14900k - Tuned for efficiency - Gaming power draw

[Vayra86]

Hi,
Yeah op got kicked off that one so he made his own intel love thread.

I don't mind tweaking in bios so more power or in this op case less power to you all out of the box really does not matter to them

People I feel sorry for is the owners/ buyers of H mobile chips instead of more expensive HX chips they are stuck with locked chips.
I found out this just in time so I canceled and went amd mobile locked but at least not a space heater

I don't get the disabling HT either frankly
The enemy is thermal defective e=cores not HT lol

Its obvious why this topic is reignited; at the same time, why not. Discuss the technicalities of undervolting Intel. Nothing wrong with that.

Live and let live

 

[freeagent]

6GHz at 50w looks pretty good to me. He isn't impressed because the box does not say AMD on it

 

[AusWolf]

6GHz at 50w looks pretty good to me. He isn't impressed because the box does not say AMD on it

I'm not impressed because it's done with only 3 cores with HT off. Regardless of the clock speed, actual application performance must be abysmal. I'm a practical person. I don't care about clock speed.

 

[freeagent]

I don't care about clock speed.

Did you know a dual core E8600 running at 4.7GHz is just as strong, but much faster than a stock Q6600?

Now you do

 

[AusWolf]

Did you know a dual core E8600 running at 4.7GHz is just as strong, but much faster than a stock Q6600?

Now you do

Yeah, but now you're talking about actual game/application performance of a fully enabled CPU, which I agree, clock speed has an effect on.

Somehow, I can't see a 3-core HT-disabled Raptor Lake CPU being useful in modern programs/games at any GHz (unless you're talking about browsing and/or video playback at most).

 

[freeagent]

Yeah, but now you're talking about actual game/application performance of a fully enabled CPU, which I agree, clock speed has an effect on.

Somehow, I can't see a 3-core HT-disabled Raptor Lake CPU being useful in modern programs/games at any GHz.

Maybe so.. but the fact that he did it with 50w is pretty neat.. don't you think?

 

[AusWolf]

Maybe so.. but the fact that he did it with 50w is pretty neat.. don't you think?

Dunno... it's still just a meaningless number to me.

I'll let the hardcore OC-enthusiasts decide.

 

[freeagent]

Dunno... it's still just a meaningless number to me.

I'll let the hardcore OC-enthusiasts decide.

It has already been decided

 

[vidal6x6]

I'am playing with a 13100T gives a cpu power compsumation about 7 14w some loads 35w

 

[Gica]

This thread feels like an extension of the other 14900k vs 7800X3D thread, but without mentioning the 7800X3D over here because if you did, a lot of people would probably laugh at you.

Sure you can tune the CPU to be more efficient, but did you actually gain anything? If anything you're going to lose performance in the grand scheme of things unless your thermal solution is just woefully not up to the task. At least the 7800X3D works well out of the box without being a space heater.

Because I can't compare otherwise, I look at the Guru review and see that the 13500 beats the 7800X3D well in both tests. In the TPU review, the same 13500 is about equal to the 7800X3D in singles and crushes it in MT. Tell me, please, how much X3D can help a 3070Ti to make it worth paying double?
X3D have their beauty, but they are definitely not worth that much money. If you don't want Intel, well, you have the cheaper and faster 7700(X) variants in applications and with very good behavior in games.
You omit or you didn't even know that the processor is not 100% in games and you don't use it to render 100% of the session. I'm writing this post with a wattmeter that doesn't jump at 30W. I'm curious what the wattmeter indicates when an AMD owner writes on TPU.
Puget has review with 14900K@125W for Content Creation, TPU for gaming. Losses between 0 and nothing.
You can't compare 7800X3D with 14900K because processors are not just for gaming.

 

[ShrimpBrime]

Dunno... it's still just a meaningless number to me.

I'll let the hardcore OC-enthusiasts decide.

Well at first 253w really really meant something to you. (Other people too!)

And now 50w means nothing!!!

 

[vidal6x6]

Well at first 253w really really meant something to you. (Other people too!)

And now 50w means nothing!!!

253w it's the power compsumation of my sound / monitor and case when gaming / idle 60w.

 

[ShrimpBrime]

Yeah, but now you're talking about actual game/application performance of a fully enabled CPU, which I agree, clock speed has an effect on.

Somehow, I can't see a 3-core HT-disabled Raptor Lake CPU being useful in modern programs/games at any GHz (unless you're talking about browsing and/or video playback at most).

Oh, here we go.

Let me try and explain.

So certain benchmarks do well with a nice high cpu frequency. So I wanted to try a rather high frequency with a core reduction. Effectively helping the benchmark.

GPUPI is not real cpu heavy, using 1 thread on the cpu. The benchmark runs most load on the GPU. It's 32B calculation both.

A lot of legacy 3D is also not cpu heavy. But cpu frequency sensitive. Such as Aquamark. Also done at 6ghz 3 cores. Hit 57w!!

And again, cause it was ignored.

On air cooling.

Edit.

For modern benchmarking, this would not be stable enough to daily.
Cinebench single core would be a modern use I suppose.

6 GHz is just an arbitrary number, which doesn't mean anything in scenarios where it matters, that is, doing actual work, or gaming. Especially with only 3 cores running with no HT. So no, I'm not impressed.

That is my work. For my benchmarking team. Each submission contributes to the team.

Many benchmarks are legacy and work with modern hardware.

