AMD Ryzen 9 5900X vs Intel Core i9-11900K: Rocket Lake and Ryzen 5000 CPU Face Off

[elgato610]

As of today the 5900X is still not available. Of more concern is the now famous, and unsolved, WHEA ERRORs
that seem to plague the entire AMD processor line. No solution and no proof that it is not a chip or design flaw.
Also, many reports that AMD processor life expectancy is 5 years as opposed to Intel's 10.

 

[Tanquen]

Any truth to Intel being faster in VMs with special functions?

 

[Tanquen]

As of today the 5900X is still not available. Of more concern is the now famous, and unsolved, WHEA ERRORs
that seem to plague the entire AMD processor line. No solution and no proof that it is not a chip or design flaw.
Also, many reports that AMD processor life expectancy is 5 years as opposed to Intel's 10.

Nonsense, neve had an AMD CPU fail, tons of other chips made at the same plants. The WHEA ERRORs are folks messing with RAM speeds. Yes AMD suggested a stable 2000 infinity fabric but my 5950x runs fine at 1800. It did take months to get a MB BIOS that stopped dropping USB devices.

 

[Graxxor Vidhelssen]

Since the AMD 5000 series is competing with hen's teeth in the rarity stakes, I find articles like this to be purely theoretical discussions of little applicability right now as they are almost impossible to get hold of at regular price and in significant numbers.
And the situation will get much worse before it gets better, with competition from auto makers for silicon wafer millimetres and the relative lack of investment to keep prices artificially afloat over the last half decade coming back to haunt the big players, it will take until 2024-25 before the current quarter trillion combined investment starts to show increased yield.

Likely the only thing that could possibly help in the meantime before the new fabs come online is a crash in the cryptocurrency markets removing some of the demand... But with players like Elon Musk buffing the crypo markets, a crash could be a few years away.

 

[TerryLaze]

TSMC is spending $100 billion. 😛

Intel is aware that ARM is the future, hence the IDM 2.0 strategy with foundry service on offer.

this article is a good read.

Chip industry battle royal: Arm throws down the gauntlet at Intel’s feet - SiliconANGLE

Chip industry battle royal: Arm throws down the gauntlet at Intel’s feet - SiliconANGLE

siliconangle.com

Intel has been making risc/arm CPUs themselves since the 80ies,look up the i960.

 

[Jim90]

As of today the 5900X is still not available. Of more concern is the now famous, and unsolved, WHEA ERRORs
that seem to plague the entire AMD processor line. No solution and no proof that it is not a chip or design flaw.
Also, many reports that AMD processor life expectancy is 5 years as opposed to Intel's 10.

Intel is clearly getting desperate if they sanction their shill payments to produce this garbage,

 

[TerryLaze]

Intel is clearly getting desperate if they sanction their shill payments to produce this garbage,

AMD is clearly getting desperate if they sanction their shill payments to produce this garbage.
See?!Works both ways!
Your comment isn't any better.

 

[Rodrigodrt]

So.

would it be worth to sidegrade to a 11700k from a 10700k in case I want to take advantage of gen4?

 

[King_V]

As of today the 5900X is still not available. Of more concern is the now famous, and unsolved, WHEA ERRORs
that seem to plague the entire AMD processor line. No solution and no proof that it is not a chip or design flaw.

I would ask for citations, but Tanquen's post already addressed this being people screwing with RAM timings, and NOT an issue that plagues the entire CPU line with "no solution and no proof that it is not a chip or design flaw" as you claim without evidence.

Also, many reports that AMD processor life expectancy is 5 years as opposed to Intel's 10.

Please provide citations.

 

[Jim90]

As of today the 5900X is still not available. Of more concern is the now famous, and unsolved, WHEA ERRORs
that seem to plague the entire AMD processor line. No solution and no proof that it is not a chip or design flaw.
Also, many reports that AMD processor life expectancy is 5 years as opposed to Intel's 10.

Intel is clearly getting desperate if they sanction their shill payments to produce this garbage,

AMD is clearly getting desperate if they sanction their shill payments to produce this garbage.
See?!Works both ways!
Your comment isn't any better.

--> "Also, many reports that AMD processor life expectancy is 5 years as opposed to Intel's 10" --> !!

--> Since Zen2 was released in July 2019 and Zen3 November last year, I would suggest that @TerryLaze's support of @elgato610's post is, shall we say, eminently foolish. Maybe use that time machine again to provide some evidence??
Again, to reiterate, Intel is clearly getting desperate if they sanction their shill payments to produce this garbage.

 

[Kutchek]

I've noticed that most, if not all of the comparisons that have been done between Intel and AMD cpus and Nvidia and AMD gpus are only compared using 1080 or 1440 settings

Is there any chance you can start using 2160 4K benchmarks and game settings considering that there are now many people out there with systems that can comfortably handle 4k gaming (and other applications) now that the new 3000 series gpus have been released (and that you use a 3090 in your test system), especially since there will also be some shiny new hdmi 2.1 monitors released very shortly which are pretty much aimed at 4k gaming?

Thanks 😊

 

[TerryLaze]

--> "Also, many reports that AMD processor life expectancy is 5 years as opposed to Intel's 10" --> !!

