7800X3D vs 14900K video by HWUB. What would you choose for gaming?

[SL2]

Steve had a troll gleam in his eyes...

This just in! HUB is back with more's gleams! [/irony]

Spoiler: Offtopic

The only problem with the power consumption excuse is that you're saving on the CPU and using it on the GPU.

My point was 100 % about cost. I was about to add 4070 as well but thought people would understand that. Like I said, it was just an example.
I've edited my post, just for you.

But even if we stick with my original examples, you'd get better gaming performance from a 7900 XT than a 7800 XT, and a 14900K won't change that fact. Saving on the CPU power draw and using that power on the GPU is a good thing.

Wasn't a problem when AMD did it with the FX-9590, I reckon.

Well, you surely missed out if that's what you bought. I got a 2600K for $140 the same year, the best deal I've ever got on a new CPU.

 

[Psychoholic]

The power limit affects efficiency only when the CPU starts hitting the limit. When you allow P cores to run at 5700MHz while gaming, they will eat a lot of power, that frequency is way out some "acceptable efficiency frequency interval".

Correct, that's where the VID undervolt comes in
But yes, with power limit only, you still have the full 180w concentrated on the P Cores.

I actually have 2 profiles saved in bios from messing around with it, one of them has P-Cores limited to 5.4ghz.

Its crazy how far motherboard vendors overvolt these things, on top of intel's insane power limits.

 

[Aquinus]

Its crazy how far motherboard vendors overvolt these things, on top of intel's insane power limits.

Oh, oh, I have a meme for that. Applies to Intel and board manufacturers equally.

 

[BoggledBeagle]

Its crazy how far motherboard vendors overvolt these things...

I always thought that they put in what the CPU asks for and the voltage safety margin is built in by CPU manufacturer.

 

[Warrior24_7]

I saw this video

and I got a strong vibe of it not being very objective. I got even a feeling that Steve will not be proud of this video after some time passes.

If you have a product A 5% stronger in task GAM than product B, and product B 110% stronger than A in task PROD, it seems more than silly making a video bashing product B because it is 4.8% weaker than product A in GAM.

And even if you mention in that video, that B is much stronger than A in PROD, it changes little about the general meaning of the video.

I wonder what would happen if he compared 7700X with 14600K (direct price competitors) and was not shy to show all aspects of CPU performance in the same manner (graphs, etc).

(And BTW 14900K can be tuned nicely to increase power efficiency for people who value it.)

I don't pay any attention to these people and you shouldn’t either! These people are “sponsored!” That means they’re taking money, products, favors, or something from AMD. Their “JOB” is to steer your buying decisions. That’s what they’re being paid to do here. They’re paid to influence you. Hence the term “influencer.” He said it in the beginning of the video, “we’re going to see which one you should buy.” He also said this is an “updated” video. Suggesting that Intel won the first head-to-head comparison. Go figure. AMD is spending A LOT of money to get your money and they’re using these YouTube hacks to do it!

 

[Vayra86]

I don't pay any attention to these people and you shouldn’t either! These people are “sponsored!” That means they’re taking money, products, favors, or something from AMD. Their “JOB” is to steer your buying decisions. That’s what they’re being paid to do here. They’re paid to influence you. Hence the term “influencer.” He said it in the beginning of the video, “we’re going to see which one you should buy.” He also said this is an “updated” video. Suggesting that Intel won the first head-to-head comparison. Go figure. AMD is spending A LOT of money to get your money and they’re using these YouTube hacks to do it!

Correct, but the numbers and conclusions aren't useless either. Especially given the fact the 'influencing' happens for every company. Its marketing. Yes. Youtube videos are built on marketing and ads.

Like most things commercial, this is how these things work. The best approach is to be conscious of that, and not give it too much credit. That said, there's no lie in that HWUB video That's another category of influencing right; the ones with outright lies or misrepresentation, which was the focal point of the OP, but in fact, there is none of that. Its true the 7800X3D is the best gaming CPU at this time, and its also true its a much better choice for gaming than an Intel 14900K.

 

[freeagent]

From what I have seen, capping power on AMD does hurt performance, I can only assume the same for Intel.

And quit it with the fanboy talk, it makes you all sound like fanboys lol.

 

[Psychoholic]

From what I have seen, capping power on AMD does hurt performance, I can only assume the same for Intel.

