This is WILD! 🔥 9800X3D OC vs 14900K OC vs 285K OC in 40 Games!

[@DannyzReviews]

Let's end off 2024 with a Bang! 💥💥💥 A few notes: -9800X3D FCLK 2167MHz - Cyberpunk data for Arrow Lake was from before the patch addressing performance issues -"32GB for AMD vs 48GB for Intel", I've made a video recently on the channel showcasing performance differences between the 2 configs, and the result was basically 0 difference. -Why no 8000MT/s for Ryzen? Because you have to run the the mem controller at half speed which can actually hurt gaming performance. Outside of synthetics or some specific production workloads you're better off sticking with 1:1 ratio for Ryzen.

[@BrawlXKing]

You'll get better performance and better low 1% if HAGS is disabled. Did this to Fortnite and now I have 300 plus better 1% lows and .1% This is something that needs to be fixed in games.

[@shaggymanson5375]

is this with the updates for the 285k?

[@TOTALpopedom]

@@BrawlXKing HAGS = ??

[@BrawlXKing]

@@TOTALpopedom Hardware accelerated GPU scheduling. HAGS. Also disable core isolation.

[@n8spL8]

FCLK should be 2067 for 1:1 w/ 6200 (2x3100). 2133 is for 6400, 2200 is 6600...so on. The FCLK is a 3:2 ratio on speed to RAM at 3100x2 = 6200. 3 x FCLK should = RAM speed = MCLK x 2 ................ 3x 2067 = 6200 = 2 x 3100. I don't believe you are running DDR 6500, no? Reddit posts seem to get this wrong, just try it on a calculator. It will lead to latency stutters to stray from 1:1 even if you do achieve some slightly higher fps w/ FCLK ocd over 1:1. Intel doesn't suffer from this, OC everything as far as you can. AMD, no need to OC past 1:1, it will in fact lead to latency. Thanks for the video.

[@arc00ta]

That poor 285K, getting abused like this in front of everyone. How embarrassing.

[@JahonCross]

People will cope that it's not a gaming cpu...yet Intel bost about gaming everytime...that is embarressing 😅

[@shambler7359]

Only if it did cost poor too 😂

[@JDD_Tech_MODS]

Waste of sand :(

[@backslide311]

Should have tested it with the proper memory. You need to run CU DIMMS

[@JDD_Tech_MODS]

@@backslide311 cu-dimms offer nothing performance wise for the Core Ultras. The slow memory latency on the IMC still bottlenecks the chips. 70-80 ns latency is pretty bad.

[@bjarnis]

The 285K is like going back several generations, why even sell it..

[@03chrisv]

Because it's not a gaming CPU. It has decent gains over a 14900k in many productivity benchmarks and programs. This whole video is pointless because it's comparing a CPU designed specifically for gaming versus high core count CPUs meant for productivity. A more fair comparison would have been a 9950x vs 14900k/285k.

[@Kyanzes]

@@03chrisv The latest flagship still shouldn't be much worse than last generations's flagship.

[@bjarnis]

​@@03chrisvthe issue is that Intel advertised it as a good gaming CPU. I understand how PC hardware works, but avarage Joe doesn't and will be disappointed if he buy it for "everything".

[@Chaos_God_of_Fate]

I think they knew that it was going to be slower and they knew they were going to eat it on this first new chip design. I suspect they have a long-term plan and they had to switch architectures to keep up. While this first generation is not very good I suspect that as the new architecture matures it'll much get better. Ya though, AMD is owning it hard these Days and I like to see it, I'm always for the underdog and AMD was for a long time. Intel could have done so much better if they hadn't stagnated with their 4-cores for a decade allowing AMD to catch up, which illustrates why this kind of competition is good for everyone!

