10900K hit 200W too but with 2 more cores, showing there's indeed a regression in power consumption per core. While it's still too early to tell, 11900K's efficiency advantages probably will all be used in extra clockspeeds... and early pricing suggests Intel is positioning it against the 5900X anyway.

While CPU temps are not as hot as Ryzen for the TDP, there's also how much heat is being expelled out of the PC to the room, this matters when one isn't using air conditioning in Malaysia. Ryzen having a small CPU die and thus has worse heat transfer to the IHS is a known issue. There's still the other issue on how to dump 200++ watts of heat out vs ~142W from the 5800X. The only way to avoid this while sticking to Intel, unfortunately, is using a 6-core and not overclocking... and the RKL benchmark really suggests overclocking is what's required to close the gap against Ryzen in performance.

As for AVX-512 - TBH while there's a use for it, most real world consumer applications that can benefit from it are better off running CUDA/OpenCL and using the GPU anyway... Hence why I think the 290W power draw on Anandtech is a bit overblown.

10900K hit 200W too but with 2 more cores, showing there's indeed a regression in power consumption per core. While it's still too early to tell, 11900K's efficiency advantages probably will all be used in extra clockspeeds... and early pricing suggests Intel is positioning it against the 5900X anyway.

While CPU temps are not as hot as Ryzen for the TDP, there's also how much heat is being expelled out of the PC to the room, this matters when one isn't using air conditioning in Malaysia. Ryzen having a small CPU die and thus has worse heat transfer to the IHS is a known issue. There's still the other issue on how to dump 200++ watts of heat out vs ~142W from the 5800X. The only way to avoid this while sticking to Intel, unfortunately, is using a 6-core and not overclocking... and the RKL benchmark really suggests overclocking is what's required to close the gap against Ryzen in performance.

As for AVX-512 - TBH while there's a use for it, most real world consumer applications that can benefit from it are better off running CUDA/OpenCL and using the GPU anyway... Hence why I think the 290W power draw on Anandtech is a bit overblown.

AVX-512, even if it's little use it is still better to have one instead of not having it right? It has some niche uses that some people need. Also with the RKL having it, it'll also mean that more apps will begin to include it in the future.

By the time there are more than a handful of mainstream apps that utilizes AVX512, RKL would be a pointless chip already.

By the time there are more than a handful of mainstream apps that utilizes AVX512, RKL would be a pointless chip already.

AVX-512, even if it's little use it is still better to have one instead of not having it right? It has some niche uses that some people need. Also with the RKL having it, it'll also mean that more apps will begin to include it in the future.

the cpu is not a cooker lol, so 250w of power draw does not equal to 250w of heat. Infact heat is technically wasted energy... so if you factor that in, the AMD is actually less effecient at using the powerdraw due to the chiplet design. The heat gone to the room is technically what you see in the temp monitoring as that is what is actually being absorbed/expelled.

So it actually makes no diff, infact due to intel’s more effecient boost algorithm, it is a much cooler cpu overall. For example, you dont run cinebench or prime95 24/7. Even for rendering, intel tends to run cooler too.. so when it comes to gaming, there’s no comparison thanks to intel’s more mature clock stepping.

By the time there are more than a handful of mainstream apps that utilizes AVX512, RKL would be a pointless chip already.

avx has long been on xeons actually. So having this may provide a reason to buy 11400F for example as budgeted office workstations.

I don't really appreciate the anand review. It's not the REVIEW they used to be like the old days. Started the review saying it's a hot hot chip and bulldozing everything else.
At best it's termed a Garbage Review and not worth even reading it.
Anand and Toms hardware - I don't follow their stuff for a very long time.
They are NOT the CREDIBLE sites they used to be (after change of ownership)
Just look at the Toms layout and the stuff there, it's garbage.

While you're right on that, there's no question that a 10700K @ 200-250W will still expel more heat than a 5800X @ 142W.

https://www.anandtech.com/show/16535/intel-...-rocket-lake/19

11700K really seems disappointing without overclocking. The core-to-core and L3 cache latency regressions vs Comet Lake means 5800X still easily beats it in gaming.

Power consumption when TDP is ignored, without manual core multiplier settings is also worse than Comet Lake - 225W for most work, 290W for AVX-512. Although to be fair, it's the only desktop chip that does AVX-512 right now.

Given the leaked pricing, I think the 8-cores appear to be a bust, 6-core value chips is where RKL will shine the most as it's much more competitively priced against the 5600X there.

The cooler we’re using on this test is arguably the best air cooling on the market – a 1.8 kilogram full copper ThermalRight Ultra Extreme, paired with a 170 CFM high static pressure fan from Silverstone. This cooler has been used for Intel’s 10-core and 18-core high-end desktop variants over the years, even the ones with AVX-512, and not skipped a beat. Because we’re seeing 104ºC here, are we failing in some way?

