alatron978

€145 for a 10400F which beats the 3600. A 3600 system is not the way to go at all.

Yeah, if all you want is 60FPS a 10400F is all you need. It will provide you with the same great experience that any higher end chip would give you There is no reason to be concerned about paring CPU's and GPU's, it just isn't a good idea. You should be paring your CPU and GPU to use load, which for 60fps in triple A games a 10400F is all you need. CPU + GPU parings is a concept that really should die, it is extremely useless with the varied gaming loads that people put their systems under.

What is your FPS target and what games do you play? You could likely get away with a 10400F or a 10100F and see no noticeable performance impact.

Knowing performance per a specific game isn't useful unless you only play one game. Back to the civ 6 example, zen 3 is 60% faster then CML, but, CML is faster on average in gaming then zen 3.

It certainly matters when you have games having 4 times the impact as other games.

Neither, you play at 4K, you don't need that much CPU power.

My beats anything AMD offers is based off of averages, because what is the point of looking at a single bench. Sure I could pull up a civ 6 bench where zen 3 is 60% faster then CML, but it doesn't average out to that. I'm not biased towards intel, the 10400F is simply cheaper and faster. And what is wrong with putting the 10900K on to a b460 board for gaming? Under gaming workloads the chip doesn't pull the 300W+ like it does in something such as prime95. The 5950X can pull a crap load of current as well can't it.

Intel benefits more from dual rank mem then amd systems, since intels mem arch is quite significantly less efficient then amd's.

That is why I made that comment. With 3600 mem on the 5800x you'd most likely beat the 10900K if it is using 2933, assuming both are running mem that isn't tuned.

I'm just stating facts, I don't care if AMD or Intel is on top, but I do care about knowing which is currently on top. I'd prefer AMD to be on top though since Intel is still using skylake which is just ancient at this point.

If you have the same mem settings on the ryzen 5000 system it would.

Yeah that is fine, 1.59V is nothing for a 9900K.

CPU's use like 1.2V~1.5V, or as in this post, 1.59V~1.32V RMS.

The 10900K does beat the 5000 series AMD chips. Note that the TPU messed up the mem oc on the 10900K, so ignore the 3800 cl16 mem results.

If they don't buy a cooler with the 10400F it would be cheaper. And the 10400F doesn't need a cooler, the stock one is not good but adequate.

Bench I showed had the 10400F 10% faster in a like for like scenario. They'd be paying less if they went for a cheaper b460 and didn't through a cooler in with the 3600. I see personally no reason to spend more money on something that is slower. Regarding cpu and mem oc yes that could easily put the 3600 10~20% above the 10400F, though I don't expect that most people would engage in these kinds of overclocking. The 3600 having a better upgrade path is also dependent on your workload, for gaming the 10400F for sure has a better upgrade path since the 10900K beats anything AMD offers and has rocketlake on the horizon.

Of course it is, what matters is that the 10400F is faster then the 3600 in a gaming workload though.

Yeah it is the same screenshot because it didn't let me delete the 2nd image. That is an average from the source which is linked above. You can also average 2 results, you don't need 3 to have an average.

That's the writers of the manual being idiots then.

Yeah I checked the bench on average it is slower.

That benchmark is pretty useless considering hwub's averaging methodology.

10400F has a stock cooler that is perfectly fine. Just cut the cooler out of the 10400F system and it'll be faster and cheaper.

No it isn't.

Boost clocks are different to per core OC. Per core OC has each core at a static frequency. Boost ratios allow higher clocks when only a certain number of cores are loaded.

The PBO scalar is a reliability scalar and thus a hazard, I wouldn't recommend running it 10X without any form of negative voltage offset. Your memory settings are also very strange.