10400f VS 3600

[alatron978]

1 minute ago, ragnarok0273 said:

B460 locks out RAM speed to 2933, remember?

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.

[boggy77]

Just now, alatron978 said:

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

why would you, though?

 

and the techpowerup benches you keep pushing are made with single rank ram. it was shown amd benefits greatly by having 4 ranks in dual channel, and all benchmarks done with this type of memory beats the equivalent intel systems by a lot.

[Princess Luna]

12 minutes ago, alatron978 said:

Bench I showed had the 10400F 10% faster in a like for like scenario.

How many % would this one be:

It's not good to deal with absolutes, all in all the statement that the processors are on pair of one another stands.

 

12 minutes ago, alatron978 said:

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.

11 minutes ago, alatron978 said:

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.

To be fair everyone who has ever had an Intel Stock Cooler under operation knows that thing is as loud as it gets, it is clearly inferior to AMD's and it can get to really bothering levels. Also again with this "something that is slower" there's no objectively faster or slower answer in here.

 

12 minutes ago, alatron978 said:

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.

Having the alternative means you can at some point bother yourself to learn and do it, this options is simply not present on Intel which be as much team blue as possible you cannot deny how much a bummer that is.

12 minutes ago, alatron978 said:

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.

Now this is the moment you kinda show your bias... the R5 5600X can be used on the Motherboard of choice by OP and that's a processor that does reach and sometimes even surpasses the i9 10900K already making your "beats anything AMD offers" claim false.

 

To add salt to injury, note you just advised OP to consider an i9 10900K, a 10 cores power house CPU, to be used on the "cheapest B460" you can buy. How's that a senseful upgrade path?

[alatron978]

Just now, boggy77 said:

why would you, though?

 

and the techpowerup benches you keep pushing are made with single rank ram. it was shown amd benefits greatly by having 4 ranks in dual channel, and all benchmarks done with this type of memory beats the equivalent intel systems by a lot.

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

[alatron978]

Just now, Princess Luna said:

How many % would this one be:

It's not good to deal with absolutes, all in all the statement that the processors are on pair of one another stands.

 

To be fair everyone who has ever had an Intel Stock Cooler under operation knows that thing is as loud as it gets, it is clearly inferior to AMD's and it can get to really bothering levels. Also again with this "something that is slower" there's no objectively faster or slower answer in here.

 

Having the alternative means you can at some point bother yourself to learn and do it, this options is simply not present on Intel which be as much team blue as possible you cannot deny how much a bummer that is.

Now this is the moment you kinda show your bias... the R5 5600X can be used on the Motherboard of choice by OP and that's a processor that does reach and sometimes even surpasses the i9 10900K already making your "beats anything AMD offers" claim false.

 

To add salt to injury, to note you just advised OP to consider an i9 10900K, a 10 cores power house CPU, to be used on the "cheapest B460" you can buy. How's that a senseful upgrade path?

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.

[Mister Woof]

20 minutes ago, Princess Luna said:

I sense some unnecessary bias in here. The reality is that the 10400 and 3600 are too close to one another to really matter, your own "better methodology" benchmarks shows a variance within 1%~2% meaning OP won't be gaining or losing significant performance regardless what they choose.

 

So we get back to the fact that if they are paying the same price, potentially cheaper with the Ryzen if they stick to stock cooler, there's the notion AMD is still a better pick thanks to not locking CPU/Memory overclocking and having a superior upgrade path (without mention actual better performance on other content creation applications should OP ever want those).

I read an entire reddit thread about people raging over them not using geomean vs. averages, including HUB interactions, and basically everyone agreed the difference is not statistically meaningful in any way but that HUB would start using geomean instead of averages to appease the whiners, without labeling it as such to not confuse people since it doesn't really matter.

 

https://www.reddit.com/r/hardware/comments/kyoc08/todays_hardware_unboxed_video_and_how_to_spot_bad/

[alatron978]

1 minute ago, Mister Woof said:

I read an entire twitter thread about people raging over them not using geomean vs. averages, including HUB interactions, and basically everyone agreed the difference is not statistically meaningful in any way but that HUB would start using geomean instead of averages to appease the whiners, without labeling it as such to not confuse people since it doesn't really matter.

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

[Mister Woof]

1 minute ago, alatron978 said:

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

It would matter if they didn't include any of the itemized results and only gave you the final average. The change from average to gemoean was 1% or less.

 

But, you are given that information specific to each game. If you skip to the end and only base your decision on that, without looking at the individual game you care about, that's your fault.

[alatron978]

Just now, Mister Woof said:

It would matter if they didn't include any of the itemized results and only gave you the final average. The change from average to gemoean was 1% or less.

 

But, you are given that information specific to each game. If you skip to the end and only base your decision on that, without looking at the individual game you care about, that's your fault.

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.

[Mister Woof]

 

You won't change my mind.

