Search Results

Search found 1226 results on 50 pages for 'improvement'.

Page 1/50 | 1 2 3 4 5 6 7 8 9 10 11 12  | Next Page >

  • A toolset for self improvement and learning [closed]

    - by Sebastian
    Possible Duplicate: I’m having trouble learning I've been working as an IT consultant for 1½ years and I am very passionate about programming. Before that I studied MSc Software Engineering and had both a part time job as a developer for a big telecom company. During that time I also took extra courses and earned a SCJP certificate. I have been continuously reading a lot of books during the last 3½ years. Now to my problem. I want to continue learning and become a really, really good developer. Apart from my daytime job as a full time java developer I have taken university courses in, for me, new languages and paradigms. Most recently, android game development and then functional programming with Scala. I've read books, went to conferences and had a couple of presentations for internal training purposes in our local office. I want to have some advice from other people who have previously been in my situation or currently are. What are you guys doing to keep improving yourselves? Here is some things that I have found are working for me: Reading books I've mostly read books about best practices for programming, OO-design, refactoring, design patterns, tdd. Software craftmanship if you like. I keep a reading list and my current book is Apprenticeship patterns. Taking courses In my country we have a really good system for taking online distance courses. I have also taken one course at coursera.org and a highly recommend that platform. Ive looked at courses at oreilly.com, industriallogic, javaspecialists.eu and they seem to be okay. If someone gives these type of courses a really good review, I can probably convince my boss. Workshops that span over a couple of days would probably be harder, but Ive seen that uncle Bob will have one about refactoring and tdd in 6months not far from here.. :) Are their possibly some online learning platforms that I dont know about? Educational videos I've bought uncle bobs videos from cleancoders.com and I highly recommend them. The only thing I dont like is that they are quite expensive and that he talks about astronomy for ~10 minutes in every episode. Getting certified I had a lot of fun and learned a lot when I studied for the SCJP. I have also done some preparation for the microsoft equivalent but never went for it. I think it is a good when selling yourself as a newly graduated student and also will boost your knowledge if your are interested in it. Now I would like others to start sharing their experiences and possibly give me some advice! BR Sebastian

    Read the article

  • Time management and self improvement

    - by Filip
    Hi, I hope I can open a discussion on this topic as this is not a specific problem. It's a topic I hope to get some ideas on how people in similar situation as mine manage their time. OK, I'm a single developer on a software project for the last 6-8 months. The project I'm working on uses several technologies, mainly .net stuff: WPF, WF, NHibernate, WCF, MySql and other third party SDKs relevant for the project nature. My experience and knowledge vary, for example I have a lot of experience in WPF but much less in WCF. I work full time on the project and im curios on how other programmers which need to multi task in many areas manage their time. I'm a very applied type of person and prefer to code instead of doing research. I feel that doing research "might" slow down the progress of the project while I recognize that research and learning more in areas which I'm not so strong will ultimately make me more productive. How would you split up your daily time in productive coding time and time to and experiment, read blogs, go through tutorials etc. I would say that Im coding about 90%+ of my day and devoting some but very little time in research and acquiring new knowledge.

    Read the article

  • Time management and self improvement

    - by Filip
    I hope I can open a discussion on this topic as this is not a specific problem. It's a topic I hope to get some ideas on how people in similar situation as mine manage their time. OK, I'm a single developer on a software project for the last 6-8 months. The project I'm working on uses several technologies, mainly .net stuff: WPF, WF, NHibernate, WCF, MySql and other third party SDKs relevant for the project nature. My experience and knowledge vary, for example I have a lot of experience in WPF but much less in WCF. I work full time on the project and im curios on how other programmers which need to multi task in many areas manage their time. I'm a very applied type of person and prefer to code instead of doing research. I feel that doing research "might" slow down the progress of the project while I recognize that research and learning more in areas which I'm not so strong will ultimately make me more productive. How would you split up your daily time in productive coding time and time to and experiment, read blogs, go through tutorials etc. I would say that Im coding about 90%+ of my day and devoting some but very little time in research and acquiring new knowledge. Thanks for your replies. I think I will adopt a gradual transition to Dominics block parts. I kinda knew that coding was taking up way to much of my time but it feels good having a first version of the project completed and ready. With a few months of focused hard work behind me I hope to get more time to experiment and expand my knowlegde. Now I only hope my boss will cut me some slack and stop pressuring me for features...

    Read the article

  • Skillset improvement in coming new year

    - by exiter2000
    Here is a little background information. I have been working for Java 10 years. The product I am working on went to live about 3 years ago. Now, the product is getting stable. After all the post-product drama, I gained a lot of knowledge about Oracle & SQL. People(mainly management) were desperated enough to give me deep oracle-related task over DBAs. I admit I considered becoming DBA but eventually decided to remain as a programmer. DBAs & Management are demanding all the DB & Query related task back to DBA, which makes me a bit sad. In short, I anticipate a lot of time next year. What would you do to improve your skillset?? I am thinking to upgrade my Java version(Not from experience though, we are using JDK1.5) to 1.6 getting certificate. Any good idea from fellow developers?? -----------Edit --------------------- How about data modeling for application? Do you guys think it is developer role??

    Read the article

  • What was the most productive improvement suggestion you ever made for your team

    - by questzen
    I suggested the testing and functional teams to use Freemind map for jotting the functional flows and test steps. There was some paranoia but our module took it up and the QA teams were surprised to see near zero review comments. There was misconception among the team that there are doing more work. I assured them that by the time others would complete their work along with comment fixes, we would be going out for team lunchs and we did. The real returns came when the developers started refering to the created document in their discussions. So share your contribution(s).

