What every Programmer should know about memory

By Jochen Voss, on 2007-10-23

I started reading Ulrich Drepper's excellent series What every Programmer should know about memory. It is published by Linux Weekly News in weekly installments. Until now, the following parts appeared:

• Commodity Hardware Today
• CPU Caches
• Virtual Memory
• NUMA Support
• What Programmers Can Do — Cache Optimization
• What Programmers Can Do — Multi-Threaded Optimizations
• Memory Performance Tools
• Future Technologies
• Appendices and Bibliography

The author shows, for example, how the naive matrix-matrix multiplication

for (i = 0; i < N; ++i)
  for (j = 0; j < N; ++j)
    for (k = 0; k < N; ++k)
      res[i][j] += A[i][k] * B[k][j];

can be improved to run 10 times faster for big matrices!