The series of mini-guides is taken from the original "ALF for Cell BE Programmer's Guide and API Reference." The rest are from IBM developerWorks.
M I N I - G U I D E S
Meet ALF (SDK 3.0)
Introduction includes descriptions of workload division, tasks, runtime framework, external components, and indication and situations best for ALF.
10 major concepts (SDK 3.0)
Explains the  computational kernel,  tasks (finalize, dependency, scheduling, instance, fixed mapping, context, events),  the task descriptor,  work blocks types and scheduling,  data partitioning (data transfer list, host data partitioning, accelerator data partitioning),  accelerator buffer management and the five types of buffers,  using work blocks and order of function calls per task instance on the accelerator,  modifying the work block parameter and context buffer when using multi-use work blocks,  datasets, and  error handling.
Basic ALF application structure (SDK 3.0)
Diagram of basic app structure, description of host/accelerator processes, samples, and source code.
Double buffering (SDK 3.0)
Description of double buffering and buffer schemes (single-buffer, three-buffer, four-buffer, and overlapped I/O buffer schemes).
Handling constraints (SDK 3.0)
SPE accelerator memory constraints and data transfer list limitations.
Optimizing applications (SDK 3.0)
Three ways to optimize applications with ALF --  using accelerator data partitioning,  using multi-use work blocks, and  what to consider for data layout design.
Accelerator buffer management (SDK 3.0)
Backgrounder on buffer management and description of/usage guide for buffer types (task context, work block parameter and context, work block input data, work block output data, and work block overlapped input and output data buffers).
When to use the overlapped I/O buffer (SDK 3.0)
To maximize the memory usage on accelerators.
M I N I - G U I D E S > H Y B R I D X 8 6
Hybrid x86: Using ALF on Cell/B.E.-x86 hybrid systems (SDK 3.0)
Description, installation and configuration, linking to libraries, constraints and limitations, building, running, and optimizing an application.
M I N I - G U I D E S > G L O S S A R I E S
Glossary: ALF error codes (SDK 3.0)
Error code listing and descriptions.
Glossary: ALF trace events (SDK 3.0)
API, performance, and SPU hooks descriptions.
Glossary: ALF attributes (SDK 3.0)
Attributes listing and descriptions.
A R T I C L E S / T U T O R I A L S
Fun with ALF: Adding large matrices together
In part one of this code examples series, use ALF to add two large matrices together on the host and accelerator partitions.
Fun with ALF: Converting I/O data
Use ALF's task context buffer as a large lookup table to convert 16-bit input to 8-bit output data.
Fun with ALF: Finding minimum and maximum values
Use ALF's task context to find minimum and maximum values in a large data set.
Fun with ALF: Determining the dot product of large vectors
Use the ALF bundled work block distribution with the task context to overcome memory limits.
Fun with ALF: Using overlapped I/O buffers to add matrices
Use the ALF overlapped I/O buffers to do matrix addition.
Fun with ALF: Using task dependency
Use ALF task dependency in a two-stage pipeline application.
Cell/B.E. SDK 3.0, Part 6: Use the ALF wizard
In this tutorial, see a great example usage scenario of the ALF IDE wizard.
C O D E S A M P L E S
Cell/B.E. SDK: Code sample directory
Where to find code samples for ALF (BlackScholes_ALF, FFT16M_ALF, hello_world, inout_buffer, inverse_matrix_ovl, matrix_add, matrix_transpose, PI, pipe_line, and task_context) and ALF Hybrid (same list as above).