x*********n
发帖数: 28013 | 1 比如。我现在
ping x。x。x。x r 1000次。没有drop
但是traceroute
4 216.80.5.50 4 msec 4 msec 4 msec
5 63.237.42.129 8 msec 4 msec 8 msec
6 205.171.139.62 [MPLS: Label 797744 Exp 0] 4 msec 4 msec 8 msec
7 * * *
8 67.14.8.194 4 msec 4 msec 8 msec
9 63.146.26.218 4 msec 4 msec 8 msec
10 12.122.84.54 [MPLS: Label 17124 Exp 0] 32 msec 28 msec 32 msec
11 12.122.2.206 [MPLS: Labels 17694/17132 Exp 0] 32 msec 28 msec 32 msec
12 12.122.2.126 [MPLS: Labels 23426/17132 Exp 0] 32 msec 32 msec 32 msec
中间的* * *能表... 阅读全帖 |
|
z**r
发帖数: 17771 | 2 NH应该是far end PE loopback，这个肯定是从IGP得到，同时也应该在LFIB里，你不激
活ipv4 neighbor，同时看一下show mpls for？
刚才俺也做了个实验，果然不需要在add ipv4下激活ipv4 neighbor，只需要在add
ipv6下激活ipv4 neighbor就可以送ipv6扩展了。
R1----R2----R3----R4
R1 and R4 are dual stack, R2 and R3 are IPv4 only, LDP enabled on all
interface
相关配置如下
!
!
!
ip cef
no ip domain lookup
ipv6 unicast-routing
ipv6 cef
!
!
interface Loopback0
ip address 10.1.1.1 255.255.255.255
!
interface Loopback1
no ip address
ipv6 address 2001::1/128
!
interface Ethernet0/0
no ip address
... 阅读全帖 |
|
r*a
发帖数: 1503 | 3 非常感谢！ 电源，显卡有什么可以推荐和它搭配的嘛？
这里就是400块的那个组合。
MSI P67A-C43 B3 Intel P67 Motherboard
Processor Socket: Intel
Processor Interface: LGA1155
Form Factor: ATX
Processors Supported: 2nd generation Intel® Core™ i3
2nd generation Intel® Core™ i5
2nd generation Intel® Core™ i7
Additional Technologies: THX TruStudio PRO
MSI ClickBIOS
Military Class II
OC Genie II
Northbridge: Intel® P67 Express
Memory Supported: DDR3 1333MHz
DDR3 1066MHz
DDR3 1600MHz (... 阅读全帖 |
|
h*b
发帖数: 209 | 4 【 以下文字转载自 Working 讨论区 】
发信人: htb (小牛 。。。妞。。。), 信区: Working
标 题: 问个面试问题
关键字: in
发信站: BBS 未名空间站 (Wed Jun 13 16:55:27 2012, 美东)
面试的时候问了下面的问题，没有回答上来：
下面的程序当N=64*1024的时候，run time是100 msec,
当N=70*1024的时候，time变成了74 msec。为什么N越大run time
越小？
如果run多次的时候，run time可能变成：
N=64： 100， 76， 500， 80， 90
N=70： 74， 80， 300， 63， 80
为什么会出现500/300这突变的数字？
把a，b,c,d,e初始化成0也会出现同样的现象。
#define N 64*1024
int a[N], b[N], c[N], d[N], e[N];
int main()
{
int i;
for (i = 0; i
a[i] = b[i] + c[i] + d[i] + e[i]... 阅读全帖 |
|
t****a
发帖数: 1212 | 5 这是很有趣的问题: 如何编写并行的recusion?
我在i5 macbook pro上做了个小试验: 并行的recusion based fib函数，用FP语言(
clojure)来实现：
单thread版本:
(defn fib [n]
(if (< n 2)
n
(apply + (map fib [(dec n) (- n 2)]))))
(time (fib 32)); "Elapsed time: 1851.765 msecs"
并行版本:
(defn pfib [n m]
(if (< n 2)
n
(if (> (- m n) 4)
(apply + (map fib [(dec n) (- n 2)]))
(apply + (pmap #(pfib % m) [(dec n) (- n 2)]))
)))
(time (pfib 32 32)); "Elapsed time: 1170.566 msecs"
用clojure写并行很容易，将map改为pmap就可以了。从速度上看，提高了接近1倍... 阅读全帖 |
|
d***a
发帖数: 13752 | 6 仔细看看丰田是怎样测试的。简单的说，就是做半径80米转圈测试，不断加速，直到车
oversteer，得到结果，Torsen-C比理论上零时间反应的打滑电控系统还好。
为什么会这样呢？因为Torsen的设计，在转圈时会把Torque更多地分配给外轮，所以比
对比系统的理论值还好。
问题何在呢？高级的电控系统，包括SH-AWD，xDrive，4matic，在车拐弯时会自动把
Torque分配给外轮，甚至可以比Torsen分配得更多。它们使用yaw sensor，不等打滑就
知道车子在拐弯。
那个图表画得也有问题，把差别放大了。仔细看，70ms延迟的打滑电控系统是58kmh时
滑出去，Torsen-C在63kmh时滑出去，只有5kmh或3.5mph的差别。
这样做研究，太黑了。
丰田的测试方法：
http://www.keepandshare.com/doc/2146555/20090122-tranfer-case-d
Fig. 5 shows the results of a response analysis of the distribution control
(yaw rate ... 阅读全帖 |
|
x*******i
发帖数: 777 | 7 Write a program for mixing soundfiles, with the following
characteristics:
写一个音频合并的程序，包括以下内容
(a) accepting any soundfile formats supported by libsndfile
输入的音频文件的格式要被libsndfile支持。
libsndfile ：a crossplatform library for soundfile manipulation.