    Read the article

  • SOA, Empowerment and Continuous Improvement

    - by Tanu Sood
    Normal 0 false false false EN-US X-NONE X-NONE /* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-qformat:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-family:"Calibri","sans-serif"; mso-bidi-font-family:"Times New Roman";} Rick Beers is Senior Director of Product Management for Oracle Fusion Middleware. Prior to joining Oracle, Rick held a variety of executive operational positions at Corning, Inc. and Bausch & Lomb. With a professional background that includes senior management positions in manufacturing, supply chain and information technology, Rick brings a unique set of experiences to cover the impact that technology can have on business models, processes and organizations. Rick will be hosting the IT Leader Editorial on a regular basis. I met my twin at Open World. We share backgrounds, experiences and even names. I hosted an invitation-only AppAdvantage Leadership Forum with an overcapacity 85 participants: 55 customers, 15 from the Oracle AppAdvantage team and 15 Partners. It was a lively, open and positive discussion of pace layered architectures and Oracle’s AppAdvantage approach to a unified view of Applications and Middleware. Rick Hassman from Pella was one of the customer panelists and during the pre event prep, Rick and I shared backgrounds and found that we had both been plant managers and led ERP deployments prior to leading IT itself. During the panel conversation I explored this with him, discussing the unique perspectives that this provides to CIO’s. He then hit on a point that I wasn’t able to fully appreciate until a week later. First though, some background. The week after the Forum, one of the participants emailed me with the following thoughts: “I am 150% behind this concept……but we are struggling with the concept of web services and the potential use of the Oracle Service Bus technology let alone moving into using the full SOA/BPM/BAM software to extend our JD Edwards application to both integrate and support business processes”. After thinking a bit I responded this way: While I certainly appreciate the degree of change and effort involved, perhaps I could offer the following: One of the underlying principles behind Oracle AppAdvantage is that more often than not, the choice between changing a business process and invasively customizing ERP represents a Hobson's Choice: neither is acceptable. In this case the third option, moving the process out of ERP, is the only acceptable one. Providing this choice typically requires end to end, real time interoperability across applications and/or services. This real time interoperability, to be sustainable over time requires a service oriented architecture. There's just no way around this. SOA adaptation is admittedly tough at the beginning. New skills, new technology and new headaches. But, like any radically new technology, it has a learning curve that drives cost down rather dramatically over time. Tough choices to be sure, but not entirely different than we face with every major technology cycle. Good points of course, but I felt that something was missing. The points were convincing, perhaps even a bit insightful, but they didn’t get at the heart of what Oracle AppAdvantage is focused upon: how the optimization of technology, applications, processes and relationships can change the very way that organizations operate. And then I thought back to the panel discussion with Rick Hassman at Oracle OpenWorld. Rick stressed that Continuous Improvement is a fundamental business strategy at Pella. I remember Continuous Improvement well as I suspect does everyone who was in American manufacturing during the 80’s. Pioneered by W. Edwards Deming in Japan (and still known alternatively as Kaizen), Continuous Improvement sets in place the business culture that we must not become complacent with success and resistant to the ongoing need for change. Many believe that this single handedly drove the renaissance in American manufacturing through the last two decades, which had become complacent during the 70’s and early 80’s. But what exactly does this have to do with SOA? It was Rick’s next point. He drew the connection that moving those business processes that need to continually change over time out of ERP and into edge applications and services enables continuous improvement by empowering people to continually strive for better ways of doing things rather than be being bound by workflows that cannot change. A compelling connection: that SOA, and the overall Oracle AppAdvantage framework of which it is an integral part, can empower people towards continuous improvement in business processes and as a result drive business leadership and business excellence. What better a case for technology innovation?

    Read the article

  • How to evaluate SEO/prominence improvement [on hold]

    - by Rober
    I will work on a website SEO and before starting with it I would like to "take a snapshot" of the present status so that I will be able to compare it with the new situation in a few months and evaluate my work and the real improvement. I don't mean whether the website is well implemented or not, but how well it is seen by Google and others. What prominence it has. I am taking some variables from Google Analytics (average day visits...), from Google Webmaster Tools (Search traffic and average position...) and some other indicators, like automatic SEO audit figures (website estimated worth, real pagerank...). What would you look at before starting SEO improvement?

