Monday, September 19th 2022

Early Core i9-13900K Review Hints that it Holds up to the "20-40" Claim
An early review of a retail Intel Core i9-13900K "Raptor Lake" 8P+16E processor shows it holding up to the rumored "20-40" claim, the idea that the processor can be up to 20% faster in gaming, and up to 40% faster in productivity, compared to the current i9-12900K. Much of the gaming performance increase is attributed to the higher IPC of the new "Raptor Cove" P-cores, and the much higher boost clocks they run at (up to 5.80 GHz); whereas the multi-threaded performance boost comes from not just the faster P-cores, but a doubling in the E-core count to 16, and improved E-core cache structures, besides higher clock speeds that they run on. For tests that scale across P-cores and E-cores, the i9-13900K behaves like a 24-core/32-thread processor, which is what it is. Among the tests included are CSGO, AIDA64, 7-Zip, WinRAR, Cinebench R15, R20, and R23; and their average, in comparison to the i9-12900K.
Sources:
ECSM_Official (bilibili), VideoCardz
39 Comments on Early Core i9-13900K Review Hints that it Holds up to the "20-40" Claim
The whole "wait it out", "look at these synthetic benchmarks" and "the latest and greatest" is more geared towards enthusiast but it's moot because there in the economic bracket where price doesn't matter.
My previous CPU was a i7 6700K Skylake CPU, I didn't want to upgrade sooner since all CPU's after 6th gen till 11th were basically still 14nm "Skylake-refresh" CPU's (albeit with more cores).
So I waited for 10nm Alder Lake. I was running Skylake since 2016.
And let me tell you that from i7 6700K to i7 12700K is a quite an upgrade and I'm satisfied for now. ;)
i was putting more emphasis on your post than counter arguing it actually.
Is there a reason AMD can't or hasn't done a AM4 5900G pairing a 5600X an 5600G together!!? Wouldn't that even allow for a hybrid mixture of PCIE 3.0/4.0 connectivity!!? Sort of a AM4 5900G big LITTLE?
- Needing a whole new CPU substrate, a significant cost
- Layout difficulties - the APU die is far, far too large to fit within the standard AM4 IOD+2xCCD layout (at ~3x the size of a CCD)
- Lack of scheduler support/tuning for cores of the same architecture but with different amounts of cache
Among others.
Also, how would you get the connectivity from both dice off the package? Do you think AM4 just has a bunch of spare PCIe pins free, that are in motherboards but not being used? And if you're using the PCIe from the APU die as PCIe, where would you find the IF links to connect it to the IOD of the rest of the chip? Those APUs don't have a ton of connectivity, so you'd need to pick - either off-package PCIe, or on-package IF to connect to the other chip. And you'd obivously go with the other as you can't just run two CPUs in a system willy-nilly. And the lack of I/O also means it wouldn't work to use an APU die as a "CCD+IOD" combo with a second CCD paired to it, as you'd have, what, four PCIe lanes left? Something like that.
We are getting to the era where unlocked multipliers don't really matter anymore because the OC headroom is way less than say, 4790K era. Back then the stock frequencies were averaging in the 3 GHz range, while overclocked you could do upwards of 4.5 GHz, that's a 28.6% increase in clock speed from 3.5 (for example). Now? We have CPUs that do 4.6 GHz+ stock, and the oc frequency is maybe 5.1 GHz all core, that's a 10.9% increase in frequency in exchange for paying extra for a motherboard that supports multiplier overclocking, and a cooler that can handle it. Why not just get a fast CPU regardless of unlocked, and pair it with a decent B or H board, and get a cooler? That way you don't have to do any of the tuning work, and it runs almost as fast without any of the headaches (also you can't void your warranty doing it this way).
That is precisely my point. Since the OC headroom is next to 0 on the higher end chips, why not buy the non K version?