[@chucklesdeclown8819]

​@@Chaos_God_of_Fateif they knew it was gonna be slower, why do they keep saying "it was not performing as expected" suggesting either performance was left on the table or they were expecting way faster results meaning their tests showed better or at least on par. Nah, it's clear to me that they done goofed a generation without checking related parties on whether performance was left on the table or not and squashing that before release. Basically if that were true, why are they trying to fix the performance?

[@Floodsye]

I upgraded from a i7-12700K and have only had the 9800X3D for a day and am already pretty blown away. I'm not even that concerned with the raw FPS it can get at 1080p, but rather that it absolutely crushes CPU bottlenecks in games that made the 12700K (and...I guess 285K, yikes) buckle at my usual 1440p. Nearly every game, at least that I frequent, has those areas like WoW's cities/raids, big cities in open world games like Horizon's Meridian, strategy games, Baldur's Gate, etc. that impact gameplay smoothness paired with 4070 Ti and, at least for me, rips me out of the immersion. This thing though? CPU bottlenecked trouble spots are gone entirely, even RTS/simulation games seem to be nothing to this beast. Seriously, impressive performance from AMD. Last time I had an AMD cpu was the FX-9590, the hottest CPU literally. Thankfully, 9800X3D's running nice and cool with 240mm liquid cooler.

[@Pannencop]

nice. im waiting for my 9800X3D...have the 12700k and i see that my 4070ti is to good for the 12700k... CPU Bottlenecks in most of the games...so i will see how good the AMD is :)

[@starlord632]

Would like to see the 0.1% lows on the graph, especially if you're going to discuss them during each game comparison.

[@The_Chad_]

I was thinking the same thing. I'd also like to see a percent difference between avg and 1%

[@GewelReal]

9800X3D would win even harder then

[@The_Chad_]

@@GewelReal Whether it would or wouldn't, it would be nice to see it data.

[@craigjoe8691]

Already been addressed. Stop reaching for straws.

[@wildwilco]

More interesting is performance per watt...

[@69ryxn]

Would have loved to see the 0.1% data in all the bar graphs.

[@wololo10]

You are losing your smoothness because of gpu power monitoring. That only affects amd, it's not a problem of the 9800x3d. You have to disable gpu power, and gpu power % on any monitoring software.

[@karlos1060]

Maybe you should watch the video! He came with that sollution and notice a couple of video's before this when he found out. He did not use that now with this video! He is even showing with some benchmark tools when the 8 core cpu suffers the results with and without him using the capture card.

[@n8spL8]

not to mention his FCLK should be 2067 for ddr 6200. straying from 1:1 on ryzen will lead to latency stutters.

[@enviable01]

9800x3d costs $479 14900k costs $589, the 14 and 13 gen had issues burning themselves out but it was "fixed", the 9800x3d outperforms the i9 on majority of titles, give me it anyday

[@AndyPro05]

9800x3d costs 1000$ right now in Moldova.

[@typeins]

@@AndyPro05 yeah moldova is not a concern for amd.. no population and nearly no one can afford it anyway.. just buy it from germany with DHL

[@ez6678]

Here in Europe it costs 900$.

[@johnscaramis2515]

@@ez6678 Just checked on Mindfactory, 569€. I assume you checked for scalper prices on Ebay?

[@AugmentedGravity]

Why are people siting scalper prizes

[@mcgmgc]

What are the numbers on your curve optimizer?

[@marcusm5127]

I am on my third i9 (14900KS) had the 13900KS and 14900K before they too broke. I am waiting for 9950X3D so I can get better gaming and non gaming perfomance. Plan to buy a 12900 or something and put in the motherboard to get a decent server out of it. I am so tired of crashes and issues. Right now I limit power, current and have the base P core multiplier set to 55x.

[@mesi0r]

im going to build new pc soon, intel's prices are very tempting atm, you say there's a lot of trouble with them? still?

[@marcusm5127]

@mesi0r Intel says it's solved with new bios releases but I don't feel like getting another i9 after this. Also most of the reviews are made with the old power unlimited bios settings so it's lower performance now. My degraded i9 can't be saved with new bios but a new one would probably be fine.