    Read the article

  • How John Got 15x Improvement Without Really Trying

    - by rchrd
    The following article was published on a Sun Microsystems website a number of years ago by John Feo. It is still useful and worth preserving. So I'm republishing it here.  How I Got 15x Improvement Without Really Trying John Feo, Sun Microsystems Taking ten "personal" program codes used in scientific and engineering research, the author was able to get from 2 to 15 times performance improvement easily by applying some simple general optimization techniques. Introduction Scientific research based on computer simulation depends on the simulation for advancement. The research can advance only as fast as the computational codes can execute. The codes' efficiency determines both the rate and quality of results. In the same amount of time, a faster program can generate more results and can carry out a more detailed simulation of physical phenomena than a slower program. Highly optimized programs help science advance quickly and insure that monies supporting scientific research are used as effectively as possible. Scientific computer codes divide into three broad categories: ISV, community, and personal. ISV codes are large, mature production codes developed and sold commercially. The codes improve slowly over time both in methods and capabilities, and they are well tuned for most vendor platforms. Since the codes are mature and complex, there are few opportunities to improve their performance solely through code optimization. Improvements of 10% to 15% are typical. Examples of ISV codes are DYNA3D, Gaussian, and Nastran. Community codes are non-commercial production codes used by a particular research field. Generally, they are developed and distributed by a single academic or research institution with assistance from the community. Most users just run the codes, but some develop new methods and extensions that feed back into the general release. The codes are available on most vendor platforms. Since these codes are younger than ISV codes, there are more opportunities to optimize the source code. Improvements of 50% are not unusual. Examples of community codes are AMBER, CHARM, BLAST, and FASTA. Personal codes are those written by single users or small research groups for their own use. These codes are not distributed, but may be passed from professor-to-student or student-to-student over several years. They form the primordial ocean of applications from which community and ISV codes emerge. Government research grants pay for the development of most personal codes. This paper reports on the nature and performance of this class of codes. Over the last year, I have looked at over two dozen personal codes from more than a dozen research institutions. The codes cover a variety of scientific fields, including astronomy, atmospheric sciences, bioinformatics, biology, chemistry, geology, and physics. The sources range from a few hundred lines to more than ten thousand lines, and are written in Fortran, Fortran 90, C, and C++. For the most part, the codes are modular, documented, and written in a clear, straightforward manner. They do not use complex language features, advanced data structures, programming tricks, or libraries. I had little trouble understanding what the codes did or how data structures were used. Most came with a makefile. Surprisingly, only one of the applications is parallel. All developers have access to parallel machines, so availability is not an issue. Several tried to parallelize their applications, but stopped after encountering difficulties. Lack of education and a perception that parallelism is difficult prevented most from trying. I parallelized several of the codes using OpenMP, and did not judge any of the codes as difficult to parallelize. Even more surprising than the lack of parallelism is the inefficiency of the codes. I was able to get large improvements in performance in a matter of a few days applying simple optimization techniques. Table 1 lists ten representative codes [names and affiliation are omitted to preserve anonymity]. Improvements on one processor range from 2x to 15.5x with a simple average of 4.75x. I did not use sophisticated performance tools or drill deep into the program's execution character as one would do when tuning ISV or community codes. Using only a profiler and source line timers, I identified inefficient sections of code and improved their performance by inspection. The changes were at a high level. I am sure there is another factor of 2 or 3 in each code, and more if the codes are parallelized. The study’s results show that personal scientific codes are running many times slower than they should and that the problem is pervasive. Computational scientists are not sloppy programmers; however, few are trained in the art of computer programming or code optimization. I found that most have a working knowledge of some programming language and standard software engineering practices; but they do not know, or think about, how to make their programs run faster. They simply do not know the standard techniques used to make codes run faster. In fact, they do not even perceive that such techniques exist. The case studies described in this paper show that applying simple, well known techniques can significantly increase the performance of personal codes. It is important that the scientific community and the Government agencies that support scientific research find ways to better educate academic scientific programmers. The inefficiency of their codes is so bad that it is retarding both the quality and progress of scientific research. # cacheperformance redundantoperations loopstructures performanceimprovement 1 x x 15.5 2 x 2.8 3 x x 2.5 4 x 2.1 5 x x 2.0 6 x 5.0 7 x 5.8 8 x 6.3 9 2.2 10 x x 3.3 Table 1 — Area of improvement and performance gains of 10 codes The remainder of the paper is organized as follows: sections 2, 3, and 4 discuss the three most common sources of inefficiencies in the codes studied. These are cache performance, redundant operations, and loop structures. Each section includes several examples. The last section summaries the work and suggests a possible solution to the issues raised. Optimizing cache performance Commodity microprocessor systems use caches to increase memory bandwidth and reduce memory latencies. Typical latencies from processor to L1, L2, local, and remote memory are 3, 10, 50, and 200 cycles, respectively. Moreover, bandwidth falls off dramatically as memory distances increase. Programs that do not use cache effectively run many times slower than programs that do. When optimizing for cache, the biggest performance gains are achieved by accessing data in cache order and reusing data to amortize the overhead of cache misses. Secondary considerations are prefetching, associativity, and replacement; however, the understanding and analysis required to optimize for the latter are probably beyond the capabilities of the non-expert. Much can be gained simply by accessing data in the correct order and maximizing data reuse. 