nizzen:

And in gaming, 12900k has 40-50% more fps per watt than 5950x (with fast memory) . Pretty funny it's almost the other way around in rendering. 🙂

I am skeptical of that number that is not what I have seen in 30 AMD vs Intel videos on youtube, but let us just say you are telling the truth. Gaming 5950x 70W vs 12900k 40W power used. Rendering 5950x 150W vs 12900k 244W power used. It is up to the user to judge if 30W difference in gaming or 90W difference in rendering is most important.

TLD LARS:

I am skeptical of that number that is not what I have seen in 30 AMD vs Intel videos on youtube, but let us just say you are telling the truth. Gaming 5950x 70W vs 12900k 40W power used. Rendering 5950x 150W vs 12900k 244W power used. It is up to the user to judge if 30W difference in gaming or 90W difference in rendering is most important.

Der8auer tested it. Look at youtube 😉 He tested 6400xmp, and I'm running 7000c30 max tweaked...

Freeman:

General Hardware discussion is full of Ryzen. And will be. We will all see new posts of AM4 :"hey guys, what do I dooo? What's wrong with my PC?" and we will see AM5 with most likely new issues. Looking at how AMD approach release of new CPU it's to push sales.

The amount of issues I had at Ryzen launch has me seriously considering going back to Intel. I still can't turn on global C states without causing instability (resulting in overall higher power usage at idle). Nevermind USB issues. TPM stuttering. Some of that may be down to my Gigabyte board, which is a vendor I would avoid for the next round as well. But the entire Ryzen platform certainly wasn't ready when 5000 released. And is still showing signs of that today.

winning.exe:

Comments like this make me wonder 😱. My PC is used mainly for highly threaded code compilation, CAD, and rendering, with only occasional gaming. This is entirely throughput based and scales nearly linearly up to 128 threads (in my testing). If I had to buy a consumer processor, I would benefit greatly from a huge number of E-cores. Is this not a “real life” scenario 😀?

Yes, what you describe is a real-life scenario. You're not really countering my point but making it. This is what I mean so yes I want to see real use cases just like yours. I'm not some fan boy that cares who wins I just want to know in a realistic scenario how things shape up as these types of benches usually don't correlate to real use cases all that well especially today when Intel has performance cores and small cores it complicates testing.

certainly anyone needing to do productivity work should wait two or three months for the teething pains of both Raphael and for Rocket Lake as both are showing new uArch, new features, new chipsets, and as of yet no optimized drivers. however some of us like the agony and the ecstasy of launching new platforms, none more so than our very own HH. i use both AMD and Intel on near daily basis and i can easily say with zero controversy or fanboy-ism that both camps have their issues. imho, anyone who does do productivity with huge files must look at the performance levels w/ four DIMMs - period and i'm not goring any oxen or breaking new ground by saying so. as far as the leaks go, there are definite issues with both camps at this all-too-early date.

Freeman:

That thing can be 10Ghz and it would not matter if it was stuttering and causing restarts on every OS. General Hardware discussion is full of Ryzen. And will be. We will all see new posts of AM4 :"hey guys, what do I dooo? What's wrong with my PC?" and we will see AM5 with most likely new issues. Looking at how AMD approach release of new CPU it's to push sales. I dare you: Say one word about driver and software support! Ever heard of : " "Welcome to AMD's User Experience Program"? The one thing every user avoids? ... It will be interesting CPU and one to be careful to seriously look at.

funny i never had an issue w/fTPM stutter as i use a dedicated TPM also... what are you on about re: sales? every company want you to buy new. Intel is far worse on this point as they've just sold AL and have to convince the same people that RL is better (it is), but is doing the exact same thing as AMD so what is the point of your comment? and Intel is incredibly weak on leveraging it's vast databases leaving the end user to parse arcane update and reference points with zero interface that's worth calling an interface

tunejunky:

funny i never had an issue w/fTPM stutter as i use a dedicated TPM also... what are you on about re: sales? every company want you to buy new. Intel is far worse on this point as they've just sold AL and have to convince the same people that RL is better (it is), but is doing the exact same thing as AMD so what is the point of your comment? and Intel is incredibly weak on leveraging it's vast databases leaving the end user to parse arcane update and reference points with zero interface that's worth calling an interface

Intel marketing strategy is different and driver support is better. Intel series is called 11gen, 12gen, 13gen. Not a 5 letter ''fancy word'' with percentage numbers on top. Point of post is at the end of my post - Interesting CPU, careful to look at. Congrats on not having any issues. The guy before who replied to my post did have some.

shady28:

Dude, you realize that score is like a median, where like 70% of the people submitting to Geekbench are using some garbage $80 DDR4-3200 C16 or C18 memory right. And you also know, the test rigs AMD used had DDR5-6000 C30 on $500 motherboards. This is technically a significantly OC RAM on *both* systems, since AMD is only rated for JEDEC 5200 and Intel JEDEC 4800, and if we want to nit-pick according to AMD with the DIMM config they used they are only rated for DDR5-3600 (yes go look). Plus it looks like they were running AMD specific sub-timings on the Intel rigs, which is lame. If you have the type of setup AMD used in their release on a 12900K, $500 motherboard $350+ RAM and 360mm AIO, you can easily knock out 2150+ single thread GB scores. There are actually 12900KS that have over 2500 on ST geekbench, and plenty of 12900K with over 2300. This was just AMD doing a marketing promo of their new chips, stop acting like it was a review.

