Search Results

Search found 233 results on 10 pages for 'parallelism'.

Page 6/10 | < Previous Page | 2 3 4 5 6 7 8 9 10  | Next Page >

  • Load Balance and Parallel Performance

    Load balancing an application workload among threads is critical to performance. However, achieving perfect load balance is non-trivial, and it depends on the parallelism within the application, workload, the number of threads, load balancing policy, and the threading implementation.

    Read the article

  • SQL Server Connections Fall 2011 - Demos

    - by Adam Machanic
    Today is the last day of the annual SQL Server Connections show in Vegas, and I've just completed my third and final talk. (Now off to find a frosty beverage or two.) This year I did three sessions: SQL302: Parallelism and Performance: Are You Getting Full Return on Your CPU Investment? Over the past five years, multi-core processors have made the jump from semi-obscure to commonplace in the data center. While servers with 16, 32, or even 64 cores were once an out-of-reach choice for all except the...(read more)

    Read the article

  • How employable am I as a programmer?

    - by dsimcha
    I'm currently a Ph.D. student in Biomedical Engineering with a concentration in computational biology and am starting to think about what I want to do after graduate school. I feel like I've accumulated a lot of programming skills while in grad school, but taken a very non-traditional path to learning all this stuff. I'm wondering whether I would have an easy time getting hired as a programmer and could fall back on that if I can't find a good job directly in my field, and if so whether I would qualify for a more prestigious position than "code monkey". Things I Have Going For Me Approximately 4 years of experience programming as part of my research. I believe I have a solid enough grasp of the fundamentals that I could pick up new languages and technologies pretty fast, and could demonstrate this in an interview. Good math and statistics skills. An extensive portfolio of open source work (and the knowledge that working on these projects implies): I wrote a statistics library in D, mostly from scratch. I wrote a parallelism library (parallel map, reduce, foreach, task parallelism, pipelining, etc.) that is currently in review for adoption by the D standard library. I wrote a 2D plotting library for D against the GTK Cairo backend. I currently use it for most of the figures I make for my research. I've contributed several major performance optimizations to the D garbage collector. (Most of these were low-hanging fruit, but it still shows my knowledge of low-level issues like memory management, pointers and bit twiddling.) I've contributed lots of miscellaneous bug fixes to the D standard library and could show the change logs to prove it. (This demonstrates my ability read other people's code.) Things I Have Going Against Me Most of my programming experience is in D and Python. I have very little to virtually no experience in the more established, "enterprise-y" languages like Java, C# and C++, though I have learned a decent amount about these languages from small, one-off projects and discussions about language design in the D community. In general I have absolutely no knowledge of "enterprise-y" technlogies. I've never used a framework before, possibly because most reusable code for scientific work and for D tends to call itself a "library" instead. I have virtually no formal computer science/software engineering training. Almost all of my knowledge comes from talking to programming geek friends, reading blogs, forums, StackOverflow, etc. I have zero professional experience with the official title of "developer", "software engineer", or something similar.

    Read the article

  • Google I/O 2010 - Batch data processing with App Engine

    Google I/O 2010 - Batch data processing with App Engine Google I/O 2010 - Batch data processing with App Engine App Engine 201 Mike Aizatsky In this session, attendees will learn how to write map() functions, how to do simple reduce() operations, how to run these over large datasets, and how App Engine is used to accomplish such parallelism. For all I/O 2010 sessions, please go to code.google.com From: GoogleDevelopers Views: 6 0 ratings Time: 38:45 More in Science & Technology

    Read the article

  • Shows how to use the new Tasks namespace to download multiple documents in parallel.

    In C# 4.0, Task parallelism is the lowest-level approach to parallelization with PFX. The classes for working at this level are defined in the System.Threading.Tasks namespace.  read moreBy Peter BrombergDid you know that DotNetSlackers also publishes .net articles written by top known .net Authors? We already have over 80 articles in several categories including Silverlight. Take a look: here.

    Read the article

  • Ideas for pet concurrent applications [closed]

    - by DJoyce
    As part of research I am doing into functional concurrent programming, I am looking to develop a prototype business application (or otherwise), that requires and supports parallelism. So I will first develop this application in java 7, followed by scala and then clojure while demonstrating the concurrent support in each language. However, I am a little short on suitable ideas and therefore i'm hoping I can get some good, interesting ideas on this thread. Thanks!

    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

  • Inappropriate Updates?