6 out of the 10 codes studied here benefited from such high level optimizations. Array Accesses The most important cache optimization is the most basic: accessing Fortran array elements in column order and C array elements in row order. Four of the ten codes—1, 2, 4, and 10—got it wrong. Compilers will restructure nested loops to optimize cache performance, but may not do so if the loop structure is too complex, or the loop body includes conditionals, complex addressing, or function calls. In code 1, the compiler failed to invert a key loop because of complex addressing do I = 0, 1010, delta_x IM = I - delta_x IP = I + delta_x do J = 5, 995, delta_x JM = J - delta_x JP = J + delta_x T1 = CA1(IP, J) + CA1(I, JP) T2 = CA1(IM, J) + CA1(I, JM) S1 = T1 + T2 - 4 * CA1(I, J) CA(I, J) = CA1(I, J) + D * S1 end do end do In code 2, the culprit is conditionals do I = 1, N do J = 1, N If (IFLAG(I,J) .EQ. 0) then T1 = Value(I, J-1) T2 = Value(I-1, J) T3 = Value(I, J) T4 = Value(I+1, J) T5 = Value(I, J+1) Value(I,J) = 0.25 * (T1 + T2 + T5 + T4) Delta = ABS(T3 - Value(I,J)) If (Delta .GT. MaxDelta) MaxDelta = Delta endif enddo enddo I fixed both programs by inverting the loops by hand. Code 10 has three-dimensional arrays and triply nested loops. The structure of the most computationally intensive loops is too complex to invert automatically or by hand. The only practical solution is to transpose the arrays so that the dimension accessed by the innermost loop is in cache order. The arrays can be transposed at construction or prior to entering a computationally intensive section of code. The former requires all array references to be modified, while the latter is cost effective only if the cost of the transpose is amortized over many accesses. I used the second approach to optimize code 10. Code 5 has four-dimensional arrays and loops are nested four deep. For all of the reasons cited above the compiler is not able to restructure three key loops. Assume C arrays and let the four dimensions of the arrays be i, j, k, and l. In the original code, the index structure of the three loops is L1: for i L2: for i L3: for i for l for l for j for k for j for k for j for k for l So only L3 accesses array elements in cache order. L1 is a very complex loop—much too complex to invert. I brought the loop into cache alignment by transposing the second and fourth dimensions of the arrays. Since the code uses a macro to compute all array indexes, I effected the transpose at construction and changed the macro appropriately. The dimensions of the new arrays are now: i, l, k, and j. L3 is a simple loop and easily inverted. L2 has a loop-carried scalar dependence in k. By promoting the scalar name that carries the dependence to an array, I was able to invert the third and fourth subloops aligning the loop with cache. Code 5 is by far the most difficult of the four codes to optimize for array accesses; but the knowledge required to fix the problems is no more than that required for the other codes. I would judge this code at the limits of, but not beyond, the capabilities of appropriately trained computational scientists. Array Strides When a cache miss occurs, a line (64 bytes) rather than just one word is loaded into the cache. If data is accessed stride 1, than the cost of the miss is amortized over 8 words. Any stride other than one reduces the cost savings. Two of the ten codes studied suffered from non-unit strides. The codes represent two important classes of "strided" codes. Code 1 employs a multi-grid algorithm to reduce time to convergence. The grids are every tenth, fifth, second, and unit element. Since time to convergence is inversely proportional to the distance between elements, coarse grids converge quickly providing good starting values for finer grids. The better starting values further reduce the time to convergence. The downside is that grids of every nth element, n > 1, introduce non-unit strides into the computation. In the original code, much of the savings of the multi-grid algorithm were lost due to this problem. I eliminated the problem by compressing (copying) coarse grids into continuous memory, and rewriting the computation as a function of the compressed grid. On convergence, I copied the final values of the compressed grid back to the original grid. The savings gained from unit stride access of the compressed grid more than paid for the cost of copying. Using compressed grids, the loop from code 1 included in the previous section becomes do j = 1, GZ do i = 1, GZ T1 = CA(i+0, j-1) + CA(i-1, j+0) T4 = CA1(i+1, j+0) + CA1(i+0, j+1) S1 = T1 + T4 - 4 * CA1(i+0, j+0) CA(i+0, j+0) = CA1(i+0, j+0) + DD * S1 enddo enddo where CA and CA1 are compressed arrays of size GZ. Code 7 traverses a list of objects selecting objects for later processing. The labels of the selected objects are stored in an array. The selection step has unit stride, but the processing steps have irregular stride. A fix is to save the parameters of the selected objects in temporary arrays as they are selected, and pass the temporary arrays to the processing functions. The fix is practical if the same parameters are used in selection as in processing, or if processing comprises a series of distinct steps which use overlapping subsets of the parameters. Both conditions are true for code 7, so I achieved significant improvement by copying parameters to temporary arrays during selection. Data reuse In the previous sections, we optimized for spatial locality. It is also important to optimize for temporal locality. Once read, a datum should be used as much as possible before it is forced from cache. Loop fusion and loop unrolling are two techniques that increase temporal locality. Unfortunately, both techniques increase register pressure—as loop bodies become larger, the number of registers required to hold temporary values grows. Once register spilling occurs, any gains evaporate quickly. For multiprocessors with small register sets or small caches, the sweet spot can be very small. In the ten codes presented here, I found no opportunities for loop fusion and only two opportunities for loop unrolling (codes 1 and 3). In code 1, unrolling the outer and inner loop one iteration increases the number of result values computed by the loop body from 1 to 4, do J = 1, GZ-2, 2 do I = 1, GZ-2, 2 T1 = CA1(i+0, j-1) + CA1(i-1, j+0) T2 = CA1(i+1, j-1) + CA1(i+0, j+0) T3 = CA1(i+0, j+0) + CA1(i-1, j+1) T4 = CA1(i+1, j+0) + CA1(i+0, j+1) T5 = CA1(i+2, j+0) + CA1(i+1, j+1) T6 = CA1(i+1, j+1) + CA1(i+0, j+2) T7 = CA1(i+2, j+1) + CA1(i+1, j+2) S1 = T1 + T4 - 4 * CA1(i+0, j+0) S2 = T2 + T5 - 4 * CA1(i+1, j+0) S3 = T3 + T6 - 4 * CA1(i+0, j+1) S4 = T4 + T7 - 4 * CA1(i+1, j+1) CA(i+0, j+0) = CA1(i+0, j+0) + DD * S1 CA(i+1, j+0) = CA1(i+1, j+0) + DD * S2 CA(i+0, j+1) = CA1(i+0, j+1) + DD * S3 CA(i+1, j+1) = CA1(i+1, j+1) + DD * S4 enddo enddo The loop body executes 12 reads, whereas as the rolled loop shown in the previous section executes 20 reads to compute the same four values. In code 3, two loops are unrolled 8 times and one loop is unrolled 4 times. Here is the before for (k = 0; k < NK[u]; k++) { sum = 0.0; for (y = 0; y < NY; y++) { sum += W[y][u][k] * delta[y]; } backprop[i++]=sum; } and after code for (k = 0; k < KK - 8; k+=8) { sum0 = 0.0; sum1 = 0.0; sum2 = 0.0; sum3 = 0.0; sum4 = 0.0; sum5 = 0.0; sum6 = 0.0; sum7 = 0.0; for (y = 0; y < NY; y++) { sum0 += W[y][0][k+0] * delta[y]; sum1 += W[y][0][k+1] * delta[y]; sum2 += W[y][0][k+2] * delta[y]; sum3 += W[y][0][k+3] * delta[y]; sum4 += W[y][0][k+4] * delta[y]; sum5 += W[y][0][k+5] * delta[y]; sum6 += W[y][0][k+6] * delta[y]; sum7 += W[y][0][k+7] * delta[y]; } backprop[k+0] = sum0; backprop[k+1] = sum1; backprop[k+2] = sum2; backprop[k+3] = sum3; backprop[k+4] = sum4; backprop[k+5] = sum5; backprop[k+6] = sum6; backprop[k+7] = sum7; } for one of the loops unrolled 8 times. Optimizing for temporal locality is the most difficult optimization considered in this paper. The concepts are not difficult, but the sweet spot is small. Identifying where the program can benefit from loop unrolling or loop fusion is not trivial. Moreover, it takes some effort to get it right. Still, educating scientific programmers about temporal locality and teaching them how to optimize for it will pay dividends. Reducing instruction count Execution time is a function of instruction count. Reduce the count and you usually reduce the time. The best solution is to use a more efficient algorithm; that is, an algorithm whose order of complexity is smaller, that converges quicker, or is more accurate. Optimizing source code without changing the algorithm yields smaller, but still significant, gains. This paper considers only the latter because the intent is to study how much better codes can run if written by programmers schooled in basic code optimization techniques. The ten codes studied