not AMD's doing they almost never have an ES in the wild unlike Intel who has double the need for them

shady28:

As you spend the next year doing AGESA updates on your new Zen 4 while the guy next door is in the game with his Intel rig, perhaps you will understand why Intel samples so many ES.

i know exactly why Intel fabs so many ES and it has little to do with timings. and you act quite certain that RL will not have issues, but it will (as will AMD). Intel unwittingly (unexpected consequence) has created an ES ecosystem that marketing has taken advantage of the last ten years. after all it keeps Intel in the news and it follows the dictum of P.T. Barnum (the world's first marketing genius). but the reason behind that is quite simple and it used to be true for AMD as well - the fabs require ES for tweaking as much as the uArch does. the real issue for Intel is lax security from Chinese/Malaysia suppliers/AIB vendors. the fact of the matter is very few of us should be seeing ES period. but social media and the need for "likes", plus "content creators" on YT have given this life. just don't go drawing conclusions on any until you know which sampling from when and under what drivers - which you won't so just leave them alone unless you can afford to burn black market money.

shady28:

You don't seem to comprehend my meaning. For one Raptor Lake is coming out on a mature platform. From what I've seen, there isn't even much difference between Z790 and Z690. I seriously doubt there will be many issues with Raptor Lake. Putting that aside, Alder Lake did not have as many issues as Zen 3 did. This is despite Alder Lake being the first to introduce DDR5 as well as PCIe 5.0, along with doubling the number of DMI lanes to the chipset vs Z590. Zen 3 on the other hand had repeat issues with USB devices that went on for ~6 to 9 months after release. These were severe to those affected, with mouse/kb and storage device disconnects, static on sound devices, and in some cases instability in general. And that was with no notable change outside of CPU, on the same socket they had been on for like 5 years. With that back drop, I would expect a more major change like Zen 4 to be a significant bug fest for at least 6 months. Raptor Lake may get some tweaks but I imagine it'll be stable with no major issues out of the box.

au contraire mon frere AL & RL both have serious RAM issues (which doesn't affect gamers much) for power & productivity when having 4 DIMMs is necessary to work at hand... which is actually a larger market (but more static) than gaming. there is as of yet no mobo capable (without some serious tweaking) of running 4 DIMMs at normal speeds with any AL/RL cpu which is exactly why those sales didn't go to Intel and why those sales went to AMD. we can get into the benefits of "moar" cores & what types but don't get me wrong, i love what Intel is doing now as opposed to three years ago. and there have been other issues as well but as i was mentioning earlier, those are typical with uArch and /or node change for both AMD & Intel.

nevcairiel:

The amount of issues I had at Ryzen launch has me seriously considering going back to Intel. I still can't turn on global C states without causing instability (resulting in overall higher power usage at idle). Nevermind USB issues. TPM stuttering. Some of that may be down to my Gigabyte board, which is a vendor I would avoid for the next round as well. But the entire Ryzen platform certainly wasn't ready when 5000 released. And is still showing signs of that today.

If its not working with stock C-states then your motherboard almost certainly broken, I would see if its covered under warranty.

shady28:

That's garbage. You first talk about a big chunk of the market that needs > 32GB in more than 2 DIMMS, then obliquely reference to enthusiasts trying to OC (those are the ones that can't get 4 to 'run' at XMP). If you run 4 DIMMS, even DDR4 is difficult to OC. AL is only rated to run at 4800. Virtually all DDR5 memory will run at JEDEC 4800 with 4 DIMMs as they are rated for, and most will run OC'd to 5200 or so with 4 DIMMS. With 2, we are seeing SK Hynix almost universally go to 6400+. The new A die SK Hynix is going over 7000 for I'd say - most on a midrange motherboard or higher. The "Complaints" you refer to are people trying to get 6Ghz DDR5 to run with 4 DIMMS, ie overclockers. That is the small part of the market you talk of. I'd bet that Zen 4 will be far, far less capable in that regard. A system that can run 'in spec' is not the same as your USB devices dropping off like Zen 3 had. Not even close to comparable.

FYI many alderlake motherboards have issues running 4800mt/s with more than 2 dimms, this isn't news either. intel flatout doesn't support 4800mt/s with 4 dimms officially on alderlake, it only supports 4000mt/s with 4 single rank dimms, and 3600mt/s with 4 dual rank dimms (yes that slow , its not a typo). Technically 4800mt/s isn't even supported with 2 dimms if the board has 4 slots. Its far from ideal, If you need lots of ram, with ddr5 you aren't running 4 jedec 4800mt/s dimms out of the box with alderlake, the complaints are warranted. https://www.tomshardware.com/features/intel-alder-lake-ram-guide-ddr4-ddr5