Performance Studies using PSTSWM

Compaq AlphaServer SC Serial Performance

(comparing performance when running multiple instances)

(fixed memory size per node)

T5 MFlop rates

Per Processor Rate	Per Node Rate

T10 MFlop rates

Per Processor Rate	Per Node Rate

T21 MFlop rates

Per Processor Rate	Per Node Rate

T42 MFlop rates

Per Processor Rate	Per Node Rate

T85 MFlop rates

Per Processor Rate	Per Node Rate

T170 MFlop rates

Per Processor Rate	Per Node Rate

DISCUSSION

• In this analysis, the experiment computational rates are graphed as a function of the total number of vertical levels on the node. Thus the experiments using more processors have a smaller problem granularity per processor, resulting in improved cache utilization. The total memory required per node is approximately constant for a given number of "node" levels. Therefore the total volume moved between memory and processors for a given number of "node" levels should also be somewhat less for the experiments using more processors, due to the improved cache utilization. However, if the total problem size is large enough to saturate the processor to memory bandwidth, then this bandwidth constrains the cumulative computational rate for the node, and the individual per processor computational rates will suffer.
• This experiment indicates an upper bound on the performance one might see when parallelizing a problem across the processors of a node, in as far as it does not include message passing or thread creation or synchronization costs.
• Note that two different metrics are displayed - per processor computational rate and per node computational rate. The per node rate is the best measure of the performance of the architecture, but the per processor rate is a convenient way to observe performance degradation.
• Summary: When a single level of the problem fits in cache (T42), there is no performance degradation in the per processor rate from using more than one processor per node. When memory must be accessed throughout the computation even for a single level (T85, T170), then there is contention, and both the per processor and per node rates decrease. However, it never makes sense to use less than 4 processors per node.

---

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:05:49 EDT.
82445 accesses since 1/2/96.