benefited from three types of "instruction reducing" optimizations. The two most prevalent were hoisting invariant memory and data operations out of inner loops. The third was eliminating unnecessary data copying. The nature of these inefficiencies is language dependent. Memory operations The semantics of C make it difficult for the compiler to determine all the invariant memory operations in a loop. The problem is particularly acute for loops in functions since the compiler may not know the values of the function's parameters at every call site when compiling the function. Most compilers support pragmas to help resolve ambiguities; however, these pragmas are not comprehensive and there is no standard syntax. To guarantee that invariant memory operations are not executed repetitively, the user has little choice but to hoist the operations by hand. The problem is not as severe in Fortran programs because in the absence of equivalence statements, it is a violation of the language's semantics for two names to share memory. Codes 3 and 5 are C programs. In both cases, the compiler did not hoist all invariant memory operations from inner loops. Consider the following loop from code 3 for (y = 0; y < NY; y++) { i = 0; for (u = 0; u < NU; u++) { for (k = 0; k < NK[u]; k++) { dW[y][u][k] += delta[y] * I1[i++]; } } } Since dW[y][u] can point to the same memory space as delta for one or more values of y and u, assignment to dW[y][u][k] may change the value of delta[y]. In reality, dW and delta do not overlap in memory, so I rewrote the loop as for (y = 0; y < NY; y++) { i = 0; Dy = delta[y]; for (u = 0; u < NU; u++) { for (k = 0; k < NK[u]; k++) { dW[y][u][k] += Dy * I1[i++]; } } } Failure to hoist invariant memory operations may be due to complex address calculations. If the compiler can not determine that the address calculation is invariant, then it can hoist neither the calculation nor the associated memory operations. As noted above, code 5 uses a macro to address four-dimensional arrays #define MAT4D(a,q,i,j,k) (double *)((a)->data + (q)*(a)->strides[0] + (i)*(a)->strides[3] + (j)*(a)->strides[2] + (k)*(a)->strides[1]) The macro is too complex for the compiler to understand and so, it does not identify any subexpressions as loop invariant. The simplest way to eliminate the address calculation from the innermost loop (over i) is to define a0 = MAT4D(a,q,0,j,k) before the loop and then replace all instances of *MAT4D(a,q,i,j,k) in the loop with a0[i] A similar problem appears in code 6, a Fortran program. The key loop in this program is do n1 = 1, nh nx1 = (n1 - 1) / nz + 1 nz1 = n1 - nz * (nx1 - 1) do n2 = 1, nh nx2 = (n2 - 1) / nz + 1 nz2 = n2 - nz * (nx2 - 1) ndx = nx2 - nx1 ndy = nz2 - nz1 gxx = grn(1,ndx,ndy) gyy = grn(2,ndx,ndy) gxy = grn(3,ndx,ndy) balance(n1,1) = balance(n1,1) + (force(n2,1) * gxx + force(n2,2) * gxy) * h1 balance(n1,2) = balance(n1,2) + (force(n2,1) * gxy + force(n2,2) * gyy)*h1 end do end do The programmer has written this loop well—there are no loop invariant operations with respect to n1 and n2. However, the loop resides within an iterative loop over time and the index calculations are independent with respect to time. Trading space for time, I precomputed the index values prior to the entering the time loop and stored the values in two arrays. I then replaced the index calculations with reads of the arrays. Data operations Ways to reduce data operations can appear in many forms. Implementing a more efficient algorithm produces the biggest gains. The closest I came to an algorithm change was in code 4. This code computes the inner product of K-vectors A(i) and B(j), 0 = i < N, 0 = j < M, for most values of i and j. Since the program computes most of the NM possible inner products, it is more efficient to compute all the inner products in one triply-nested loop rather than one at a time when needed. The savings accrue from reading A(i) once for all B(j) vectors and from loop unrolling. for (i = 0; i < N; i+=8) { for (j = 0; j < M; j++) { sum0 = 0.0; sum1 = 0.0; sum2 = 0.0; sum3 = 0.0; sum4 = 0.0; sum5 = 0.0; sum6 = 0.0; sum7 = 0.0; for (k = 0; k < K; k++) { sum0 += A[i+0][k] * B[j][k]; sum1 += A[i+1][k] * B[j][k]; sum2 += A[i+2][k] * B[j][k]; sum3 += A[i+3][k] * B[j][k]; sum4 += A[i+4][k] * B[j][k]; sum5 += A[i+5][k] * B[j][k]; sum6 += A[i+6][k] * B[j][k]; sum7 += A[i+7][k] * B[j][k]; } C[i+0][j] = sum0; C[i+1][j] = sum1; C[i+2][j] = sum2; C[i+3][j] = sum3; C[i+4][j] = sum4; C[i+5][j] = sum5; C[i+6][j] = sum6; C[i+7][j] = sum7; }} This change requires knowledge of a typical run; i.e., that most inner products are computed. The reasons for the change, however, derive from basic optimization concepts. It is the type of change easily made at development time by a knowledgeable programmer. In code 5, we have the data version of the index optimization in code 6. Here a very expensive computation is a function of the loop indices and so cannot be hoisted out of the loop; however, the computation is invariant with respect to an outer iterative loop over time. We can compute its value for each iteration of the computation loop prior to entering the time loop and save the values in an array. The increase in memory required to store the values is small in comparison to the large savings in time. The main loop in Code 8 is doubly nested. The inner loop includes a series of guarded computations; some are a function of the inner loop index but not the outer loop index while others are a function of the outer loop index but not the inner loop index for (j = 0; j < N; j++) { for (i = 0; i < M; i++) { r = i * hrmax; R = A[j]; temp = (PRM[3] == 0.0) ? 1.0 : pow(r, PRM[3]); high = temp * kcoeff * B[j] * PRM[2] * PRM[4]; low = high * PRM[6] * PRM[6] / (1.0 + pow(PRM[4] * PRM[6], 2.0)); kap = (R > PRM[6]) ? high * R * R / (1.0 + pow(PRM[4]*r, 2.0) : low * pow(R/PRM[6], PRM[5]); < rest of loop omitted > }} Note that the value of temp is invariant to j. Thus, we can hoist the computation for temp out of the loop and save its values in an array. for (i = 0; i < M; i++) { r = i * hrmax; TEMP[i] = pow(r, PRM[3]); } [N.B. – the case for PRM[3] = 0 is omitted and will be reintroduced later.] We now hoist out of the inner loop the computations invariant to i. Since the conditional guarding the value of kap is invariant to i, it behooves us to hoist the computation out of the inner loop, thereby executing the guard once rather than M times. The final version of the code is for (j = 0; j < N; j++) { R = rig[j] / 1000.; tmp1 = kcoeff * par[2] * beta[j] * par[4]; tmp2 = 1.0 + (par[4] * par[4] * par[6] * par[6]); tmp3 = 1.0 + (par[4] * par[4] * R * R); tmp4 = par[6] * par[6] / tmp2; tmp5 = R * R / tmp3; tmp6 = pow(R / par[6], par[5]); if ((par[3] == 0.0) && (R > par[6])) { for (i = 1; i <= imax1; i++) KAP[i] = tmp1 * tmp5; } else if ((par[3] == 0.0) && (R <= par[6])) { for (i = 1; i <= imax1; i++) KAP[i] = tmp1 * tmp4 * tmp6; } else if ((par[3] != 0.0) && (R > par[6])) { for (i = 1; i <= imax1; i++) KAP[i] = tmp1 * TEMP[i] * tmp5; } else if ((par[3] != 0.0) && (R <= par[6])) { for (i = 1; i <= imax1; i++) KAP[i] = tmp1 * TEMP[i] * tmp4 * tmp6; } for (i = 0; i < M; i++) { kap = KAP[i]; r = i * hrmax; < rest of loop omitted > } } Maybe not the prettiest piece of code, but certainly much more efficient than the original loop, Copy operations Several programs unnecessarily copy data from one data structure to another. This problem occurs in both Fortran and C programs, although it manifests itself differently in the two languages. Code 1 declares two arrays—one for old values and one for new values. At the end of each iteration, the array of new values is copied to the array of old values to reset the data structures for the next iteration. This problem occurs in Fortran programs not included in this study and in both Fortran 77 and Fortran 90 code. Introducing pointers to the arrays and swapping pointer values is an obvious way to eliminate the copying; but pointers is not a