    - by Tony Davis
    A recent Simple-talk article by Kathi Kellenberger dissected the fastest SQL solution, submitted by Peter Larsson as part of Phil Factor's SQL Speed Phreak challenge, to the classic "running total" problem. In its analysis of the code, the article re-ignited a heated debate regarding the techniques that should, and should not, be deemed acceptable in your search for fast SQL code. Peter's code for running total calculation uses a variation of a somewhat contentious technique, sometimes referred to as a "quirky update": SET @Subscribers = Subscribers = @Subscribers + PeopleJoined - PeopleLeft This form of the UPDATE statement, @variable = column = expression, is documented and it allows you to set a variable to the value returned by the expression. Microsoft does not guarantee the order in which rows are updated in this technique because, in relational theory, a table doesn’t have a natural order to its rows and the UPDATE statement has no means of specifying the order. Traditionally, in cases where a specific order is requires, such as for running aggregate calculations, programmers who used the technique have relied on the fact that the UPDATE statement, without the WHERE clause, is executed in the order imposed by the clustered index, or in heap order, if there isn’t one. Peter wasn’t satisfied with this, and so used the ingenious device of assuring the order of the UPDATE by the use of an "ordered CTE", based on an underlying temporary staging table (a heap). However, in either case, the ordering is still not guaranteed and, in addition, would be broken under conditions of parallelism, or partitioning. Many argue, with validity, that this reliance on a given order where none can ever be guaranteed is an abuse of basic relational principles, and so is a bad practice; perhaps even irresponsible. More importantly, Microsoft doesn't wish to support the technique and offers no guarantee that it will always work. If you put it into production and it breaks in a later version, you can't file a bug. As such, many believe that the technique should never be tolerated in a production system, under any circumstances. Is this attitude justified? After all, both forms of the technique, using a clustered index to guarantee the order or using an ordered CTE, have been tested rigorously and are proven to be robust; although not guaranteed by Microsoft, the ordering is reliable, provided none of the conditions that are known to break it are violated. In Peter's particular case, the technique is being applied to a temporary table, where the developer has full control of the data ordering, and indexing, and knows that the table will never be subject to parallelism or partitioning. It might be argued that, in such circumstances, the technique is not really "quirky" at all and to ban it from your systems would server no real purpose other than to deprive yourself of a reliable technique that has uses that extend well beyond the running total calculations. Of course, it is doubly important that such a technique, including its unsupported status and the assumptions that underpin its success, is fully and clearly documented, preferably even when posting it online in a competition or forum post. Ultimately, however, this technique has been available to programmers throughout the time Sybase and SQL Server has existed, and so cannot be lightly cast aside, even if one sympathises with Microsoft for the awkwardness of maintaining an archaic way of doing updates. After all, a Table hint could easily be devised that, if specified in the WITH (<Table_Hint_Limited>) clause, could be used to request the database engine to do the update in the conventional order. Then perhaps everyone would be satisfied. Cheers, Tony.

    Read the article

  • DirectX: Game loop order, draw first and then handle input?

    - by Ricket
    I was just reading through the DirectX documentation and encountered something interesting in the page for IDirect3DDevice9::BeginScene : To enable maximal parallelism between the CPU and the graphics accelerator, it is advantageous to call IDirect3DDevice9::EndScene as far ahead of calling present as possible. I've been accustomed to writing my game loop to handle input and such, then draw. Do I have it backwards? Maybe the game loop should be more like this: (semi-pseudocode, obviously) while(running) { d3ddev->Clear(...); d3ddev->BeginScene(); // draw things d3ddev->EndScene(); // handle input // do any other processing // play sounds, etc. d3ddev->Present(NULL, NULL, NULL, NULL); } According to that sentence of the documentation, this loop would "enable maximal parallelism". Is this commonly done? Are there any downsides to ordering the game loop like this? I see no real problem with it after the first iteration... And I know the best way to know the actual speed increase of something like this is to actually benchmark it, but has anyone else already tried this and can you attest to any actual speed increase?

    Read the article

  • Data Warehouse Best Practices

    - by jean-pierre.dijcks
    In our quest to share our endless wisdom (ahem…) one of the things we figured might be handy is recording some of the best practices for data warehousing. And so we did. And, we did some more… We now have recreated our websites on Oracle Technology Network and have a separate page for best practices, parallelism and other cool topics related to data warehousing. But the main topic of this post is the set of recorded best practices. Here is what is available (and it is a series that ties together but can be read independently), applicable for almost any database version: Partitioning 3NF schema design for a data warehouse Star schema design Data Loading Parallel Execution Optimizer and Stats management The best practices page has a lot of other useful information so have a look here.

