Those chips also always had hyperthreading on the P-cores
true but doesn’t Ultra have a shorter pipeline (according to Intel, maybe it’s not true), maybe they are suffering from low priority scheduling going to P instead of E and impacting execution. Not sure but the horse is out the barn on the line. I guess we will know the answer with Ultra II, Ultra Pro, Ultra Man, or whatever intel calls the next generation and see if they do bring back HT.
let’s not forget its security issues
a huge part of CPU improvements are security issues. one of the big ones was speculative execution which is still getting exploited
Sal, check out this interesting article on PC gamer
Titled - Intel’s Arrow Lake chips aren’t winning any awards for gaming performance but I think its new E-cores deserve a gold star
"So, to that end, after I’d finished testing and reviewing the Core Ultra 9 285K, I ran our suite of CPU gaming benchmarks again (with the same test setup), but this time in two configurations: (1) All E-cores disabled, so just eight P-cores, and (2) one P-core and seven E-cores.
This way the CPU would always have eight threads on offer, but in the second configuration, the E-cores would be called upon to do a lot more work than normal…I don’t think we’ll ever see Intel release a desktop CPU that’s almost entirely E-cores but if the next generation of Core Ultra processors can have them running a bit faster, Team Blue could well be back in with a fighting chance of snatching the gaming crown from AMD.
Until then, just stick with Raptor Lake or any of AMD’s AM5 socket chips. Arrow Lake has disappointed an awful lot of people, but those Skymont E-cores are darned impressive to me."
what was the result between the two tests?
looked at the article by googling - i wonder if reviewers like HUB retested the very poorly performing games with e cores disabled if things would improve?
i think homeworld and plague tale are 2 standouts
the PCGamer results showed most benefited from e-cores, but a few didn’t. and from the suite of benchmarks from HUB there were a few standouts that just performed very poorly. They should retest those with different configurations.
i still don’t really buy the architecture just sucking hard on some games. I think it’s gotta be software
1 Like
really depends on the game
CP2077 & BG3 played out as expected, all cores did the best with 1-p & 7e coming in last but holding their own
Homeworld, Metro, TW3 - 1-p & 7e held their own showing almost the same performance as full cores and just 8 - pcores
some reviewers have theorized it has to do with moving the memory controller and increasing memory latency (bad memory benchmarks). Maybe, we won’t know until some former Intel engineer comes out and says “yeah we screwed the pooch on that move” or Intel does a complete redesign in certain areas.
1 Like
Did you just look at the images and not read what I said? First off I was at 4k Max Settings. Second 13900k was running 7200mhz Ram while the 7950x3D was at 6000mhz. When I dropped the 13900k down to the same 6000mhz as the 7950x3D was at the frame rate tanked! Started seeing an average in Valdraken of 65fps. When I bumped the ram speed on the 13900k thats when I started getting the 100+ FPS
So dont think there wasnt a MASSIVE difference between the 13900k and X3D chip. There was! I mean I had to run 7200mhz ram JUST TO COMPETE with it!
It did a few months ago when it was selling at $320 
AMD stopped production on it in prep for the 9800x3D this shortened supply and sky rocketed the price of the 7800x3D
It actually hit that price point in mid Sept as well (hit it in June of this year) that said, $320 is still not a “budget” CPU price point in my opinion. When I think of “budget” gaming CPUs I think of under $200 and usually around $150. If I’m spending over $300 on a CPU I expect top performance (maybe not the best but certainly near the top).
If your game engine is is heavily nested & looped while dumping a ton of rendering on one or two cores (most but especially online) than latency is obviously a major impact. In some gaems HT(also SMT) increases it while others it helps, TPU has often said in their tests HT did not help gaming benchmarks and often made then worse when the CPU had plenty of overhead. My point (and your benchmarks as well) is that the design of the X3D allows information to get from the cache (with you few misses and pipeline issues) to the cores. With the 13900k once you reduced the RAM speed, it made an obvious impact on the CPU latency that was harder to overcome with a longer pipeline and any HT issues.
This is why anyone who ever says “X” amount of core or threads is good enough, not enough, just enough, etc., really it’s the performance of “X” CPU is good enough, not enough, just enough for your needs.
I’m well aware of all of this but the gentleman above claimed I didn’t see much of an uplift between the two CPUs and I was explaining that they’re most definitely was. Just not in the Benchmark images so I want to explain why
1 Like
Out of curiosity, hows the boot time on the 7950x3D?
18sec if you don’t have memory context restore enabled. Instant if you do. I know on some motherboards its longer but my motherboard has one of the fastest boot times available. RoG Strix x670e Gaming-E
1 Like
Interesting
you read stories about two minute boot times (worst case) with MCR on or off, ASUS being the worst of the mobos for boot time (although they are the most popular), dual rank memory having longer boot times.
I talked to someone who reviews mobos and he said he still sees it across all mobos even with BIOS updates. I find it interesting AMD can’t get a handle on it but nice to know it’s not impacting everyone.
On the original bios the boot time was absolutely horrible. New AGESA updates fixed that
Its their mid to lower tier that are absolutely horrible on boot. Seems to be the same on the new x870 platform
But if you check out HardwareunBoxed motherboard review video it shows all the various motherboard boot times. For new x870. The office motherboards also had some of the best ram stability
Honestly since it’s memory training causing the boot time this is probably Ram related. I tried it team group ram kit and my boot time doubled. I am on top of the line G.skill kit that is on the motherboard qvl list. The team group kit was not on the qvl list and like I said it doubled my boot time. Both were 6000mhz CL30 I wound up keeping the G.skill just because I like the look of the aesthetic better
This is more in line with people have told you. This is not the first instance I’ve heard of non-qvl list Ram increasing boot times on AMD
1 Like
my cousin complains of slow boot times on his 7800X3D
but he says he leaves it on all the time anyway so I don’t know how big of an issue that really is.
My intel system boots up in what feels like 10s or less, but I often find myself leaving it on by accident most times and I just don’t really care much to remember to turn it off
agree its ram training repeatedly on startup
2 Likes
In my experience the two contributing factors to this is choice of motherboard and RAM kit
what DOES make me mad though is the monitor flickering upon wake…i think it’s a gsync issue but man it’s a good thing i don’t have epilepsy because the amount of times it has to flicker to wake tf up is jarring