(b) taking only uncompressed PCM format, in any (integer or floating-
point)
precision (8-bit (signed/unsigned), 16-bit, 24-bit, 32-bit, floats,
doubles).
接受未经压缩过的PCM编码格式的音频，这些音频的量化精度为浮点数或者整数（每个
sample分为
1 -100 就是整数，每个sample的值只有100个可能； 每个sample 分为1.00-10... 阅读全帖 |
|
t****a
发帖数: 1212 | 8 第三题很有趣，偶来给个clojure的解
基本思路是
1. 对T做预处理，构建graph，同时，记录每个字符a-z所出现的位置
2. 由于S长度<100，数量却很大(100M)，意味着有很多重复，使用memoize function会
大大减少计算量
下面的程序在1/1000大小的S数据集上运行需要1.7秒。观察了一下时间和数据集大小趋
势，在完整数据集上几分钟内也可以产生结果
(def alphabet (map char (range (int \a) (inc (int \z)))))
(def T (vec (take 10000000 (repeatedly #(rand-nth alphabet)))))
(def S (take 100000 (repeatedly (fn [] (take (inc (int (* (rand) 100))) (
repeatedly #(rand-nth alphabet)))))))
(def T-ix (group-by #(T %) (range (count T))))
(def T-map (let [pairs (dist... 阅读全帖 |
|
t****a
发帖数: 1212 | 9 第三题很有趣，偶来给个clojure的解
基本思路是
1. 对T做预处理，构建graph，同时，记录每个字符a-z所出现的位置
2. 由于S长度<100，数量却很大(100M)，意味着有很多重复，使用memoize function会
大大减少计算量
下面的程序在1/1000大小的S数据集上运行需要1.7秒。观察了一下时间和数据集大小趋
势，在完整数据集上几分钟内也可以产生结果
(def alphabet (map char (range (int \a) (inc (int \z)))))
(def T (vec (take 10000000 (repeatedly #(rand-nth alphabet)))))
(def S (take 100000 (repeatedly (fn [] (take (inc (int (* (rand) 100))) (
repeatedly #(rand-nth alphabet)))))))
(def T-ix (group-by #(T %) (range (count T))))
(def T-map (let [pairs (dist... 阅读全帖 |
|
p*****3
发帖数: 488 | 10
抛个砖头吧。
假设一台机器精确到1msec, 每秒可处理1000个request, 那么10000/s需要10台机器。
首先需要同步10台机器的时间(GPS之类的)，否则时间都不一样没法保证先后顺序啊。
既然每秒10000个request, 那么可能在同一毫秒时间内不同机器同时收到request, 那
么就这么编号:这样如果时间顺序有差不管在同
一台机器还是不同机器上id顺序都可以保证，时间相同id也不会重复,还是
decentralized,根据request rate来看，不说完全不浪费，id不会跳跃太大。 time
synchronize也不用每个request就同步一次，10分钟同步一次。再说那些request时间
差个1，2毫秒估计也没啥问题，网络延迟啥的比这玩意厉害多了，没必要那么strict |
|
p*****3
发帖数: 488 | 11
如果以毫秒为最小单位的话建一个int a[1000]，就是一个size是1000的循环队列, 维
护一个count = 0; 和一个curIt = msec%1000;
如果每个毫秒都有call的话 ==> 如果curIt改变了， count = count - a[curIt], 如
果call的不是很频繁，可能需要clear多个slot, 但是amortize的效率还可以，可能他
说的提示就是这个意思 |
|
c***p
发帖数: 1042 | 12 有4个可以推荐,杂志是MSEC.