    Read the article

  • Auto DOP and Concurrency

    - by jean-pierre.dijcks
    After spending some time in the cloud, I figured it is time to come down to earth and start discussing some of the new Auto DOP features some more. As Database Machines (the v2 machine runs Oracle Database 11.2) are effectively selling like hotcakes, it makes some sense to talk about the new parallel features in more detail. For basic understanding make sure you have read the initial post. The focus there is on Auto DOP and queuing, which is to some extend the focus here. But now I want to discuss the concurrency a little and explain some of the relevant parameters and their impact, specifically in a situation with concurrency on the system. The goal of Auto DOP The idea behind calculating the Automatic Degree of Parallelism is to find the highest possible DOP (ideal DOP) that still scales. In other words, if we were to increase the DOP even more  above a certain DOP we would see a tailing off of the performance curve and the resource cost / performance would become less optimal. Therefore the ideal DOP is the best resource/performance point for that statement. The goal of Queuing On a normal production system we should see statements running concurrently. On a Database Machine we typically see high concurrency rates, so we need to find a way to deal with both high DOP’s and high concurrency. Queuing is intended to make sure we Don’t throttle down a DOP because other statements are running on the system Stay within the physical limits of a system’s processing power Instead of making statements go at a lower DOP we queue them to make sure they will get all the resources they want to run efficiently without trashing the system. The theory – and hopefully – practice is that by giving a statement the optimal DOP the sum of all statements runs faster with queuing than without queuing. Increasing the Number of Potential Parallel Statements To determine how many statements we will consider running in parallel a single parameter should be looked at. That parameter is called PARALLEL_MIN_TIME_THRESHOLD. The default value is set to 10 seconds. So far there is nothing new here…, but do realize that anything serial (e.g. that stays under the threshold) goes straight into processing as is not considered in the rest of this post. Now, if you have a system where you have two groups of queries, serial short running and potentially parallel long running ones, you may want to worry only about the long running ones with this parallel statement threshold. As an example, lets assume the short running stuff runs on average between 1 and 15 seconds in serial (and the business is quite happy with that). The long running stuff is in the realm of 1 – 5 minutes. It might be a good choice to set the threshold to somewhere north of 30 seconds. That way the short running queries all run serial as they do today (if it ain’t broken, don’t fix it) and allows the long running ones to be evaluated for (higher degrees of) parallelism. This makes sense because the longer running ones are (at least in theory) more interesting to unleash a parallel processing model on and the benefits of running these in parallel are much more significant (again, that is mostly the case). Setting a Maximum DOP for a Statement Now that you know how to control how many of your statements are considered to run in parallel, lets talk about the specific degree of any given statement that will be evaluated. As the initial post describes this is controlled by PARALLEL_DEGREE_LIMIT. This parameter controls the degree on the entire cluster and by default it is CPU (meaning it equals Default DOP). For the sake of an example, let’s say our Default DOP is 32. Looking at our 5 minute queries from the previous paragraph, the limit to 32 means that none of the statements that are evaluated for Auto DOP ever runs at more than DOP of 32. Concurrently Running a High DOP A basic assumption about running high DOP statements at high concurrency is that you at some point in time (and this is true on any parallel processing platform!) will run into a resource limitation. And yes, you can then buy more hardware (e.g. expand the Database Machine in Oracle’s case), but that is not the point of this post… The goal is to find a balance between the highest possible DOP for each statement and the number of statements running concurrently, but with an emphasis on running each statement at that highest efficiency DOP. The PARALLEL_SERVER_TARGET parameter is the all important concurrency slider here. Setting this parameter to a higher number means more statements get to run at their maximum parallel degree before queuing kicks in.  PARALLEL_SERVER_TARGET is set per instance (so needs to be set to the same value on all 8 nodes in a full rack Database Machine). Just as a side note, this parameter is set in processes, not in DOP, which equates to 4* Default DOP (2 processes for a DOP, default value is 2 * Default DOP, hence a default of 4 * Default DOP). Let’s say we have PARALLEL_SERVER_TARGET set to 128. With our limit set to 32 (the default) we are able to run 4 statements concurrently at the highest DOP possible on this system before we start queuing. If these 4 statements are running, any next statement will be queued. To run a system at high concurrency the PARALLEL_SERVER_TARGET should be raised from its default to be much closer (start with 60% or so) to PARALLEL_MAX_SERVERS. By using both PARALLEL_SERVER_TARGET and PARALLEL_DEGREE_LIMIT you can control easily how many statements run concurrently at good DOPs without excessive queuing. Because each workload is a little different, it makes sense to plan ahead and look at these parameters and set these based on your requirements.

