With the release of Alder Lake less than a week away and the "Lakes" thread having turned into a nightmare to navigate I thought it might be a good time to start a discussion thread solely for Alder Lake.
So we're basically arguing about less than 1% of desktop users?99% of desktop users have absolutely no need for a 12900k or 5950x.
And yet the 12900k are sold out, to wait for it..... desktop users! So, maybe there are those who will take every ounce of performance they can get? This is the same argument people make saying you can pair a 5950x with a B550 board. So who buys the more expensive x570 boards? The 5600x users? There's a market for these things and they're typically called enthusiasts. They want every fps they can get, they will push the hardware as far as it will go, and they will spend all they can to acquire the best. So what's your point again?If all you do is lightly threaded stuff, why in the world would you buy a 12900k or a 5950x when there are much cheaper options that will do just as well?
Not true. See above. The 12900k plays in the 5950x's territory, not the other way around. Even the 12600k is faster in most gaming, and other bursty workloads. So don't try to get things twisted. The 12900k dominates EVERYTHING that's not a 5950x in select parallel workloads.if the argument is that most workloads don't need that many cores, then that's the end of the discussion right there because there's no need to talk about either CPU anymore and it's a whole different discussion around SKUs with less cores.
First, you qualify that statement with "select paralell workloads". Then you assume that is at PL2 of 241 watts. If not 241 watts, it LOOSES most of those workloads. But you keep areguing about desktop users, who will not generally buy a 5950x or a 12900k.Not true. See above. The 12900k plays in the 5950x's territory, not the other way around. Even the 12600k is faster in most gaming, and other bursty workloads. So don't try to get things twisted. The 12900k dominates EVERYTHING that's not a 5950x in select parallel workloads.
So we're basically arguing about less than 1% of desktop users?
/Thread
And yet the 12900k are sold out, to wait for it..... desktop users!
So, maybe there are those who will take every ounce of performance they can get? This is the same argument people make saying you can pair a 5950x with a B550 board. So who buys the more expensive x570 boards? The 5600x users? There's a market for these things and they're typically called enthusiasts. They want every fps they can get, they will push the hardware as far as it will go, and they will spend all they can to acquire the best. So what's your point again?
Not true. See above. The 12900k plays in the 5950x's territory, not the other way around. Even the 12600k is faster in most gaming, and other bursty workloads. So don't try to get things twisted. The 12900k dominates EVERYTHING that's not a 5950x in select parallel workloads.
How not? It dominates in everything and even surpasses the 5950x in certain parallel workloads, yet it's not a good processor to get? It comes in at a close 2nd in multithreaded workloads on desktop. in single and lightly threaded loads on the desktop, where's the 5950x? The 12900k rendered the 5900x obsolete. Notice no one even talks about it? Greater multithreaded performance than the 5900x and king in everything else on desktop. What's not to like? The 5950x loses to an i5 in most codes that are not embarrassingly parallel. Is there a more niche processor on the desktop right now? I've always called the 5950x an HEDT processor, and Alder Lake has exposed it for what it is. It was already losing to RKL in some single thread tests, and now ADL i5s as well. That's not a great place for a flagship processor to be.I just don't see a good argument for getting a 12900k unless you don't care much about highly threaded workloads and just want the best for lower threaded workloads and don't care that you are paying hundreds more for a few percentage points faster than a 12700k at best.
How not? It dominates in everything and even surpasses the 5950x in certain parallel workloads, yet it's not a good processor to get? It comes in at a close 2nd in multithreaded workloads on desktop. in single and lightly threaded loads on the desktop, where's the 5950x? The 12900k rendered the 5900x obsolete. Notice no one even talks about it? Greater multithreaded performance than the 5900x and king in everything else on desktop. What's not to like? The 5950x loses to an i5 in most codes that are not embarrassingly parallel. Is there a more niche processor on the desktop right now? I've always called the 5950x an HEDT processor, and Alder Lake has exposed it for what it is. It was already losing to RKL in some single thread tests, and now ADL i5s as well. That's not a great place for a flagship processor to be.
Comparing pre-production leaks for Alder Lake used to market a Lenovo desktop's power delivery + cooling (I.E Lenovo were well incentivized to play up the results), vs Computerbase results for Zen 3, when Computerbase also tested Alder Lake at various power limits.. Yikes.So I hate to link every review I saw that says the 5950x does better than the 12900k at 125 watts in highly threaded workloads ? I don't have the time. If you don't read enough, thats your problem. Again, see Hitmans post above, he knows. A lot of people know, except you.
