Not bad performance, but almost 100W more for the same multi score of 5950X?
I know many people are biased, but who cares who has more FPS at FHD when talking about hundreds?
I see a power hungry product that barely beats the competition in performance and tries to capitalize on price.
I would not consider upgrading from my 2600 to this, I'd be more prone to buy a 5600/G or something and wait out for affordable DDR5.
Nothing to see here folks, I'm very surprised that the P cores just match zen3 in IPC. Means that the 3d vcache refresh is probably all that amd need to do to compete/ exceed the performance of this line of products.
Not bad performance, but almost 100W more for the same multi score of 5950X?
I know many people are biased, but who cares who has more FPS at FHD when talking about hundreds?
I see a power hungry product that barely beats the competition in performance and tries to capitalize on price.
I would not consider upgrading from my 2600 to this, I'd be more prone to buy a 5600/G or something and wait out for affordable DDR5.
187w for 29451MT CB23 on my 5950x 4.75Ghz@1.35v (using CTR 2.1 rc5) but yeah I have a higher score at a lower temp too (68°C but I have a 3x360mm rad setup lol)
from the 8 more threads for sure that said if the intel had 8 more threads it would be above for sure
it's not that power hungry don't be fooled, AMD frequencies are still overall low compared to intel and that's where the watts come from
Ryzen from day one "cheats" with fake high clocks that last 1us but aren't really used when Intel can have a flat 5ghz line forever with OC (I have a 9900k on my left and a 5950x on my right running as I type this)
those fps do matter because there are more games than the "reviewer friendly" ones that all influencers and websites use and those don't run at 300fps more like 60-100 and then those 10fps more matter a lot especially when you've paid like 3000€ of hardware my ultra modded skyrim struggles on everything but clearly is smoother on intel, not because of the graphics but the amount of background scripts running that really don't like slow cpus/memory
very interesting...might buy a ddr4 board if it's that bad, you can't even buy 6200 here only 4800 or 5200
hmm... maybe not so fast
I would expect DDR5 to pull away, once teething probs are fixed and especially once devs start utilizing newly available bandwidth
hmm... maybe not so fast
I would expect DDR5 to pull away, once teething probs are fixed and especially once devs start utilizing newly available bandwidth
True that means better skip this generation and sit on ddr4 for 2 years more.
Seeing as that's likely a game benchmark, that doesn't prove much seeing as the 12900K isn't that much better in games overall.
Show me a graph of DDR5 vs DDR4 where the 12900K leads in the charts by a significant margin (like in the Jetstream, Kraken, or 7-Zip benchmarks) - that's a way to prove whether DDR5 is doing most of the lifting or not.
hmm... maybe not so fast
I would expect DDR5 to pull away, once teething probs are fixed and especially once devs start utilizing newly available bandwidth
I wouldn't be so sure, the jedec timings of ddr5 are alot looser than any memory type going back to DDR, , there has been a "cap" of around 15ns for the past 20 years, this time around its exceeding 18ns... I suspect that this trade-off is going to be compensated by the very large l3 or l4 caches on future cpus.
Most consumer applications aren't memory bound atm, as evidenced by alderlake's lack luster performance, I would've thought that handbrake would get a big speed up since it tends to scale with memory , but nope...
@user1
even so. are you really willing to invest in old(er) RAM platform. DDR4 surely wont get _better in time, and it could easily get worse compared to DDR5
Meet Intel's new flagship CPU, the Core i9 12900K. It is based on the Alder Lake architecture and is reviewed here. This time around, Intel was back at the drawing board, creating a completely new ar...
Review: Core i9 12900K processor
I will be a bit harsh and say that these results wont reflect the ones that people are actually going to get, due to the ddr5 memory used being so slow... no one buying a 12900k is gonna buy poo 5200 cl40 modules, unless it's a prebuilt... so you can probably add 10-15% performance to these results, just from using proper modules.
@user1
even so. are you really willing to invest in old(er) RAM platform. DDR4 surely wont get _better in time, and it could easily get worse compared to DDR5
not suggesting that , but what I am suggesting is that on an early platform like alderlake, there will probably be near 0 benefit to using ddr5 (for its lifespan), especially if you use one of those fancy new ddr4 5066 kits. I would probably wait for another socket change before considering it, especially since higher density memory is not available yet.
Looks about right to me. Power consumption aside, we have real competition again. Too bad the overall shortages will mess up prices regardless (also of course the new DDR5). The CPU price itself is pretty good, but those motherboards... daaaayum!
To extrapolate further If you already own fast ddr4, then it makes no sense to buy ddr5.
If you don't own ddr4, it still doesn't make much sense because you will likely lose performance with ddr5 compared to a fast kit of ddr4, and the ddr4 is probably cheaper. the only way it would make sense is if you need a ton of ram, like >256gb, in which case , the ddr5 is going to offer higher densities in the future, but are you going to need that much on a platform like alderlake? By the time ddr5 is able to be utilized properly(ie cpus with gigantic caches and cores with high enough throughput), you'll need a new motherboard and that ddr5 you bought is going to be extreme trash, since its barely offering more bandwidth than ddr4 can.
I don't see any issues with availability at MSRP - yet. The same couldn't have been said for Zen3, which was gone minutes of release. Motherboards and memory are a different story, though. Both are pricey.
I will be a bit harsh and say that these results wont reflect the ones that people are actually going to get, due to the ddr5 memory used being so slow... no one buying a 12900k is gonna buy poo 5200 cl40 modules, unless it's a prebuilt... so you can probably add 10-15% performance to these results, just from using proper modules.
Other people already tested super expensive DDR 6200 memory and the difference with DDR 4 is marginal in most cases. If you gonna spend all this money you might as well wait for the next TR cpus imo. It doesn't make much sense to pay a premium for super expensive DDR 5 memory right now.
FINALLY!!!! LET TO WARS COMMENCE!!! :D