    Read the article

  • SQL SERVER – SQLServer Quiz 2011 – Do you know your execution plan – Two questions – One Answer

    - by pinaldave
    My friend Jacob Sebastian has SQL Server Quiz 2011 launched. This time when he asked me to come up with quiz question – I wanted to come up with something which is new and make participant to think about it. After carefully thinking I come with question which I really like to solve myself. Here is the details: 1) Using Single table only Once in Single SELECT statement generate execution plan which have JOIN operator. Explain the reason for the same. 2) Using Single table only Once in Single SELECT statement generate execution plan which have parallelism operator. Explain the reason for the same. Bonus: Create a single query which satisfy both of the above statement. To answer this question and win exciting gifts please visit the SQL Server Quiz website. Reference: Pinal Dave (http://blog.SQLAuthority.com)   Filed under: Pinal Dave, PostADay, Readers Contribution, Readers Question, SQL, SQL Authority, SQL Puzzle, SQL Query, SQL Scripts, SQL Server, SQL Tips and Tricks, SQLServer, T SQL, Technology

    Read the article

  • Parallel Computing Platform Developer Lab

    This is an exciting announcement that I must share: "Microsoft Developer & Platform Evangelism, in collaboration with the Microsoft Parallel Computing Platform product team, is hosting a developer lab at the Platform Adoption Center on April 12-15, 2010.  This event is for Microsoft Partners and Customers seeking to incorporate either .NET Framework 4 or Visual C++ 2010 parallelism features into their new or existing applications, and to gain expertise with new Visual Studio 2010 tools...Did you know that DotNetSlackers also publishes .net articles written by top known .net Authors? We already have over 80 articles in several categories including Silverlight. Take a look: here.

    Read the article

  • A better way to do concurrent programming

    - by Alex.Davies
    Programming to take advantage of multicore processors is hard. If you let multiple threads access the same memory, bad things happen. To avoid this, you use the lock keyword, but if you use that in the wrong way, your code deadlocks. It's all a nightmare. Luckily, there's a better way - Actors. They're really easy to think about. They're really safe (if you follow a couple of simple rules). And high-performance, type-safe actors are now available for .NET by using this open-source library: http://code.google.com/p/n-act/ Have a look at the site for details. I'll blog with more reasons to use actors and tips and tricks to get the best parallelism from them soon.

    Read the article

  • SQL Saturday #220 - Atlanta - Pre-Con Scholarship Winners!

    - by Most Valuable Yak (Rob Volk)
    A few weeks ago, AtlantaMDF offered scholarships for each of our upcoming Pre-conference sessions at SQL Saturday #220. We would like to congratulate the winners! David Thomas SQL Server Security http://sqlsecurity.eventbrite.com/ Vince Bible Surfing the Multicore Wave: Processors, Parallelism, and Performance http://surfmulticore.eventbrite.com/ Mostafa Maged Languages of BI http://languagesofbi.eventbrite.com/ Daphne Adams Practical Self-Service BI with PowerPivot for Excel http://selfservicebi.eventbrite.com/ Tim Lawrence The DBA Skills Upgrade Toolkit http://dbatoolkit.eventbrite.com/ Thanks to everyone who applied! And once again we must thank Idera's generous sponsorship, and the time and effort made by Bobby Dimmick (w|t) and Brian Kelley (w|t) of Midlands PASS for judging all the applicants. Don't forget, there's still time to attend the Pre-Cons on May 17, 2013! Click on the EventBrite links for more details and to register!

    Read the article

  • Slides for Parallel Debugging windows

    Recently I gave a talk at our Microsoft Shanghai offices on Parallel Programming so I had to update my existing Beta1 deck to Beta2 content. Specifically for Parallel Tasks and Parallel Stacks, I used 5 slides to accompany the demo.In case you are giving talks on parallelism within Visual Studio 2010, please feel free to download and use the updated parallel debugger slides (pptx).TIP: The slides have animations so be sure to F5 the deck for the full benefit and they also have text in the Comments section so be sure to see them at design time too. Comments about this post welcome at the original blog.

    Read the article

  • What do you look for in a scripting language?

