Intel Shows That Their 9th Gen Core CPU Lineup Is Faster Than AMD Ryzen 3000 In Everything Except Cinebench

[PotatoWithEarsOnSide]

I take no exception to this post, and you are not entirely wrong. However, it brings me to an observation, and that is: Man, how the worm turns.

Dead platform lacking the newest features the competition's boards have. Check

Need for aftermarket cooling and even physical mods. Check

Also requires the more expensive boards. Check

CPU maker spreads disinformation and/or misinformation trying to make their products look better than they are. Check

Oh they get it. But it serves the purpose of obfuscating what this discussion is actually about. I.E. Intel going full Baghdad Bob again.

Could have sworn that his name was Chemical/Comical Ali.

 

[TheELF]

Look, nobody said anything about CPU utilization. He said the programs are bottlenecked by something else. When that happens, it means it is the bottleneck (GPU, RAM, storage, whatever) that determines application performance. Not the CPU.

It's the CPU that determines how well the GPU, RAM, storage or whatever will run,the faster the CPU can talk to the other devices the faster they will run, how is that not part of "the performance" everybody is talking about?

Are you just parroting Intel's marketing?

When you compare a 3900X to a 9900K...

Web (Anandtech link) - used instead of TPU because they used the re-ran test using updated BIOS after 7/7
3900X is better at:
WebXPRT 3
WebXPRT15
Speedometer 2
Google Octane 2.0
Mozilla Kraken 1.1

9900K is better at:
?

Wow man,did you even look at those numbers?
The 9700k is on the same level as the 3900x or even beating it...you are thinking that the 9900k is worse than the 9700k?

That's what you get if you only look at locked TDP results,sure they are important but you should also know how much more potential there is for someone that cares.
If you don't care about temps or power or anything else you can overclock the 9900k by a rather large margin,you have to look at those numbers as well.

 

[Abwx]

Wow man,did you even look at those numbers?
...you are thinking that the 9900k is worse than the 9700k?

That's what you get if you only look at locked TDP results,

Yes, it is worse, that s what happen when you look.....at what exactly?..

 

[amrnuke]

It's the CPU that determines how well the GPU, RAM, storage or whatever will run,the faster the CPU can talk to the other devices the faster they will run, how is that not part of "the performance" everybody is talking about?


Wow man,did you even look at those numbers?
The 9700k is on the same level as the 3900x or even beating it...you are thinking that the 9900k is worse than the 9700k?

That's what you get if you only look at locked TDP results,sure they are important but you should also know how much more potential there is for someone that cares.
If you don't care about temps or power or anything else you can overclock the 9900k by a rather large margin,you have to look at those numbers as well.

Not sure why you're trying to argue against my statement. Perhaps you didn't bother to read the context.

The results are the results. Despite a clock speed advantage of 8.6% over the 3900X (and a clock speed advantage over the 9700K and the 3700X), the 9900K LOSES to all of them in all the web tests.

I posted what I posted in response to a guy who seemed to have the stance that clock speed is all that matters for basic productivity like browsing the web, watching videos, and using office suites. He is wrong. And if you're trying to argue that he is correct, you are wrong as well.

 

[maddie]

Not sure why you're trying to argue against my statement. Perhaps you didn't bother to read the context.

The results are the results. Despite a clock speed advantage of 8.6% over the 3900X (and a clock speed advantage over the 9700K and the 3700X), the 9900K LOSES to all of them in all the web tests.

I posted what I posted in response to a guy who seemed to have the stance that clock speed is all that matters for basic productivity like browsing the web, watching videos, and using office suites. He is wrong. And if you're trying to argue that he is correct, you are wrong as well.

Ha. Don't you know that theory always beats empirical data nowadays?

 

[Guru]

Going with a 3600 cpu next time around, and for about $40 buying a decent cooler, though with that said might add $10 more and go for the 3600x and get the better cooler and slightly faster performance.

DDR4 ram is still freaking expensive, so its preventing me from upgrading my pc.

 

[TheELF]

The results are the results. Despite a clock speed advantage of 8.6% over the 3900X (and a clock speed advantage over the 9700K and the 3700X), the 9900K LOSES to all of them in all the web tests.

Yeah it shows that you have no idea about computers.
The 9900k and the 9700k have the same TDP rating,this means that the 9900k has to feed twice the threads with the same power if you lock the CPU to this TDP.
So the 9900k does not have a clock speed advantage in this benchmark!
Quite the opposite, because it has to feed twice the threads with the same power it runs all cores at lower clocks.
This is why so many benchmarks are very deceitful.

 

[Abwx]

Yeah it shows that you have no idea about computers.
The 9900k and the 9700k have the same TDP rating,this means that the 9900k has to feed twice the threads with the same power if you lock the CPU to this TDP.
So the 9900k does not have a clock speed advantage in this benchmark!
Quite the opposite, because it has to feed twice the threads with the same power it runs all cores at lower clocks.
This is why so many benchmarks are very deceitful.