[@user-xyz1000]

😂 nobody will accept, that 14900k is a reaktor and nothing Microcode can fix the one Circiut that dont like high voltage and Temperatures! when you dont have a Costum Loop with two or More 360er Radiators, you have to Clock him down to 5,6 or 5,7 Ghz for Handle the Temp! othervise he jump so fast over 90 degrees you dont can show it! 😂 my 14900KS was running good castrated with the lequid freezer 3 to 5,7 ghz and -0,1 offset undervolting. NOW: he runs to 6GHZ, -0,6 offset at all P-cores stabl. + with Costum Loop ✌️ The Last Tipp: Set a Maximum Temp (overtemptrothling) for save him to burn out at ca. 86° When he often do Benchmarks or he came often to 90° - 100° its the same as selfdestruktion! chears all 14900/KS Burners 🤘 stay safe!

[@mesi0r]

@@marcusm5127 well Intel claimed there was no problem until they were pushed against the wall. On AMD side situation doesnt look great as well. I have no idea what to use for my PC right now.

[@marcusm5127]

@@mesi0r I would certainly go with AMD right now. Amd got flack for not improving enough while Intel has critical faliures and their new arrow lake line is slower than 14th gen. I am waiting for 9950x3d so I can get better gaming and multitasking performance than my 14900KS. For general computing the 9900X looks great.

[@thegamergg200O]

Thank you so much for providing deeply detailed tests, that's what most benchmarks misses. You're the goat!

[@evofreak1304]

uhm i have 184 fps average and 130 as 0.1 low in cyberpunk 2077. with a 265k . this is tested without updates in this game?

[@DannyzReviews]

The arrow lake data is from before that patch.

[@KneelbfZod]

@@DannyzReviews you should probably explicitly call this out.

[@warrenskeen9203]

Did you not watch or read any reviews? Good to see Intel is starting to fix some of the issues.

[@Jwalker76]

You have to keep in mind there's alot of work in these benchmarks. Sometimes new versions of software comes out while your testing and starting again from scratch isn't happening.

[@tomthomas3499]

If the other chip were tested using the same version then it shouldn't be a problem, though i do agree Cyberpunk update should be a content on it's own to point out how much improvement gained for the 285k

[@jeremyhaw7949]

What resolution were these games tested?

[@ZenFpsx]

6:41 1080p

[@jeremyhaw7949]

@@ZenFpsx thank you! I thought so, just couldnt find where that information was located.

[@NBWDOUGHBOY]

Great video Danny. Very Thorough. I'll be building a PC next month. This really helps with that decision.

[@h1tzzYT]

Happy holidays Danny, thanks for these videos as mainstream media does not provide properly tuned configurations of cpus👍

[@DannyzReviews]

Happy holidays!

[@johnscaramis2515]

Because that requires time. And tuning a CPU is a niche in the niche "PC gaming"

[@h1tzzYT]

@@johnscaramis2515 Everything worth of doing requires time. There is no excuse to not explore cpu tuning if you are a reviewer, it makes a monumental difference to gaming experience and performance. Tuned ram alone, basically gives you a new gen of cpu performance and sometimes even better due to much improved 1% low and 0.1% low fps.

[@johnscaramis2515]

@@h1tzzYT What about "niche inside a niche" aren't you able to understand? Becuase if you would be able to understand what this phrase means, you would understand why mainstream channels don't do it. And besides: each CPU will give different results. So even if they would have time to to the tuning, people will not be able to replicate. And as I said: tuning is a niche inside the niche PC gaming. Most people simply slap their CPU in and don't bother. In how many gaming PCS do you think XMP/EXPO is even enabled? And it's nice if you want to do it, but as always some people are not able or willing to understand why certain things are not done how they think it must be done the right way.

[@johnscaramis2515]

@@h1tzzYT Besides, if we move to 1440p or 4k gaming, all those hours, days or weeks of tuning are basically... useless.