 

[Mister Woof]

We're splitting hairs here and derailing the intent of the thread by bitching about HUB's methodology which amounts to a whole lot of nothing.

 

OP: Gaming performance is about the same, with some games being better than others on one or another CPU

 

3600 Pros:

Professional tasks/non gaming tasks are better

PCIE4 w/B550 board

More upgrade choices

Doesn't need an aftermarket cooler

 

10400 Pros:

Gaming is just as good as the Ryzen

Costs less

Costs less now than it did 6 months ago

Doesn't need an aftermarket cooler

 

Whatever you get, it will be fine, and you likely won't really regret it. Even in the games that have large differences, the average framerate is already so high it doesn't really matter too much unless you're a 360hz gamer, in which case you're looking at the wrong CPUs anyway.

Regarding Civ6/Zen 3 - that doesn't matter here. We're comparing the 10400 to the 3600.

 

 

[freeagent]

Crazy. My 3600XT with stock cooler never hit 90 once. And it was boosting as it should. Case flow is very important.

 

I saved the 90c temps for when I got an AM4 mount for my cooler 

[Mister Woof]

1 hour ago, WikiForce said:

from what i know, ryzen CPUs do tend to run hot so it might require an aftermarket cooler. I have discord friend who's 3600 was hitting 90C+ with stock cooler and he had to buy a decent aftermarket cooler to gets temps under control

Don't have any personal experience with the Stealth (only the Spire and Prism which are good), but while 90 sounds ominous, if that's under some synthetic stress test that's not altogether unusual or concerning.

 

Intel's stock cooler isn't much better, but it still manages to handle the 10400.

 

Of course everything's relative. Put everything in a mini-itx with no airflow, yeah that might be more of an issue than say, in a big case.

 

It would affect both CPUs relatively the same.

[SavageNeo]

It is odd that people say that 10900k beats everything when it is proven that ryzen 5000 beats the 10900k at gaming.

 

 

I have seen some odd claims that are not true by @alatron978. He should get his facts straight before commenting on public subforums/threads.

I have built around 20-25 pcs in total. I always use the Ryzen stock cooler since it is better than Intel's cooler. from my experience anyway. 

[PrinnyExplodes]

The honest truth is performance is very similar for both. Go with whatever cpu+ram+mobo is cheaper and buy a better graphics card.

[LeapFrogMasterRace]

You guys are crazy a 10400f is the choice hands down unless you are doing compute stuff. The only games a 3600 is faster in are the ones that no one plays like total war. Heck it even beats it in Tomb Raider, GTAV, PUB G, Assasins Creed  battlefield ETC. This AMD ball licking needs to stop. 

[Mister Woof]

34 minutes ago, LeapFrogMasterRace said:

You guys are crazy a 10400f is the choice hands down unless you are doing compute stuff. The only games a 3600 is faster in are the ones that no one plays like total war. Heck it even beats it in Tomb Raider, GTAV, PUB G, Assasins Creed  battlefield ETC. This AMD ball licking needs to stop. 

I think it's better for the value-perspective. Especially given 10400F prices are lower than they were at launch, whereas Ryzen parts are marked up.

 

You can throw it on the cheapest H410 you can find, and it will perform just fine without throttling and stay at max all-core turbo.

 

$170 10400f

$65 H410

$65 16gb DDR4 at least 2666

 

~$300 total

 

Will perform imperceptibly different to a Ryzen 5 3600 on any motherboard with any RAM which will be at minimum $30-50+ more. In regards to compute - if your budget is this low, you aren't making money off the work anyway and the difference doesn't matter since if it did you'd buy a 3900x or something better.

 

But then again, if we're arguing about $30-50, it doesn't really matter. It's really not that important of a thing to pick apart.

 

Get the one that's available that you like, IMO. Won't really matter either way.

 

[_lucalink]

1 hour ago, Mister Woof said:

I think it's better for the value-perspective. Especially given 10400F prices are lower than they were at launch, whereas Ryzen parts are marked up.

 

You can throw it on the cheapest H410 you can find, and it will perform just fine without throttling and stay at max all-core turbo.

 

$170 10400f

$65 H410

$65 16gb DDR4 at least 2666

 

~$300 total

 

Will perform imperceptibly different to a Ryzen 5 3600 on any motherboard with any RAM which will be at minimum $30-50+ more. In regards to compute - if your budget is this low, you aren't making money off the work anyway and the difference doesn't matter since if it did you'd buy a 3900x or something better.

 

But then again, if we're arguing about $30-50, it doesn't really matter. It's really not that important of a thing to pick apart.

 

Get the one that's available that you like, IMO. Won't really matter either way.

 

Especially since 10400f is £130 where I live compared to £190 for the 3600

 

[Nena Trinity]

For gaming its preety much a tie, Hardware Unboxed 22 game average was same how ever as a CPU for traditional work the Ryzen will crush the i5-10400F in every way. If they are same price grab the 3600 if the 10400f is cheaper its worth considering. UwU