Edit: Only one application, but it is highly threaded. Cinebench r20. The 5950x beats it at 142 vs 241 for Alderlake. The 5950x beats the 125 watt 12900k even at 88 watt. (ECO) I really don't want to look up every application using the same test. But everyone knows the A12900k is highly crippled at 125 watt vs 241 watt. "Up to 36%" is many places.
OK, I don't get it.... A stock 5950x @142 watts beats a 12900k@241 watts, 99 to 90. @ 88watts, the 5950x equals the 125 watt 12900k. So it wins both scenarios (if you consider wattage), you just proved my point. What am I missing ?Comparing pre-production leaks for Alder Lake used to market a Lenovo desktop's power delivery + cooling (I.E Lenovo were well incentivized to play up the results), vs Computerbase results for Zen 3, when Computerbase also tested Alder Lake at various power limits.. Yikes.
View attachment 53092
Even in Cinebench R20 12900k @125w PL2 is ~15% slower than 250w and that's the biggest difference tested there.
OK, I don't get it.... A stock 5950x @142 watts beats a 12900k@241 watts, 99 to 90. @ 88watts, the 5950x equals the 125 watt 12900k. So it wins both scenarios (if you consider wattage), you just proved my point. What am I missing ?
I really don't want to look up every application using the same test. But everyone knows the A12900k is highly crippled at 125 watt vs 241 watt. "Up to 36%" is many places.
I googled for an hour, and only found the one reference to MAX 36% diff. I will defer to you, now that you have this. But it also proves my point to many other posters here. In production or productivity apps, the 5950s wins at stock even with the 12900k at 241 watts. At 125 watt, it ties the 5950x at 88 watt, so either way, the 12900k looses if you consider wattage.I take issue with you saying this:
There's been plenty of reviews testing Alder Lake at various power limits and no benchmark even comes close to a 36% delta between 125 vs 241w. ~15% delta is closer to peak with 10% as average.
You imply running a 12900k@125w makes it pointless when it is still a strong showing in MT even at those power levels.
So the 5900x is even less so yet it was probably the most valuable of AMD's offerings last round. This is how the narrative keeps changing on this forum with every Intel win. Answer this question: How far is the 12900k from 5950x in multithreaded workloads? You make it sound like it's not in single digits. It loses some and wins some. I'm surprised you could be typing this with a straight face. What was your verdict of the 5900x, which is actually the true competitor of the 12900k based on pricing?Again, the 12700k does everything the 12900k does at hundreds of dollars less and significantly lower power when fully loaded. If all you really want it for is gaming, well, a 12600k is realistically just as good for even less. If you want that extra, literally, few percent higher scores on a benchmark, get the 12900k, that's its only market. If you actually want to do parallel production work, get the 5950x. No one's saying it doesn't perform, just that its sandwiched between two better options and which of the two you should go with depends on what you want to use it for.
See the above benchmarks. In productivity, even at 241 watt, it looses by 10%, and even more @125 watt, it gets beaten by the 142 watt 5900x with 4 less cores, and at 88 watt ECO ties the 5950x. Where do you get that it wins some in productivity ? and by what specific % ? As you see it looses quite handily on average.So the 5900x is even less so yet it was probably the most valuable of AMD's offerings last round. This is how the narrative keeps changing on this forum with every Intel win. Answer this question: How far is the 12900k from 5950x in multithreaded workloads? You make it sound like it's not in single digits. It loses some and wins some. I'm surprised you could be typing this with a straight face. What was your verdict of the 5900x, which is actually the true competitor of the 12900k based on pricing?
That's one set of reviews..... Let's try something more real world...See the above benchmarks. In productivity, even at 241 watt, it looses by 10%, and even more @125 watt, it gets beaten by the 142 watt 5900x with 4 less cores, and at 88 watt ECO ties the 5950x. Where do you get that it wins some in productivity ? and by what specific % ? As you see it looses quite handily on average.
Edit: I do see that Premiere and Photoshop are missing from the above, and those may be won by the 12900k, but by how much ?
Compared to the AMD Ryzen 5000 series processors, the Intel 12th Gen CPUs hold a strong lead overall, although the exact amount varies based on the specific application. As a complete average across all our tests, the Core i9 12900K scored about 8% faster than the Ryzen 5900X with DDR4 memory, or roughly 17% faster with DDR5. Intel actually takes an even further lead at the i7 and i5 levels, with the Core i7 12700K beating the Ryzen 5800X by 16% on average, and the Core i5 12600K beating the Ryzen 5600X by a large 26%. And based on what we saw with the i9 12900K, we would expect those numbers to increase by roughly another 10% on average if you were to utilize DDR5 memory.