    - by Jon Purdy
    I'm writing a little embedded language for another project. While game development was not its original intent, it's starting to look like a good fit, and I figure I'll develop it in that vein at some point. Without revealing any details (to avoid bias), I'm curious to know: What features do you love in a scripting language for game development? If you've used Lua, Python, or another embedded language such as Tcl or Guile as your primary scripting language in a game project, what aspects did you find the most useful? Language features (lambdas, classes, parallelism) Implementation features (performance optimisations, JIT, hardware acceleration) Integration features (C, C++, or .NET bindings) Or something entirely different?

    Read the article

  • C++ AMP recording and slides

    - by Daniel Moth
    Yesterday we announced C++ Accelerated Massive Parallelism. Many of you want to know more about the API instead of just meta information. I will trickle more code over the coming months leading up to the date when we will share actual bits. Until you have bits in your hand, it is only your curiosity that is blocked, so I ask you to be patient with that and allow me to release this on our own schedule ;-) You can now watch my 45-minute session introducing C++ AMP on channel9. You will also want to download the slides (pdf), because they are not readable in the recording. Comments about this post welcome at the original blog.

    Read the article

  • Technical Computing

      Today, Microsoft announced our Technical Computing initiative.    Through the Technical Computing initiative, we will enable scientists, engineers and analysts to more easily model the world at much greater fidelity.  The Technical Computing initiative will address a wide range of users.  One of the most critical elements is to help developers create applications that can take advantage of parallelism on their desktop, in a cluster, and in public and private clouds. ...Did you know that DotNetSlackers also publishes .net articles written by top known .net Authors? We already have over 80 articles in several categories including Silverlight. Take a look: here.

    Read the article

  • Video: Analyzing Big Data using Oracle R Enterprise

    - by Sherry LaMonica
    Learn how Oracle R Enterprise is used to generate new insight and new value to business, answering not only what happened, but why it happened. View this YouTube Oracle Channel video overview describing how analyzing big data using Oracle R Enterprise is different from other analytics tools at Oracle. Oracle R Enterprise (ORE),  a component of the Oracle Advanced Analytics Option, couples the wealth of analytics packages in R with the performance, scalability, and security of Oracle Database. ORE executes base R functions transparently on database data without having to pull data from Oracle Database. As an embedded component of the database, Oracle R Enterprise can run your R script and open source packages via embedded R where the database manages the data served to the R engine and user-controlled data parallelism. The result is faster and more secure access to data. ORE also works with the full suite of in-database analytics, providing integrated results to the analyst.

    Read the article

  • Is there any working implementation of reverse mode automatic differentiation for Haskell?

    - by Ian Fiske
    The closest-related implementation in Haskell I have seen is the forward mode at http://hackage.haskell.org/packages/archive/fad/1.0/doc/html/Numeric-FAD.html. The closest related related research appears to be reverse mode for another functional language related to Scheme at http://www.bcl.hamilton.ie/~qobi/stalingrad/. I see reverse mode in Haskell as kind of a holy grail for a lot of tasks, with the hopes that it could use Haskell's nested data parallelism to gain a nice speedup in heavy numerical optimization.

    Read the article

  • Canonical pattern reference in Actors programming model.

    - by Bubba88
    Hello! Is there a source, which I could use to learn some of the most used and popular practices regarding Actor-/Agent-oriented programming. My primary concern is about parallelism and distribution limited to the mentioned scheme - Actors, message passing. Should I begin with Erlang documentation or maybe there is any kind of book that describes the most important building blocks when programming Actor-oriented? Thank you! (Most useful examples would be in Scala or F#)

    Read the article

  • Is it OK to try to use Plinq in all Linq queries?

    - by Tony_Henrich
    I read that PLinq will automatically use non parallel Linq if it finds PLinq to be more expensive. So I figured then why not use PLinq for everything (when possible) and let the runtime decide which one to use. The apps will be deployed to multicore servers and I am OK to develop a little more code to deal with parallelism. What are the pitfalls of using plinq as a default?

    Read the article

  • re: adding more threads to forkjoinpool

    - by paintcan
    Word up y'all I recently successfully experimented with Scala futures, got future { my shiznit } all over da place. I'm pleased as punch w/ the gains I'm seeing from the parallelism and whatnot, but I'm only seeing 4 worker threads. Wanna see some more. I've been looking all over for how I can crank up the number of threads to 11, but no luck. Help me out doods

    Read the article

< Previous Page | 2 3 4 5 6 7 8 9 10  | Next Page >