And quit it with the fanboy talk, it makes you all sound like fanboys lol.

makes no difference in light loads but of course it does for all core loads.
Going from 253W to 180W took my cinebench score from around 38,500 to around 37000.

 

[Upgrayedd]

TPU uses a 14900K in their bench system. That should say SOMETHING, though it could use some faster ram. I could've bought whatever in a new PC I'm putting together but I picked a 14700K, 4 little e cores weren't worth the price premium. I just don't even really consider AMD either, too many weird little problems over the years.

 

[BoggledBeagle]

From what I have seen, capping power on AMD does hurt performance, I can only assume the same for Intel.

I just ran a Cyberpunk built-in benchmark on a 4070 for two frequency configurations: stock 5700/4400 and lowered 5000/4000.
Power draw during benchmark was in the first case 104-117W and in the second 55-62W.

Lowering a frequency by 10-13% nearly halved the power draw. And the FPS are THE SAME to a single digit !!! I increased the efficiency twice.

The power limit of 150W has in this case no effect at all, it just ensures that no matter what my cooler can handle the CPU easily.

Note that I have HT off because with 24 cores I believe I do not need it at all.

Intel CPUs are really not that horrible, they are just stupidly set by the manufacturer.

 

[Upgrayedd]

I just ran a Cyberpunk built-in benchmark on a 4070 for two frequency configurations: stock 5700/4400 and lowered 5000/4000.
Power draw during benchmark was in the first case 104-117W and in the second 55-62W.

Lowering a frequency by 10-13% nearly halved the power draw. And the FPS are THE SAME to a single digit !!! I increased the efficiency twice.

The power limit of 150W has in this case no effect at all, it just ensures that no matter what my cooler can handle the CPU easily.

Note that I have HT off because with 24 cores I believe I do need it at all.

Intel CPUs are really not that horrible, they are just stupidly set by the manufacturer.

Did you try a few mouse quick mouse whips or going to high density npc areas and doing a test path? Built in benches aren't the real deal.

 

[A Computer Guy]

I just ran a Cyberpunk built-in benchmark on a 4070 for two frequency configurations: stock 5700/4400 and lowered 5000/4000.
Power draw during benchmark was in the first case 104-117W and in the second 55-62W.

Lowering a frequency by 10-13% nearly halved the power draw. And the FPS are THE SAME to a single digit !!! I increased the efficiency twice.

The power limit of 150W has in this case no effect at all, it just ensures that no matter what my cooler can handle the CPU easily.

Note that I have HT off because with 24 cores I believe I do not need it at all.

Intel CPUs are really not that horrible, they are just stupidly set by the manufacturer.

Are you sure you increased the efficiency? If you lowered the frequency and the FPS didn't change doesn't that just mean your already running the game at its maximum frame time with your hardware?
In other words are you GPU limited allowing you to scale back your excess CPU overhead (frequency & power)?

 

[BoggledBeagle]

I just ran a Cyberpunk built-in benchmark on a 4070 for two frequency configurations: stock 5700/4400 and lowered 5000/4000.
Power draw during benchmark was in the first case 104-117W and in the second 55-62W. Lowering a frequency by 10-13% nearly halved the power draw. And the FPS are THE SAME to a single digit !!!

I added one more data point, so the results are:

5700/4400MHz - 104-117W - av. fps 75 - max fps 101
5000/4000MHz - 55-62W - av. fps 75 - max fps 101
4500/3600MHz - 42-48W - av. fps 75 - max fps 100

So that 24 core CPU can fully serve a 4070 in this game and settings with just 50W of power. Perhaps I should repeat the test with less extreme settings...

EDIT: Ok, now Cyberpunk with RT off, DLSS off, settings high:

P cores MHz/E cores MHz - power - avg/min/max fps

5700/4400 - 122-160W - 90/73/122 *
5000/4000 - 66-83W - 94/73/126
4500/3600 - 48-63W - 95/58/126

*game drew more than 160W before I got to the benchmark and then it overheated during benchmark under my air cooler, I removed a power limit for this test

It seems that the lowest settings impacted min FPS, but then again, insane stock frequencies with my 4070 card doubled power draw for no benefit compared to 5000/4000 settings (in my case even worsening performance due to overheating).