需要的xdjm, 站内发姓名和邮箱. |
|
c***p
发帖数: 1042 | 13 杂志是MSEC, 需要的站内发姓名和工作邮箱. |
|
G***a
发帖数: 27294 | 14 不是：）思维是没有＂语言＂的，其实根本没有＂用＊＊语思考＂这种事情存在
能formulate出一句话，就已经进入phonological loop，比真正思维已经至少慢了１０
０多msec了
这只能证明你那天早上英文的semantic map更加active一点
不能证明你思考模式，fundamentally改变了。 |
|
b*********s
发帖数: 6757 | 15 "另外很多4.0/4。5的wsn球速也很快"
sure, they are capable of, but not for average rally speed. Kinda hard to
quantify this, (reaction time + stroke time + movement time) / (ball travel
time), my assumption is reaction time and stroke time is in unit of mSec,
where is movement time and ball travel time is in unit of Sec. Would be
interesting if you can create a model or do a thesis on this. :)
With that said, I saw the HL of Halep vs JJ. I thought both had more
opportunity to run around. I would guess this h... 阅读全帖 |
|
a*****s
发帖数: 2663 | 16 一般人都是DIY，比如用麦克风trigger快门，当然由于sub-msec快门和flash同步很难
，你可能需要在dark room里面完成拍摄。 |
|
b***i
发帖数: 3043 | 17 start=System.currentTimeMillis();
while (total+200
Thread.sleep(200);
Program.UpdateOutput();
System.out.println("waited " + total);
if ((pgm.demo && GUIWindow.demo2End) ||
(GUIWindow.emergency && ! pgm.demo))
break;
total = System.currentTimeMillis()-start;
这样的代码有问题否？
会不会total变成负数？比如start = 一个很大的数，然后System.currentTimeMillis(
)的新值变刚过0？ |
|
y**b
发帖数: 10166 | 18 请问一般用什么通用的方法和工具？
有几个问题请教：
（1）PAPI_MEM_SCY(Cycles Stalled Waiting for Memory Access,好处是可以嵌在代
码里面，尤其适合并行计算)似乎不支持sandybridge/ivybridge/haswell几种处理器，
是不是意味着基于PAPI的TAU,PerfSuite,HPCToolkit也没法做这种测试？
（2）用perf进行了测试，结果如下
perf stat -p 48382 sleep 10
Performance counter stats for process id '48382':
4821.604141 task-clock (msec) # 0.963 CPUs utilized
[100.00%]
1,218 context-switches # 0.253 K/sec
[100.00%]
0 ... 阅读全帖 |
|
y**b
发帖数: 10166 | 19 perf不错，好像不能测试mpi程序。请看这个例子：
perf stat -p 48382 sleep 10
Performance counter stats for process id '48382':
4821.604141 task-clock (msec) # 0.963 CPUs utilized
[100.00%]
1,218 context-switches # 0.253 K/sec
[100.00%]
0 cpu-migrations # 0.000 K/sec
[100.00%]
0 page-faults # 0.000 K/sec
17,312,623,873 cycles ... 阅读全帖 |
|
c*****s
发帖数: 18 | 20 【 以下文字转载自 Chemistry 讨论区 】
发信人: carinos (Carinos), 信区: Chemistry
标 题: PhD openings in Materials Science, Engineering, and Commercialization Program
发信站: BBS 未名空间站 (Mon Nov 14 00:52:52 2011, 美东)
PhD openings available in Materials Science, Engineering, and
Commercialization Program at Texas State University
Candidates will be trained in both traditional materials science and
engineering, and technology commercialization, thus potentially a better
chance for an industry job after graduation
Basic ... 阅读全帖 |
|
c*****s
发帖数: 18 | 21 PhD openings available in Materials Science, Engineering, and
Commercialization Program at Texas State University
Candidates will be trained in both traditional materials science and
engineering, and technology commercialization, thus potentially a better
chance for an industry job after graduation
Basic requirements: Master Degree by the end of 2011, GRE 1150 or higher,
TOEFL 78 or higher
Stipend: $32,000 (need pay ca. $7,000 tuition per year), thus net stipend:
ca. $25,000
Please see details ... 阅读全帖 |
|
s*****0
发帖数: 244 | 22 Ph.D. program in Materials science and Engineering at Texas State University
We have Ph.D. students openings for Materials science and Engineering at
Texas State University. The stipend is $32,000/year (need pay ca. $7,000
tuition per year), thus net stipend at ca. $25,000/year. See details at http://www.msec.txstate.edu/Ph-D-degree/admissions.html
Requirements:
1) Since this is a 3-year program, applicants should possess a Master’s
Degree from an accredited college or university in the field... 阅读全帖 |
|
c*****s
发帖数: 18 | 23 【 以下文字转载自 Chemistry 讨论区 】
发信人: carinos (Carinos), 信区: Chemistry
标 题: PhD openings in Materials Science, Engineering, and Commercialization Program
发信站: BBS 未名空间站 (Mon Nov 14 00:52:52 2011, 美东)
PhD openings available in Materials Science, Engineering, and
Commercialization Program at Texas State University
Candidates will be trained in both traditional materials science and
engineering, and technology commercialization, thus potentially a better
chance for an industry job after graduation
Basic ... 阅读全帖 |
|
W******g
发帖数: 143 | 24 Cardioverter Use
Selection of synchronized or nonsynchronized mode is the next critical step.