[@jeffphilp7430]

Great video, many thanks for all the hard work. Would love to see a 9800X3d specific setup and tuning video, especially how you setup your PBO and Ram within the MSI bios.

[@AugmentedGravity]

5800X3D, 7800X3D and 9800X3D are absolutely goated at this point. Its almost not funny how damn good CPU’s AMD has produced.

[@Iboster1]

Add the 5700x3d to the list. Insane performance for a 150$ cpu

[@raynel8495]

It looks like you need a DEDICATED recording PC with OBS.

[@drewmorg.]

Nvenc works pretty good these days. Especially on a 4000 series.

[@kendragon2827]

Dedicated stream pc with capture card is the way to go. That will remove that 0.1% low when you compare a 16 core cpu to a 32 core. But yes the 285k is pure trash.

[@paulnguyen843]

This is one of the most detailed and well thought out comparison videos. Dude you worked so hard on this.

[@JsGarage]

For me the easy, right out of the box, stable, high performance the 9800x3D offers is priceless. Used to love tweaking my PCs now I just want it to work and be rock solid stable for gaming and work loads alike. Loving my 9800X3D so far.

[@Bojcha76]

Did you try "core tuning" to > Legacy ?

[@OGSuperDirty]

It doesn't do shit

[@johnhughes9766]

Frame cap x has the option to calculate standard deviation So tighter average and 0.1 percent lows show as more consistent gameplay Worth a look it’s a good metric to show consistent frame pacing

[@pctechloonie]

Thanks a lot for this testing even if I'll still keep using my old I7-2600K. Was very interesting to see how they compared when OC'd :D

[@LaneBatman-c2v]

The obvious solution for you for gaming is to drop in a 5700x3d and call it a day for another decade… I went from a 3600 to 5800x3d a few years back and it was insane to see the dif even at 4k. Like double fps. Sure, I also use a 4090. But the cpu was the issue. Still using an x370 mobo too… just make sure your bios and mobo are compatible first.. and sell the 2600 for 50 or whatever. So only 150-175 to get the huge performance upgrade

[@pctechloonie]

​@@LaneBatman-c2v mostly play older games at 60FPS with a 1060 6GB + indie games so the upgrade really isn't worth it yet. But sure 5700X3D value is tough to beat and could be a good pairing for a future upgrade if I can find a deal on a used Am4 bundle. Think you're mistaking my old sandy bridge i7 with a ryzen 2600 lol.

[@LaneBatman-c2v]

@ you are correct hahaha… I just read too fast and saw what jumped out. But the bundles at Microcenter are insane if you have one close enough to visit. I find it pretty amazing that AMD is still offering newer chipsets that work on x370 mobos… and also that Intel has dropped the ball into a black hole. The mismanagement and terrible foresight is jaw dropping. I called it out years ago with the initial release of x3d and how its power efficiency was double or better than Intel. And the cases where Intel was marginally faster was silly as they take double the power. Why bother? And AMD while not perfect in their releases has been miles ahead of Intel for the past few years. And now it’s literally like Intel fell into a black hole. So many mistakes in so little time. I hope they’re saved for several reasons. Nothing great comes from markets where one company dominates. My advice to you is to wait til you’re fully ready to spec a new build. And go AMD for sure if it’s in the next decade haha… the newer the platform the better as it’ll be supported for longer. I don’t see Intel rebounding in that time. I find it more probable for Nvidia or someone else jumping on the gap in the market…

[@pctechloonie]

@@LaneBatman-c2v thanks for the advice and sure AMD is the only viable option for gaming RN. unfortunately residing in Denmark I do not have the privilege of Microcenter deals :C + 25% VAT is also pretty steep.

[@michaelthompson9798]

Thanks for another great video Danny. Happy New Year to you and fellow commentors below :D

[@oysterhead5150]

Great video Danny, I can't imagine with the microcode that the 285K will get a boost to really help it competitive that much more, it would have to be close to 30% increase, which is a pipe dream.