I am done with testing. I hope this shows that these CPUs (14700K and 14900K) need to be adjusted and that they are not that horrible after that. The stock frequencies are simply too high for the process Intel is using to make these CPUs.

EDIT 2: It seems that the dip in min FPS was caused by E cores running at 3600MHz, combination 4500/4000 was fine.

 

[AusWolf]

Correct, but the numbers and conclusions aren't useless either. Especially given the fact the 'influencing' happens for every company. Its marketing. Yes. Youtube videos are built on marketing and ads.

Like most things commercial, this is how these things work. The best approach is to be conscious of that, and not give it too much credit. That said, there's no lie in that HWUB video That's another category of influencing right; the ones with outright lies or misrepresentation, which was the focal point of the OP, but in fact, there is none of that. Its true the 7800X3D is the best gaming CPU at this time, and its also true its a much better choice for gaming than an Intel 14900K.

Well said. Not listening to these "influencers" purely on the premise that they are influencers (God, I hate this word) is just as moronic as believing their every word without question. A wise man (or woman) can watch a video or read a review while being able to differentiate between useful, useless and false information and clickbait.

 

[freeagent]

They influence the weak

The rest of us have brians

 

[thesmokingman]

This thread is entertaining. The conclusion is clear, yet some keeping coming back like they have a dog in the fight.

 

[Frick]

They influence the weak

The rest of us have brians

The correct spelling is Brian, you apostate.

 

[SL2]

“sponsored!”

Explain to me what happened between these two videos then.

They were posted four days apart. And don't give me some BS about Intel just paid them more lol. There are some severe fancreatures in this thread, simple as that.

"If I don't like what I see then it must be a lie.."

 

[A Computer Guy]

Explain to me what happened between these two videos then.

They were posted four days apart. And don't give me some BS about Intel just paid them more lol. There are some severe fancreatures in this thread, simple as that.

"If I don't like what I see then it must be a lie.."

I don't understand the rub. In the first video FRS and XeSS are not used in the benchmarking so these videos are totally different.

 

[fancucker]

The 14900K provides

1. Superior RAM scaling and higher RAM freq. support
2. Thunderbolt
3. Better motherboard availability
4. Quicksync, a phenomenal technology
5. Better workstation capabilities and all-roundedness
6. More stable software support and execution (check out Jayztwocents and why he switched)
7. Far more stable system after build
8. The opportunity to support a country currently experiencing brutalities

The choice is clear.

 

[nguyen]

Dear @W1zzard , we need a 50 games benchmarks between 14900K vs 7800X3D

 

[oxrufiioxo]

Explain to me what happened between these two videos then.

They were posted four days apart. And don't give me some BS about Intel just paid them more lol. There are some severe fancreatures in this thread, simple as that.

"If I don't like what I see then it must be a lie.."

Honestly nothing wrong with it other than fanboys championing it every chance they get FSR sucks.

 

[BoggledBeagle]

The 14900K provides
...
Better motherboard availability
...
More stable software support and execution (check out Jayztwocents and why he switched)
...
Far more stable system after build

While I agree with the motherboard availability problem (there are not many X670E motherboards available and they are very expensive), I had a weird feeling from Jays video. What he described seemed like a faulty CPU or motherboard and he should have troubleshot the problem and solve in a standard way. I think a part of his switch was that he got some of those parts for free and he could not use a warranty repair/exchange, and that he had some Intel parts again available for use (possibly again for free).

I doubt that instability of AM5 systems is very wide spread, because you would see a lot more complaints. What I see in customer feedback are/were mostly complaints about long boot times, and sometimes people even give up setting up the systems because they do not know about the memory training and they think the system is broken. The greedy motherboard manufacturers refusing to provide post code displays on motherboards do not help here at all.

I must say that setting up the Intel systems has been very easy and problem free recently for me, I just got one problem caused by a faulty SSD, and it was not an instability, just error spamming in system event viewer.

 

[DemonicRyzen666]

Honestly nothing wrong with it other than fanboys championing it every chance they get FSR sucks.

All upscaling is bad, regardless of who or how it's being made. It's just stifling real performance gains now at this point.

I don't even use AA & AF while gaming.

 

[chrcoluk]

This thread is entertaining. The conclusion is clear, yet some keeping coming back like they have a dog in the fight.

I seen it all coming, amd vs intel is pure fanboy stuff now. Even to the point youtubers are now pushing out troll videos to fuel it.