Those tests use few threads, otherwise the 7700K would be far behind, and even if it was using 8 threads why should a 9900K use more power than (a not as well binned...) 9700K loaded with 8T..?...

Besides the 9900K boost up to 119W even when set at 95W in the bios, so there s absolutly no power limiting factor in the web benches.

As for frequency being a factor the Ryzen scores say that there s not as much difference in the scores than in the frequencies, the cache/core seems to count as well.

 

[Thala]

Those tests use few threads, otherwise the 7700K would be far behind, and even if it was using 8 threads why should a 9900K use more power than (a not as well binned...) 9700K loaded with 8T..?...

Besides the 9900K boost up to 119W even when set at 95W in the bios, so there s absolutly no power limiting factor in the web benches.

As for frequency being a factor the Ryzen scores say that there s not as much difference in the scores than in the frequencies, the cache/core seems to count as well.

Still it raises the question why the 9900K is significantly slower than the 9700K - this seems at least counter intuitive or suspicious.

 

[Thunder 57]

Still it raises the question why the 9900K is significantly slower than the 9700K - this seems at least counter intuitive or suspicious.

Oh come on, don't use so much common sense! My guess is because some things still rarely run slower when SMT is enabled.

 

[DrMrLordX]

Will he laugh the same way, when he sold 8700K, bought slower CPU, and clocking worse, than 8700K?

He already said he gained fps in CS:GO, among other things. You sure he got a slower CPU?

It's the CPU that determines how well the GPU, RAM, storage or whatever will run

That is not the case. Unless you're somehow implying that AMD's platform and/or memory controller are so hosed that they can't handle RAM correctly or otherwise slow down device I/O. Which would be terribly wrong.

When the CPU is not the bottleneck, the CPU is not the bottleneck.

 

[Abwx]

Still it raises the question why the 9900K is significantly slower than the 9700K - this seems at least counter intuitive or suspicious.

The three SMT enabled CPUs show relative results that are in line with their frequency difference.

Dunno From where come the discrepancy with the non SMT CPU but the 8700K has comparable ST perf as a 9700K, yet it is well behind, perhaps that security patches have more impact on SMT enabled CPUs since those security flaws are precisely a SMT uncautious implementation, and that the 9700K doesnt need to run the SMT related patches.

 

[moinmoin]

Going with a 3600 cpu next time around, and for about $40 buying a decent cooler, though with that said might add $10 more and go for the 3600x and get the better cooler and slightly faster performance.

Just a note of warning, the stock cooler included in 3600x (Wraith Spire 2, all aluminum) is no longer the same as the one included in e.g. 2600x (Wraith Spire, aluminum with a copper core, that was the good one).

 

[amrnuke]

Yeah it shows that you have no idea about computers.

Ironic.

The 9900k and the 9700k have the same TDP rating,this means that the 9900k has to feed twice the threads with the same power if you lock the CPU to this TDP.

If you lock the CPU to the TDP and basically ONLY supply it with 95W of power, in theory, you might be right. Anandtech and TPU and TechSpot and TomsHardware and every other review I've seen have NOT done this. So it's still a might.

But what is undoubtedly true is that your theoretical TDP limitation is NOT how the review benchmarks were done. That is also NOT the real world. That is a synthetic world. In the 9900K review you can see that the 9900K runs at peak draw of 166-169W and the 9700K at 123-125W. They operated with the same cooling setup. Somehow, both chips drew more power than their TDP! Gasp! So the 9900K did NOT have to feed twice the threads with the same power, because they didn't lock the CPU to the TDP. I'm not even sure why you brought that point up.

So the 9900k does not have a clock speed advantage in this benchmark! Quite the opposite, because it has to feed twice the threads with the same power it runs all cores at lower clocks.

That's a bold statement! I assume you have a copy of the charts showing clock speeds on each core during the office test benchmarks to back up your claim? And while we're at it, power draw figures during those same tests? If you're making the claim without evidence, I'm sorry, but I will dismiss your claim without any evidence as well.

This is why so many benchmarks are very deceitful.

No. It would be far more deceitful for reviewers to lock all the TDPs to their listed rating, and then claim a performance king. How well do you think the 9900K would do if reviewers only let it draw 95W? Not only is that unrealistic (in that if you slap a cheap cooler on it, it will still happily draw 125+W without any thermal throttling), that would be incredibly stupid and a waste of time, since no one cares how the chip performs at its TDP. They care how it performs when cooled well (because anyone spending half a grand on a CPU is probably not going to put a passive heatsink on it).

Circling back, I'll just lob this one back to you:

Yeah it shows that you have no idea about computers.

 

[PotatoWithEarsOnSide]

I don't think it is really a debate about whether either has knowledge about computers, rather a debate about which of you can apply logic and common sense.