So I'm actually doing stuff with it. That contributes to a data base. Where benchmark scores come from. The world records Splave, The Stilt, Bones, FastTrack and many others mean something. At least to them!

What would be impressive? More dryice screen shots? That's not even practicle cooling.

What is the best way to demonstrate, I will try to accommodate. (I'll need my water loop I bet, fine, I'll set it back up.)

 

[freeagent]

Ha!

That is the funny thing about the internet.. you never really know who you are talking to until they tell you something

 

[Vayra86]

Dunno... it's still just a meaningless number to me.

I'll let the hardcore OC-enthusiasts decide.

Well, given the fact you can do this on air, I say, this is pretty neat, I mean it shows how far these Intel procs stretch, both towards the top and the bottom, all at the same time. I don't think Zen has nearly that tweakability; you can tweak it, but not to these extremes.

Say you love Starcraft. 6 Ghz 13th gen core will play that game so hard.

You can't compare 7800X3D with 14900K because processors are not just for gaming.

Of course you can. It all depends on your own use case. In my view, a good CPU for my use case is one that doesn't fall short on the 'other' stuff I do on it, while it excels most in the stuff I do most on it.

So sure thing, there are gaming CPUs, and there are workstation oriented CPUs, and you wouldn't use a Threadripper to game, nor a Xeon Platinum or an EPYC.

What happened in the mainstream corner of CPUs is that HEDT and MSDT segments kinda blended together. The only real limitation is RAM when it comes to performance. You can get blazing fast storage on either segment and you can get a wild number of cores too. And both Intel and AMD's top end are really HEDT class CPUs, you don't need nor want them really for gaming. So you can compare 14900K and 7800X3D fine - one is clearly more versatile than the other, while the latter is clearly better suited for gaming. Its as simple as that. And then, its also pretty damn obvious you need to tweak a 14900K downwards to get it closer to the perf/w of an 8 core 7800X3D. What's cool about the 14900K though is that you can; and you can't make an X3D grow cores.

I think its pretty cool we're in a CPU situation now where there really aren't clear winners, there are simply numerous different tools to optimize on different tasks.

 

[ShrimpBrime]

Ha!

That is the funny thing about the internet.. you never really know who you are talking to until they tell you something

Gonna go to the hardware store in a little while. Get some more tubes. Sink drain in the kitchen needs replaced too.

I should really be more realistic here. After all, they are talking gaming.

Ok then, what game is going to utilize my 24 threads in total where leaving everything stock would be a benefit?

 

[Vayra86]

Gonna go to the hardware store in a little while. Get some more tubes. Sink drain in the kitchen needs replaced too.

I should really be more realistic here. After all, they are talking gaming.

Ok then, what game is going to utilize my 24 threads in total where leaving everything stock would be a benefit?

Damn. MS Flight Simulator?

Although the verdict I believe is already out, that more cores while 'utilized' don't really improve perf that much. Its that way with most games. Some game logic/data just won't pass the pipe faster no matter what.

 

[ShrimpBrime]

Damn. MS Flight Simulator?

Although the verdict I believe is already out, that more cores while 'utilized' don't really improve perf that much. Its that way with most games. Some game logic/data just won't pass the pipe faster no matter what.

Ok Flight Sim for thread count.

What would be another demanding game that would not use all threads but would like Frequency best?

 

[Vayra86]

Ok Flight Sim for thread count.

What would be another demanding game that would not use all threads but would like Frequency best?

Stellaris > 100 years of building. Cities Skylines, similarly, just load up some huge project

Another one might be Riftbreaker for some multi threaded DX12 goodness, I mean those alien hordes are damn intense on CPU

Total War Warhammer 2 is another nice one, count the end turn button > start of next turn, that's all CPU.

 

[Super Firm Tofu]

Gonna go to the hardware store in a little while. Get some more tubes. Sink drain in the kitchen needs replaced too.

I should really be more realistic here. After all, they are talking gaming.

Ok then, what game is going to utilize my 24 threads in total where leaving everything stock would be a benefit?

Cyberpunk 2077 has been known to use all/most available threads.

 

[Vayra86]

Cyberpunk 2077 has been know to use all/most available threads.

View attachment 325762

I'd be very interested how much all those cores really matter without a GPU limitation.

 

[ShrimpBrime]

I'd be very interested how much all those cores really matter without a GPU limitation.

That's easy, lower the resolution and detail.

Want to test 1 game.

All the games you've mentioned, I'd have to buy. Because I do benchmarking usually.

See if a bencher can tweak in some game performance.

Default testing is where I'll start.
Tweaked for FPS and wattage reduction is where I stop.

 

[ThrashZone]

Here's the Direct Difference between a K and non K model.

This is cooled via Wraith Prism, not clamp, diamond nano paste.

50w 6ghz anyone?

Note the date time. That was just a couple minutes ago..... And I just typed this out to you on it.
Edit: Central US time.

Hi,
Those silly world record creators with 9GHz that high hell they could of done it with 100w

 

[Vario]

Ok Flight Sim for thread count.

What would be another demanding game that would not use all threads but would like Frequency best?

Space Engineers

 

[Super Firm Tofu]

I'd be very interested how much all those cores really matter without a GPU limitation.

Searched through some older screenshots. GW2 is usually always CPU limited. It's almost impossible to benchmark do to the nature of online and never the same scene twice. Still utilizing all the cores, but pegs P-cores 3 and 4 and still only 40fps.