me make the point clear.
You can use some functions provided by nvidia in matlab, such as fft. But
that's almost all what you can utilize---this set of functions is very
limited. Imagine that you want to have a parallel version of some function,
say the popular svd, that can run on gpu but that is not shipped with the
cuda plugin. So what do you do? One solution is to write a C implementation
(of svd) that fully exploits the CUBLAS capacity, then mex it, and run your
function in matlab. Hoor |
|
s********1
发帖数: 581 | 31 我需要做parallel programming. 希望所装的系统能够马上支持，不需要下在其他插件。 |
|
l********a
发帖数: 1154 | 32 如果问题规模不大,处理时间不长,用parallel toolbox效果是相反的,反而会变慢. |
|
L*******r
发帖数: 310 | 33 今天在ubuntu上装了一个intel@Parallel Studio XE 2011
本来指望能用它来学习fortran的。
装成功后，我怎么也找不到这个程序，这是什么情况？它躲在哪里？ |
|
s*******1
发帖数: 135 | 34 I have difficulty in every question of LSAT, esp. parallel reasoning.
Fortunately I can get help from here.
Some people believe witnessing violence in movies will discharge aggressive
energy. Does watching someone else eat fill one's own stomach?
D. Some people think that economist can control inflation. Can
meteorologists make the sun shine?
E. Some people think workaholics are compensating for a lack of
interpersonal skills. However, aren't most doctors workaholics?
I chose D instantly, then s |
|
c*******l
发帖数: 4801 | 35 【 以下文字转载自 EE 讨论区 】
发信人: chernobyl (小水葱 golden playboy,金童玉女之金童), 信区: EE
标 题: Asymmetric angle Parallel Beam in CT image reconstruction
发信站: BBS 未名空间站 (Wed Dec 13 12:59:43 2006)
if i use algebraic method, should I treat asymmetric just like symmetric,
updates all cells falling in the ray, just use different way of defining
each projection for angles?
Thanks. |
|
l***a
发帖数: 12410 | 36 sort和means还有一些其它的proc号称是自动take advantage of multi-processor，我
理解就是不需要给threads=yes的option，sas就可以自动使用多线程跑这些proc吧。
如果这样的话，对于这一类优化过的proc，是不是没必要再用rsubmit/endrsubmit来人
工安排parallel processing？ |
|
d*******o
发帖数: 493 | 37 parallel data processing是个难题，像流行的Python都加了多核锁，禁止了这方面的
运用。我不觉得现在的SAS可以加几个参数可以实现。算法和数据结构都要改，SAS必须
要用新的procedure和跟现在不一样的数据结构。 |
|
c*****l
发帖数: 297 | 38 MOST OF THE ALGORITHM IS POLYNOMIAL Time,
you don't need to use paralleled computing. no significant difference
in running time |
|
发帖数: 1 | 39 美东生统phd，之前有bioinfo的ms，phd课题是开发nonparametric的统计方法用在
sparse network searching上。工作找了一段时间了，发现越来越难，之前面了一家大
的药厂的非clinical的职位，到了第三轮onsite，最后反馈说...solid background
in statistics...but need more experience in gene expression and epigenetic
modification disease biology 所以就没面上（此处生物背景加分，可惜我没），其
他的都没到onsite。
熟练使用R和linux服务器parallel并行计算，基本统计模型，mixed，glm，glimmix,
longitudinal等都会但没机会实战，SAS会但不熟用的不多，github，shiny都会。
Machine learning的很多model也会，做过相关project，PCA，tree based methods，
etc。现在学习time series和python，乐于学习新技能。
我... 阅读全帖 |
|
w*******y
发帖数: 60932 | 40 Amazon has reduced the price on Parallels 5.0, the best virtual application
enabling you to run Windows on Your Mac.
Purchase Price: $39.99 & this item ships for FREE with Super Saver Shipping
With Rebate $19.99 after $20.00 rebate
|
|
|
|
|
|
f**d
发帖数: 768 | 45 这是一本计算神经科学的优秀著作，全文拷贝这里（图和公式缺），有兴趣的同学可以
阅读
如需要，我可以分享PDF文件（－－仅供个人学习，无商业用途）
From Computer to Brain
William W. Lytton
From Computer to Brain
Foundations of Computational Neuroscience
Springer
William W. Lytton, M.D.
Associate Professor, State University of New York, Downstato, Brooklyn, NY
Visiting Associate Professor, University of Wisconsin, Madison
Visiting Associate Professor, Polytechnic University, Brooklyn, NY
Staff Neurologist., Kings County Hospital, Brooklyn, NY
In From Computer to Brain: ... 阅读全帖 |
|
l*****a
发帖数: 38403 | 46 这里是各种译文
Revelation 12:12 NIV
New International Version
Therefore rejoice, you heavens and you who dwell in them! But woe to the
earth and the sea, because the devil has gone down to you! He is filled with
fury, because he knows that his time is short."
Read Revelation 12 NIV | Read Revelation 12:12 NIV in parallel
Revelation 12:12 ASV
American Standard Version
Therefore rejoice, O heavens, and ye that dwell in them. Woe for the earth
and for the sea: because the devil is gone down unto you, h... 阅读全帖 |
|
l*****a
发帖数: 38403 | 47 你这版本，神的众子为什么翻译成天使，另外据说圣经是有圣灵随着你们，是大有神迹
，不可能有错误冲突滴？ 另外你这没回答，我问的是在创造人类之前就造了上亿天使
的原文是什么，另外按你这逻辑，晨星也是耶稣的称号，你是不是可以认为，有很多耶
稣啊？
真有意思阿
Bible VersionsJob 38Job 38:7Compare Translations
Job 38:7 NIV
New International Version
while the morning stars sang together and all the angels shouted for joy?
Read Job 38 NIV | Read Job 38:7 NIV in parallel
Job 38:7 ASV
American Standard Version
When the morning stars sang together, And all the sons of God shouted for
joy?
Read Job 38 ASV | Read Job 38:7 ASV in p... 阅读全帖 |
|
G*******s
发帖数: 4956 | 48 Blinded By Tradition:
An Open Letter to Dave Hunt
Regarding His Newly Published Attack Upon the Reformation, What Love Is This
? Calvinism's Misrepresentation of God
James White
Updated 5/16/02, see bottom of file
Dear Dave,
In the period of time since I finally received my own copy of your book (you
may recall I scanned through it while standing at your table in St. Louis a
t the PFO Conference) I have gone through a number of different emotions. A
t first I was just going to do one Dividing ... 阅读全帖 |
|