feature that many Fortran programmers know well or are comfortable using. An easy solution not involving pointers is to extend the dimension of the value array by 1 and use the last dimension to differentiate between arrays at different times. For example, if the data space is N x N, declare the array (N, N, 2). Then store the problem’s initial values in (_, _, 2) and define the scalar names new = 2 and old = 1. At the start of each iteration, swap old and new to reset the arrays. The old–new copy problem did not appear in any C program. In programs that had new and old values, the code swapped pointers to reset data structures. Where unnecessary coping did occur is in structure assignment and parameter passing. Structures in C are handled much like scalars. Assignment causes the data space of the right-hand name to be copied to the data space of the left-hand name. Similarly, when a structure is passed to a function, the data space of the actual parameter is copied to the data space of the formal parameter. If the structure is large and the assignment or function call is in an inner loop, then copying costs can grow quite large. While none of the ten programs considered here manifested this problem, it did occur in programs not included in the study. A simple fix is always to refer to structures via pointers. Optimizing loop structures Since scientific programs spend almost all their time in loops, efficient loops are the key to good performance. Conditionals, function calls, little instruction level parallelism, and large numbers of temporary values make it difficult for the compiler to generate tightly packed, highly efficient code. Conditionals and function calls introduce jumps that disrupt code flow. Users should eliminate or isolate conditionls to their own loops as much as possible. Often logical expressions can be substituted for if-then-else statements. For example, code 2 includes the following snippet MaxDelta = 0.0 do J = 1, N do I = 1, M < code omitted > Delta = abs(OldValue ? NewValue) if (Delta > MaxDelta) MaxDelta = Delta enddo enddo if (MaxDelta .gt. 0.001) goto 200 Since the only use of MaxDelta is to control the jump to 200 and all that matters is whether or not it is greater than 0.001, I made MaxDelta a boolean and rewrote the snippet as MaxDelta = .false. do J = 1, N do I = 1, M < code omitted > Delta = abs(OldValue ? NewValue) MaxDelta = MaxDelta .or. (Delta .gt. 0.001) enddo enddo if (MaxDelta) goto 200 thereby, eliminating the conditional expression from the inner loop. A microprocessor can execute many instructions per instruction cycle. Typically, it can execute one or more memory, floating point, integer, and jump operations. To be executed simultaneously, the operations must be independent. Thick loops tend to have more instruction level parallelism than thin loops. Moreover, they reduce memory traffice by maximizing data reuse. Loop unrolling and loop fusion are two techniques to increase the size of loop bodies. Several of the codes studied benefitted from loop unrolling, but none benefitted from loop fusion. This observation is not too surpising since it is the general tendency of programmers to write thick loops. As loops become thicker, the number of temporary values grows, increasing register pressure. If registers spill, then memory traffic increases and code flow is disrupted. A thick loop with many temporary values may execute slower than an equivalent series of thin loops. The biggest gain will be achieved if the thick loop can be split into a series of independent loops eliminating the need to write and read temporary arrays. I found such an occasion in code 10 where I split the loop do i = 1, n do j = 1, m A24(j,i)= S24(j,i) * T24(j,i) + S25(j,i) * U25(j,i) B24(j,i)= S24(j,i) * T25(j,i) + S25(j,i) * U24(j,i) A25(j,i)= S24(j,i) * C24(j,i) + S25(j,i) * V24(j,i) B25(j,i)= S24(j,i) * U25(j,i) + S25(j,i) * V25(j,i) C24(j,i)= S26(j,i) * T26(j,i) + S27(j,i) * U26(j,i) D24(j,i)= S26(j,i) * T27(j,i) + S27(j,i) * V26(j,i) C25(j,i)= S27(j,i) * S28(j,i) + S26(j,i) * U28(j,i) D25(j,i)= S27(j,i) * T28(j,i) + S26(j,i) * V28(j,i) end do end do into two disjoint loops do i = 1, n do j = 1, m A24(j,i)= S24(j,i) * T24(j,i) + S25(j,i) * U25(j,i) B24(j,i)= S24(j,i) * T25(j,i) + S25(j,i) * U24(j,i) A25(j,i)= S24(j,i) * C24(j,i) + S25(j,i) * V24(j,i) B25(j,i)= S24(j,i) * U25(j,i) + S25(j,i) * V25(j,i) end do end do do i = 1, n do j = 1, m C24(j,i)= S26(j,i) * T26(j,i) + S27(j,i) * U26(j,i) D24(j,i)= S26(j,i) * T27(j,i) + S27(j,i) * V26(j,i) C25(j,i)= S27(j,i) * S28(j,i) + S26(j,i) * U28(j,i) D25(j,i)= S27(j,i) * T28(j,i) + S26(j,i) * V28(j,i) end do end do Conclusions Over the course of the last year, I have had the opportunity to work with over two dozen academic scientific programmers at leading research universities. Their research interests span a broad range of scientific fields. Except for two programs that relied almost exclusively on library routines (matrix multiply and fast Fourier transform), I was able to improve significantly the single processor performance of all codes. Improvements range from 2x to 15.5x with a simple average of 4.75x. Changes to the source code were at a very high level. I did not use sophisticated techniques or programming tools to discover inefficiencies or effect the changes. Only one code was parallel despite the availability of parallel systems to all developers. Clearly, we have a problem—personal scientific research codes are highly inefficient and not running parallel. The developers are unaware of simple optimization techniques to make programs run faster. They lack education in the art of code optimization and parallel programming. I do not believe we can fix the problem by publishing additional books or training manuals. To date, the developers in questions have not studied the books or manual available, and are unlikely to do so in the future. Short courses are a possible solution, but I believe they are too concentrated to be much use. The general concepts can be taught in a three or four day course, but that is not enough time for students to practice what they learn and acquire the experience to apply and extend the concepts to their codes. Practice is the key to becoming proficient at optimization. I recommend that graduate students be required to take a semester length course in optimization and parallel programming. We would never give someone access to state-of-the-art scientific equipment costing hundreds of thousands of dollars without first requiring them to demonstrate that they know how to use the equipment. Yet the criterion for time on state-of-the-art supercomputers is at most an interesting project. Requestors are never asked to demonstrate that they know how to use the system, or can use the system effectively. A semester course would teach them the required skills. Government agencies that fund academic scientific research pay for most of the computer systems supporting scientific research as well as the development of most personal scientific codes. These agencies should require graduate schools to offer a course in optimization and parallel programming as a requirement for funding. About the Author John Feo received his Ph.D. in Computer Science from The University of Texas at Austin in 1986. After graduate school, Dr. Feo worked at Lawrence Livermore National Laboratory where he was the Group Leader of the Computer Research Group and principal investigator of the Sisal Language Project. In 1997, Dr. Feo joined Tera Computer Company where he was project manager for the MTA, and oversaw the programming and evaluation of the MTA at the San Diego Supercomputer Center. In 2000, Dr. Feo joined Sun Microsystems as an HPC application specialist. He works with university research groups to optimize and parallelize scientific codes. Dr. Feo has published over two dozen research articles in the areas of parallel parallel programming, parallel programming languages, and application performance.

