MSI: You can't count on Comet Lake-S overclocking, most of the processors are working at the limit

All processors lend themselves to overclocking in different ways: some are able to conquer higher frequencies, others - lower ones. Ahead of the launch of the Comet Lake-S processors, MSI decided to formalize their overclocking potential by testing samples received from Intel.

MSI: You can't count on Comet Lake-S overclocking, most of the processors are working at the limit

As a motherboard manufacturer, MSI certainly received a lot of engineering and test samples of the new Comet Lake-S generation processors, so a large sample was involved in the overclocking experiment, and the resulting statistics should reflect a close to the real state of affairs. The Taiwanese manufacturer tested three groups of processors: six-core Core i5-10600K and 10600KF, eight-core Core i7-10700K and 10700KF and ten-core Core i9-10900K and 10900KF.

MSI: You can't count on Comet Lake-S overclocking, most of the processors are working at the limit

The results were quite unexpected. Of all the six-core Core i5-10600K (KF) processors tested, only 2% were able to run at higher frequencies than Intel claims (MSI Level A). More than half of the chips - 52%, were able to work only with the frequencies declared in the specifications (Level B). And 31% of the tested processors showed lower frequencies during overclocking compared to the passport ones (Level C). Apparently, there is another category of samples, but MSI does not say anything about it. The situation is similar with the eight-core Core i7-10700K (KF): 5% refers to the overclocked Level A group, 58% to the medium Level B and 32% to the number of Level C processors that work worse in overclocking than nominal.

Here it is worth explaining what the inability of processors to work with the declared frequencies means in MSI terminology. It seems that the company refers to the Level C category those chips that could not maintain stability under load when manually overclocked to the declared maximum turbo frequency for all cores. That is, with the removed restrictions on energy consumption.

But with the flagship ten-core, the situation is somewhat different. Here, 27% of the Core i9-10900K (KF) chips immediately succumbed to overclocking. The same number were not able to work with the declared characteristics, and another 35% exactly followed the nominal frequencies even during overclocking. This gives enthusiasts some hope for interesting records with these chips, which, however, will obviously have to be selected in a special way.

MSI: You can't count on Comet Lake-S overclocking, most of the processors are working at the limit

Along the way, MSI provides data on the power consumption and operating voltage of the above-mentioned new generation Core processors, depending on overclocking (X-axis indicates the multiplier value) in the Cinebench R20 multi-threaded test. As expected, Core i5 (blue) consumes the least - from about 130 to 210 watts. Core i9 (green) showed the highest appetite in most cases: from 190 to 275 watts. And a little behind the Core i7 (orange) flagships: the consumption of such processors is in the range from 175 to 280 watts. The operating voltage range is the widest for the flagship: from less than 1,0 to 1,35 V. The narrowest gap for the Core i5: from 1,1 to almost 1,3 V.

MSI: You can't count on Comet Lake-S overclocking, most of the processors are working at the limit

Finally, MSI provided data on how the power subsystem (VRM) of its motherboards heats up and, more importantly, how much the Core i9-10900K consumes when running at stock frequencies and overclocking. Under normal conditions, the processor requires about 205 watts of power, and the temperature of the VRM on the Z490 Gaming Edge WiFi board reaches 73,5 ° C. When overclocked across all cores to 5,1 GHz, power consumption reaches 255 W, and the temperature of the VRM is 86,5 ° C. By the way, in these experiments, a two-section LSS Corsair H115i was used to cool the processor.



Source: 3dnews.ru

Add a comment