Honestly Intel flopped this generation. They’re trying to bait consumers by accepting preorders 2 weeks before NDA (riding on AMD’s chip shortage) and no wonder. A hotter chip that’s pricier across the board and can’t even beat their AMD equivalent in 90% of benchmarks. Unless the 11700k is somehow magically the only bad apple in the barrel, I say wait for alder lake instead.

Slower than the 10700K in some games and consumes nearly twice the amount of power than the 5800X





Also, looks like you need a high end AIO or a custom water cooling loop, if this is true:

Only during constant high synthetic powerload. Average and gaming usage ryzens generally will expel more heat especially during light loads and idle.

And even for max load, the 5800x will also fly to 90C while intel 10900K will fly to 93C. So even if the intel cpu is draining 100w more, it’s only dispelling about 3C more... and i would also reiterate intel’s more effecient clock stepping that allows the cpu to run 60C at long continuous loads while AMD goes full power even for youtube.

so i would say intel is like a car with 8speed gearbox and an inefficient engine while amd is like a 3 speed gearbox with a high effecient engine. But thanks to the wider rev range, in the long term, the intel car will be using fewer fuel and running cooler.

Youtube video will not go to full power on AMD, not sure what's wrong with your browser... unless you're talking about 8K.
4K Costa Rica on VP9 software decoding mode not really that cpu intensive, even my old r3 1300x can play it, just not smoothly due to slow cpu.

bruh.. that thermalright cooler cannot even cool the nehalem i7 920 suffeciently lol. You can say it performs like a modern 120mm aio at best. It even got replaced by sunbeam CCF in its era.

this is zen2 onward’s issue. newer agesa has solved the aggresive boosting specifically on chrome but still very dependent on what is running at the background. For example my foobar playing FLAC is ok.. but when play APE or some ALAC, the cpu will go hulkmode and go full throttle 1.4v 3.9ghz all core.

Sunbeam CCF isn't really better than the Thermalright Ultra Extreme let alone the pure copper version, just that it's a lot cheaper. I would know, I have one at home. No idea if it'll still fit on AM4 clips though, has there been a height change in CPU dies?
Maybe that's why I never encountered this - my CD collection rips + bandcamp buys are all in FLAC. Never looked up CPU usage differences much with foobar audio codecs, maybe I'll check that later.

this is zen2 onward’s issue. newer agesa has solved the aggresive boosting specifically on chrome but still very dependent on what is running at the background. However for certain B450/x470 boards that dont often get support then too bad. For example my foobar playing FLAC is ok.. but when play APE or some ALAC, the cpu will go hulkmode and go full throttle 1.4v 3.9ghz all core.

If you early adopter of the zen2 and used their first few bios revisions...even dragging a folder or resizing a window will cause the cpu to fly up to 80C for 1 sec and then fall back down.

This feedback has been an ongoing complain till this day. Many users chose to ignore already and just accepted it, because this aggressive boosting behavior as quoted by AMD, is what made the cpu responsive.

If you ask me, apart from difficulty hitting 3200mhz ram, the 1000 series and 2000 series ryzens are actually way more stable as compared to 3000 series.

bruh.. that thermalright cooler cannot even cool the nehalem i7 920 suffeciently lol. You can say it performs like a modern 120mm aio at best. It even got replaced by sunbeam CCF in its era.

So 60-70C flutter with avx2 load is damn good. For comparison, a ryzen 3600 hits 75C on a 120mm aio with 20C ambient.

And this early version weak gaming benchmark was uncovered by bilibili that the gpu is running on pcie gen1, and they couldnt do anything on the bios to make it gen4 because the bios is not yet finalized.

I think they made a followup video with a newer bios that works, but the gpu memory bandwith is half or something when compared to 10900k on the same board. Meaning there’s still alot of performance still being locked away.

and even in their gaming benchmark, can see the intel cpu, even the 10900k is actually running 10C cooler in games.

heck.. even a china newtuber is doing a much better job at reviewing than anandtech. At least theyre unbias and honest enough to find out why there’s a performance drop from last gen.

I don't really appreciate the anand review. It's not the REVIEW they used to be like the old days. Started the review saying it's a hot hot chip and bulldozing everything else.
At best it's termed a Garbage Review and not worth even reading it.
Anand and Toms hardware - I don't follow their stuff for a very long time.
They are NOT the CREDIBLE sites they used to be (after change of ownership)
Just look at the Toms layout and the stuff there, it's garbage.

Sounds like a bad software optimization, and many of the windows visual effects are running off Desktop Widow Manager that utilize GPU power, and even webpage rendering also fully utilize your GPU.
it's either cause by bad OS code, GPU driver.

Im on 3700x now, never encounter such issue, my cpu never go to full throttle unless i run ychruncer, prime95 or 7zip benchmark.

If you open task manager and check the Desktop Window Manager, it use your GPU engine actually, since Windows 7 era.