--> Since Zen2 was released in July 2019 and Zen3 November last year, I would suggest that @TerryLaze's support of @elgato610's post is, shall we say, eminently foolish. Maybe use that time machine again to provide some evidence??
Again, to reiterate, Intel is clearly getting desperate if they sanction their shill payments to produce this garbage.

You do know that you can push a CPU or a car or a shoe way above normal use to simulate many years worth of normal use in a much smaller time right.
Somebody not knowing basic principles of how the industry tests things is shall we say at least foolish.

He didn't show any proof so his bad but you didn't show any proof either so why do you think that you are better than him?!

Here is a simple example, he makes his point and shows how fast a mobo will start to fail in 30 minutes flat.
At 5:50 he states that caps are rated for 5000 hours at those temps so you can do some simple math depending on how many hours a day the board is going to be used.
(A user will not care if it's the mobo failing they will still think it's the CPU. )

View: https://www.youtube.com/watch?v=vLMd-5yxTAc

 

[sdmf1974]

What does he mean the 11900k was readily available at lauch?! Heck Im still waiting for this chip to launch. Ive been watching since March 30th and havent seen one anywhere aside from 3rd party sellers with 60% positive feedback ratings and priced at over $1000!!!!!!

 

[JarredWaltonGPU]

I've noticed that most, if not all of the comparisons that have been done between Intel and AMD cpus and Nvidia and AMD gpus are only compared using 1080 or 1440 settings

Is there any chance you can start using 2160 4K benchmarks and game settings considering that there are now many people out there with systems that can comfortably handle 4k gaming (and other applications) now that the new 3000 series gpus have been released (and that you use a 3090 in your test system), especially since there will also be some shiny new hdmi 2.1 monitors released very shortly which are pretty much aimed at 4k gaming?

Thanks 😊

I looked at CPU scaling with the RTX 3080: https://www.tomshardware.com/features/nvidia-geforce-rtx-3080-ampere-cpu-scaling-benchmarks
RTX 3090 is a bit faster, but it doesn't move the bottleneck from the GPU at 4K, at all. You can see in my testing that at 4K ultra, there are many cases where the Core i9-9900K came out on top -- probably due to maturity of the platform. But by "on top" I mean differences of 1% or so, nothing to write home about. I didn't specifically test the 5900X in that article, but it's unlikely to radically alter the positioning at 4K.

 

[everettfsargent]

You do know that you can push a CPU or a car or a shoe way above normal use to simulate many years worth of normal use in a much smaller time right.
Somebody not knowing basic principles of how the industry tests things is shall we say at least foolish.

He didn't show any proof so his bad but you didn't show any proof either so why do you think that you are better than him?!

Here is a simple example, he makes his point and shows how fast a mobo will start to fail in 30 minutes flat.
At 5:50 he states that caps are rated for 5000 hours at those temps so you can do some simple math depending on how many hours a day the board is going to be used.
(A user will not care if it's the mobo failing they will still think it's the CPU. )

View: https://www.youtube.com/watch?v=vLMd-5yxTAc

Look up "scale effects" "structural mechanics" and the Easter Bunny is real. Also a YouTube video is NOT how the so-called industry tests their equipment (ISO and ASTM are examples of standards testing). The Easter Bunny is a point of logic, claims made must always be supported by the claimant and not the other party (pushing them to prove a never ending negative which is not possible while it only takes one objective observation to prove a positive). ...
https://en.wikipedia.org/wiki/Russell's_teapot

Russell's teapot - Wikipedia

en.wikipedia.org

Suggesting a linear relationship when clearly the YouTube clip shows a highly non-linear relationship is not the same as cars, shoes, motherboards, CPU's or whatever. One develops a relationship between time and spatial scales using something called similitude and dimensional analysis (scale effects are always an issue in accelerated testing or build an aeroplane ten kilometers long given known specific strength, specific modulus and gravity). Push anything close to or past its structural limits is bound to fail the device much more quickly then when adequate safety factors are included in its design up front. Otherwise known as standard practices.

 

[Kutchek]

I looked at CPU scaling with the RTX 3080: https://www.tomshardware.com/features/nvidia-geforce-rtx-3080-ampere-cpu-scaling-benchmarks
RTX 3090 is a bit faster, but it doesn't move the bottleneck from the GPU at 4K, at all. You can see in my testing that at 4K ultra, there are many cases where the Core i9-9900K came out on top -- probably due to maturity of the platform. But by "on top" I mean differences of 1% or so, nothing to write home about. I didn't specifically test the 5900X in that article, but it's unlikely to radically alter the positioning at 4K.

Thanks for the reply 😊 That's an interesting read, and also quite a surprise that even the old 4770k was not sooo far behind the current 10900k and previous 9900k at 4k ultra settings 😳

I do have another question, if I may?

With the Nvidia gpus, it doesn't look like you are using them to their full potential, by disabling the RT and DLSS, which seems like they are being put deliberately at a disadvantage when they have a capabilities of working so much better than the test suggest... If you're doing a true comparison, surely you should use the hardware with its optimum settings and not disable functions because their competitors don't have these extra bits... I doubt many gamers would NOT use these extras because they feel that it would be unfair 🤷‍♂️
People want the best performance that they can afford, with all the bells and whistles enabled, and would likely want to see comparisons with them enabled too

I know for sure that I want that, at least

Thanks again

Jason