[@KatsuKingTV]

Your affiliate link for the 9800x3d doesn't work. It directs you to the 7800x3d

[@DannyzReviews]

Probably since it's out stock 😭

[@techwandoUS]

Awesome video man. Benchmarking like that and bar graphs are so much work. It's also wild to see that Core Ultra 285K just so far behind. Saw a few people selling Ultra setups locally, thought about picking one up, but I might wait for 385 or whatever they call it.

[@DannyzReviews]

Ya totally agree, I was super intrigued by Arrow Lake at first, there's a lot you can tune with it but even then it barely makes a difference. For intel's sake I hope there's some magic sauce in that "gaming update" they're planning on releasing.

[@techwandoUS]

@DannyzReviews yeah hopefully they can give back to the small amount of people that invested in the platform

[@MisterWoes]

Unbiased, thorough work. Your channel is going to keep growing. Thank you for your efforts.

[@njasicko]

Running my 9800x3d OC as well (106.2 ELK, 5.725 Single core and 5.65 All core, IF 2300, custom neg-pos curve tune) DDR 5 A die @ 6400 CL26 with tight timings. Your 9800x3d test system is leaving performance on the table by not using a MB with an external clock gen. Newest amd agesa 1.2.0.2b with legacy game mode selected got me down to 59.1 ns in AIDA64. Thanks for the reviews and time put into this video, great work.

[@FierceSmurfN8r]

Seems like his 9800x3d is running a bit hotter than what I have seen as well. Mine is consistently in the low to mid 50s while playing games and hardly ever goes into the 60s.

[@OGSuperDirty]

Are you using pbo or co?

[@FierceSmurfN8r]

@@OGSuperDirty My settings are below. X3D Turbo Mode Disable SMT PBO positive 200 Overdrive scalar 5x Curve optimizer all cores negative 40

[@Mateoz]

@@FierceSmurfN8r SMT off will give you cooler temps.

[@FierceSmurfN8r]

@@Mateoz yeah honestly forgot i disabled it. with those settings i get better performance and temps in tarkov so worth a shot.

[@marekbambule5129]

Great video !!! I have 14900KS after my 14900KF "died"...I would like see some "tuning" of your ?

[@fated2k]

This must have taken forever to put together! Great video

[@band0lero]

4:55 I just subbed because of what you said here. The hardware enthusiast community is plagued with know-it-all commenters that are never satisfied with however way you test it, yet they don't move a finger to benchmark their own setup and make videos. They just keep being insufferable in the comment sections/reddit.

[@technik330]

Quick question, have you tested the 285K on CUDIMM memory? Thanks

[@randomcomments5034]

This is super good to finally have someone willing to OC and share benchmarks. It really is a shame that monolithic dies are on the way out. The latency and responsiveness of them are just so much better. All I want is a monolithic vCache chip with a strong memory controller :(

[@jaynayy]

Danny that's a lot of work! thank you for this!

[@Chaos_God_of_Fate]

Nice vid. Sub earned ;) I just upgraded from a 4790k to a 12900k and so far I can't see how I could need any more speed unless it was crunching hard 100% of the time and that extra speed paid for the hardware. I love this 'new' CPU. It'd have been nice to get a better Processor, but budgets and all- I ended up getting this upgrade for about $150-200; I got a bundle with a decent mid-z790 MB and 48gb of ddr5 for $410 so it's hard to say exactly how much the CPU was.

[@clayridell6374]

Man i want to thank you... That video just make my mind blow.... 😂 thanks alot for tests

[@robertmyers6488]

Good review. Would you consider taking the 9700x against the 9700x3d under tweaked conditions?

[@grasz]

The 1% low dips are from MSI afterburner PSU monitoring. This needs to be talked about more.

[@psychosis7325]

Would love to see a summary of the point one percent data as 200fps average out of any of them is great to point the immersion breaking stutter is what really matters IMO.