In the synchronized mode, the cardioverter searches for a large positive or
negative deflection, which it interprets as the R- or S-wave. It then
automatically discharges an electric current that lasts less than 4 msec,
avoiding the vulnerable period during repolarization when VF can be easily
induced. When the cardioverter is set to synchronize, a brief delay will
occur after the buttons are pushed for |
|
H****B
发帖数: 33 | 25 There is nothing wrong with you. Be confident. Do a long externship. It
will definately help. Get up early and push the button in the millisecond.
It is really something, you know. It is the msec that is a life changer
for most of us. Since the number that fits the interview selection criteria
is way more than enough, programs go down by the time of application
submission on the minutes. Remember, get up early.
Many matched the second time around. So, be confident. Relax. |
|
c*****s
发帖数: 18 | 26 【 以下文字转载自 Chemistry 讨论区 】
发信人: carinos (Carinos), 信区: Chemistry
标 题: PhD openings in Materials Science, Engineering, and Commercialization Program
发信站: BBS 未名空间站 (Mon Nov 14 00:52:52 2011, 美东)
PhD openings available in Materials Science, Engineering, and
Commercialization Program at Texas State University
Candidates will be trained in both traditional materials science and
engineering, and technology commercialization, thus potentially a better
chance for an industry job after graduation
Basic ... 阅读全帖 |
|
a*******n
发帖数: 82 | 27 78 yo female was transferred for evaluation of torsades. The patient has one
year history of AF and has been taking amiodarone and coumadine. For the
past 5 months, she has been feeling severe nausea. She was admitted to xxx
Hospital for evaluation. There, her Hgb was 7.8 and TSH was over 40. Then
Amiodarone was stopped for presumed toxicity to thyroid gland. And coumadin
was stopped for suspicious GI bleeding. Levothyroxine was started.She was
also given 2 units of pRBC. EGD was done and found ... 阅读全帖 |
|
a*******n
发帖数: 82 | 28 They are all valuable points!
She was given Zofran, Phenergan and compazine in the other hospital.
MRI of brain was done for evaluation of severe nausea.
For her PMH/PSH, significant for iron-deficiency anemia, treated with iron
supplements in the past; HTN; DLP; h/o C. diff colitis 6 months ago; pAF
with RVR 8 months ago; yearly normal mammogram; normal colonoscopy 3 years
ago; chronic lymphedema from lower extremities; lung nodule with stable CT
followup. She was on dig, amiodarone, coumadin, ... 阅读全帖 |
|
a*******n
发帖数: 82 | 29 78 yo female was transferred for evaluation of torsades. The patient has one
year history of AF and has been taking amiodarone and coumadine. For the
past 5 months, she has been feeling severe nausea. She was admitted to xxx
Hospital for evaluation. There, her Hgb was 7.8 and TSH was over 40. Then
Amiodarone was stopped for presumed toxicity to thyroid gland. And coumadin
was stopped for suspicious GI bleeding. Levothyroxine was started.She was
also given 2 units of pRBC. EGD was done and found ... 阅读全帖 |
|
a*******n
发帖数: 82 | 30 They are all valuable points!
She was given Zofran, Phenergan and compazine in the other hospital.
MRI of brain was done for evaluation of severe nausea.
For her PMH/PSH, significant for iron-deficiency anemia, treated with iron
supplements in the past; HTN; DLP; h/o C. diff colitis 6 months ago; pAF
with RVR 8 months ago; yearly normal mammogram; normal colonoscopy 3 years
ago; chronic lymphedema from lower extremities; lung nodule with stable CT
followup. She was on dig, amiodarone, coumadin, ... 阅读全帖 |
|
w*******y
发帖数: 60932 | 31 Solid State Drives > SATA SSD
OCZ TECHNOLOGY 120GB VERTEX PLUS CA ONLINE BACKUP
OCZ Vertex Plus 120GB 2.5" SATA II SSD Bundle
More OCZ Products
Be the first to write a review
0
inShare
Item Number: O261-6401 OB
Model:
Expiration Dates Vary
See Terms & Conditions.
Restrictions Apply. NOTICE: Enrollment in ARS is required to qualify for
rebate. MAIL IN REBATE NOT REQUIRED See form
Shipping: In Stock
Details
List Price: $194.99
Instant Savings: - $45.00
Pric... 阅读全帖 |
|
|
|
