Chess engines are extremely computation-intensive applications and place high demands on processor and main memory. The consequences are increased power consumption and correspondingly higher temperatures of these system components. If your chess engine crashes or simply stops calculating after some time of computation, e.g. during infinite analysis, this is often due to the cooling of the system no longer working properly and the specifications being exceeded. Mainly computers with an AMD processor are affected by this problem.
The already difficult cooling of processors in computers is additionally exacerbated by an intensified competition among manufacturers for the highest clock rates. The processors often run at the upper limit. Under unfavourable conditions, such as elevated outside temperatures in summer, the components can then finally overheat.
Consequence: The computer either crashes or, in the worst case, can even be damaged.
Owners of a computer with an Athlon processor from AMD should be particularly cautious. These chips do not have a protective circuit that automatically shuts down the computer when overheating is imminent. In terms of heat development and stability, the Pentium processors from Intel offer advantages. In general, however, you should also ensure here that processor and graphics cards are operated at the intended clock frequency.
If Komodo & Co. regularly experience "hiccups" on your computer during play or even crash, this is often due to inadequate cooling of the system components mentioned. According to our experience from support, mainly computers with AMD processors and high-clocked notebooks with insufficient ventilation are affected.
As an affected user you can try the following approaches:
Ensure adequate cooling of the system. Optimal are computers equipped with water cooling. This allows the computer to deliver full peak performance even at very warm outside temperatures without the permissible temperature being exceeded. The conversion is not uncritical and should only be carried out by specialists.
For test purposes you can underclock the processor in the BIOS and, after consultation with the dealer, lower the core voltage. With this method, possible temperature problems caused by the processor can usually be narrowed down or localised.
Under certain circumstances, crashes can also originate from defective memory modules. Here, possible problems can be narrowed down by removing half of the memory and testing the different modules individually.