    Read the article

  • Best place to request Ubuntu for a minor improvement (In Unity dash search)

    - by mac
    Which is the best place to request Ubuntu for a minor improvement? My request feature is this : In Ubuntu dash when I search for "Upd" it gives me update manager and some other files. Now when I click enter by default the first entry will be selected. Can we make this a slightly better experience by highlighting the first item in search results which will be selected by default if we press enter - Just like in Gnome shell Search for upd in unity dash Search for upd in gnome-shell If you notice, update manager is highlighted by default in gnome shell and appears more intuitive. Can we implement the same in Unity ? Sorry for posting this in askubuntu. I just wanted to know which is the best place to discuss this. Thanks

    Read the article

  • Process Improvement and the Data Professional

    - by BuckWoody
    Don’t be afraid of that title – I’m not talking about Six Sigma or anything super-formal here. In many organizations, there are more folks in other IT roles than in the Data Professional area. In other words, there are more developers, system administrators and so on than there are the “DBA” role. That means we often have more to do than the time we need to do it. And, oddly enough, the first thing that is sacrificed is process improvement – the little things we need to do to make the day go faster in the first place. Then we get even more behind, the work piles up and…well, you know all about that. Earlier I challenged you to find 10-30 minutes a day to study. Some folks wrote back and asked “where do I start”? Well, why not be super-efficient and combine that time with learning how to make yourself more efficient? Try out a new scripting language, learn a new tool that automates things or find out ways others have automated their systems. In general, find out what you’re doing and how, and then see if that can be improved. It’s kind of like doing a performance tuning gig on yourself! If you’re pressed for time, look for bite-sized articles (like the ones I’ve done here for PowerShell and SQL Server) that you can follow in a “serial” fashion. In a short time you’ll have a new set of knowledge you can use to make your day faster. Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • ASP.Net 4.5 Garbage Collection Improvement

    - by Aligned
    Originally posted on: http://geekswithblogs.net/Aligned/archive/2013/06/24/asp.net-4.5-garbage-collection-improvement.aspxI just read Five Great .NET Framework 4.5 Features on CodeProject by Shivprasad koirala. Feature 5 in his article mentions the GC background cleanup and has a good explanation of the work the GC has to do for ASP.Net on the server. “Garbage collector is one real heavy task in a .NET application. And it becomes heavier when it is an ASP.NET application. ASP.NET applications run on the server and a lot of clients send requests to the server thus creating loads of objects, making the GC really work hard for cleaning up unwanted objects.” “To overcome the above problem, server GC was introduced. In server GC there is one more thread created which runs in the background. This thread works in the background and keeps cleaning…objects thus minimizing the load on the main GC thread. Due to double GC threads running, the main application threads are less suspended, thus increasing application throughput. To enable server GC, we need to use the gcServer XML tag and enable it to true.” <configuration> <runtime> <gcServer enabled="true"/> </runtime> </configuration> This is not done by default. The MSDN information page says “There are only two garbage collection options, workstation or server. For single-processor computers, the default workstation garbage collection should be the fastest option. Either workstation or server can be used for two-processor computers. Server garbage collection should be the fastest option for more than two processors. Use the GCSettingsIsServerGC property to determine if server garbage collection is enabled.” “In the .NET Framework 4 and earlier versions, concurrent garbage collection is not available when server garbage collection is enabled. Starting with the .NET Framework 4.5, server garbage collection is concurrent. To use non-concurrent server garbage collection, set the <gcServer> element to true and the <gcConcurrent> element to false. “ So if you’re using ASP.Net 4.5 and have a multi-core server, you should try turning on the Server Garbage Collection and do some profiling to see if it improves the performance of your site.

    Read the article

  • Performance Improvement: Session State

    Performance is critical to today's successful applications and web sites. If you design with an awareness of the session state management challenges you can always change your strategies to match your performance needs.

    Read the article

  • Performance Improvement: Session State

    Performance is critical to today's successful applications and web sites. If you design with an awareness of the session state management challenges you can always change your strategies to match your performance needs.

    Read the article

  • SQL SERVER – Online Index Rebuilding Index Improvement in SQL Server 2012

    - by pinaldave
    Have you ever faced situation when you see something working and you feel it should not be working? Well, I had similar moments few days ago. I know that SQL Server 2008 supports online indexing. However, I also know that I cannot rebuild index ONLINE if I have used VARCHAR(MAX), NVARCHAR(MAX) or few other data types. While I held my belief very strongly I came across situation, where I had to go online and do little bit reading from Book Online. Here is the similar example. First of all – run following code in SQL Server 2008 or SQL Server 2008 R2. USE TempDB GO CREATE TABLE TestTable (ID INT, FirstCol NVARCHAR(10), SecondCol NVARCHAR(MAX)) GO CREATE CLUSTERED INDEX [IX_TestTable] ON TestTable (ID) GO CREATE NONCLUSTERED INDEX [IX_TestTable_Cols] ON TestTable (FirstCol) INCLUDE (SecondCol) GO USE [tempdb] GO ALTER INDEX [IX_TestTable_Cols] ON [dbo].[TestTable] REBUILD WITH (ONLINE = ON) GO DROP TABLE TestTable GO Now run the same code in SQL Server 2012 version. Observe the difference between both of the execution. You will be get following resultset. In SQL Server 2008/R2 it will throw following error: Msg 2725, Level 16, State 2, Line 1 An online operation cannot be performed for index ‘IX_TestTable_Cols’ because the index contains column ‘SecondCol’ of data type text, ntext, image, varchar(max), nvarchar(max), varbinary(max), xml, or large CLR type. For a non-clustered index, the column could be an include column of the index. For a clustered index, the column could be any column of the table. If DROP_EXISTING is used, the column could be part of a new or old index. The operation must be performed offline. In SQL Server 2012 it will run successfully and will not throw any error. Command(s) completed successfully. I always thought it will throw an error if there is VARCHAR(MAX) or NVARCHAR(MAX) used in table schema definition. When I saw this result it was clear to me that it will be for sure not bug enhancement in SQL Server 2012. For matter for the fact, I always wanted this feature to be added in SQL Server Engine as this will enable ONLINE Index Rebuilding for mission critical tables which needs to be always online. I quickly searched online and landed on Jacob Sebastian’s blog where he has blogged about it as well. Well, is there any other new feature in SQL Server 2012 which gave you good surprise? Reference: Pinal Dave (http://blog.sqlauthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Index, SQL Query, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • Wine Security - Improvement by second user account?

    - by F. K.
    Team, I'm considering installing wine - but still hesitant for security reasons. As far as I found out, malicious code could reach ~/.wine and all my personal data with my user-priviledges - but not farther than that. So - would it be any safer to create a second user account on my machine and install wine there? That way, the second user would only have reading rights to my files. Is there a way to install wine totally confined to that user - so that I can't execute .exe files from my original account? Thanks in advance! PS - I'm running Ubuntu 11.10 64bit if that matters.

    Read the article

  • SQL SERVER – Video – Performance Improvement in Columnstore Index

    - by pinaldave
    I earlier wrote an article about SQL SERVER – Fundamentals of Columnstore Index and it got very well accepted in community. However, one of the suggestion I keep on receiving for that article is that many of the reader wanted to see columnstore index in the action but they were not able to do that. Some of the readers did not install SQL Server 2012 or some did not have good machine to recreate the big table involved in the demo. For the same reason, I have created small video for that. I have written two more article on columstore index. Please read them as followup to the video: SQL SERVER – How to Ignore Columnstore Index Usage in Query SQL SERVER – Updating Data in A Columnstore Index Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: Pinal Dave, PostADay, SQL, SQL Authority, SQL Index, SQL Performance, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, T SQL, Technology, Video

    Read the article

  • ArvinMeritor Sees Business Improvement: Uses Oracle Demand Management, Supply Chain Planning and Tra

    - by [email protected]
    As manufacturers begin repositioning for the economic recovery, they are reevaluating their supply chain networks, extending lean into their supply chains and making logistics visibility a priority. ArvinMeritor leveraged Oracle's Demantra, ASCP and Transportation Management applications to: Optimize operations execution by building consensus-driven demand, sales and operations plans Slash transportation costs by rationalizing shippers, optimizing routes and improving delivery performance Demantra for demand management, forecasting, sales and operations planning and global trade management Advanced Supply Chain Planning for material and capacity planning across global distribution and manufacturing facilities based on consensus forecasts, sales orders, production status, purchase orders, and inventory policy recommendations Transportation Management for transportation planning, execution, freight payment, and business process automation on a single application across all modes of transportation, from full truckload to complex multileg air, ocean, and rail shipments Oracle hosted an 'open-house/showcase" on March 30th, 2010 atArvinMeritor Global Headquarters 2135 West Maple RoadTroy, MI 48084 

    Read the article

  • Alexa Traffic Rankings Continuously Inaccurate - Room For Improvement

    There is growing news on the internet that the Alexa Traffic Rankings are not accurate and cannot be used to boast traffic rankings for the average site owner. The main flaw that I have found with Alexa, is their main method of collecting user traffic data, which is mainly through an Alexa Toolbar that collects data from users' browsers that have the toolbars installed. This is my experiment and the results.

    Read the article

  • Browser privacy improvement implications for websites

    - by phq
    On https://panopticlick.eff.org/ EFF let you test the number of uniquely identifying bits that the browser gives a website. Among these are HTTP header fields such as User-Agent, Accept, Accept-Language and later perhaps ETAG and If-Modified-Since. Also there is a lot of Information that javascript can get from the browser such as time-zone, screen resolution, complete list of fonts and plugins available. My first impression is, is all this information really usable/used on a majority of all websites? For example, how many sites does really send different content-types depending on the http accept header, or what fonts are available(I thought css had taken care of this)? Let's say of these headers/js functionality on day would be gone. Which ones would; never be noticed they were gone? impact user experience? impact server performance? immediately reimplemented because the Internet cannot work without it? Extra credit for differentiating between what can be done, what should be done and what is done in most situations.

    Read the article

1 2 3 4 5 6 7 8 9 10 11 12  | Next Page >