[@undeadfiesta9811]

If anyone does notice a dip from any of these CPU’s try any typed of frame gen, it helps with the jitters

[@builderphill1361]

Seems like a lot of hardcore overclockers found the video lol. Some good critiques but love the video! Keep up the great work

[@TheRealHypervizor]

From my testing the 0.1% low issue is a Windows scheduler problem. When I play games on Linux I notice my framerates are way smoother and I've seen other people post benchmarks on linux and their 0.1% lows are much closer to 1% lows compared to on Windows.

[@Zloy_Tipok]

Hello there! I'm sorry, but I don't speak English very well. Can you please tell me which resolution the games were tested on - 1080p or 1440p?

[@johnpaulbacon8320]

Thank's for this very detailed and wondrful video of this type. :)

[@0newithmisery]

Where are the tests that compare things like video editing and, quite simply, the duration of the boot process?

[@ibscofieldracing]

That’s amazing 🤩. I just get a 13900k for a good price but it draw so much power while gaming… is it possible to run it at 5.2 ghz while consuming less power. I have to run it at 5GHz to reduce the power consumption a bit … maybe fix voltage? But how much is enough to get good stability. Thanks

[@eugeneslepov3884]

Great job highlighting 1% and 0.1% lows. People do underestimate that metric. I know I am not in position to request or even propose content for your channel, however I can’t help it :) I think it would be great to have a detailed analysis of frame time graph since if we are talking about smoothness of gameplay 1% lows are not good enough in my opinion. I think more important is how big is deviation of frame times across median. So even if 1% lows of x3D is better but its average is like x2 higher than 1 lows, that wont feel smooth. I think Gamers Nexus try to do something like that but not good enough. That sort of content could be really unique for your channel

[@qweasdzxc]

You tested all of this right before the alleged "fix" for Arrow Lake. They're just now releasing the "fix". MSI released the BIOS for the MPG Z890 EDGE TI WIFI yesterday, with Intel Planned Release 5 (PR5) & ME firmware updated. "(Recommend updated together with the latest Intel APO software, Intel DTT, and ME driver for optimized condition.)". Could you take a look if this "fix" results in any improvements?

[@RipperDraco]

I'm very intrested to see how you tuned your 14900k to not drop frequency with Intels default settings.

[@reesesman8821]

Just got a 9800x3d a month ago and its been an undervolted beast live to see a video on undervolting it

[@Lightning---K.]

I am missing the newest Microsoft Flight Simulator and DCS ( Digital Combat Simulator ) in this comparison. Both CPU´s : AMD Ryzen 7 9800 X3D and Intel Core i9 14900 K should perform excellent, but which one is the better choise ?? Please make a short video about those two CPU´s and both Jet/Plane Simulations !!! That would be great !!! When using a VR headset instead of a 4K Monitor framerates are much lower than in any other game!!! Especially in DCS it is depending on the amount of active vehicles and aircraft in the current map. So a nice intense comparison testing video would be awesome !!!

[@cisco150]

Does this reflect in 4K as well.

[@fransoualee6529]

Generally no as it is more gpu bottleneck. Expect gaps to be smaller as you move up in resolution

[@michigantoga]

It's a no brainer at this point, has been for years. The 7800X3D is in second place, then the 14900K or 5800X3D

[@johnscaramis2515]

Especially if you also take into account power draw and derived from that cooling requirements.

[@christycho9871]

I play at 4K 120hz on my Gsync compatible TV. Without the need to go over 120fps, the 14900K was a great purchase at $385 brand new on black Friday. I power limited it to 200 watts. Don't need the insanely high frame rates. It was a huge performance leap above my old 9700K.The 98003D couldn't be found in stock anyway.

[@clarence0075]

You definitely put a lot of work in thank you for this 👍🏾👍🏾

[@ScottieMckinley787]

Serious question: Was the 285 intended for workstations?

[@michaellin1393]

Wait am I blind? How did you get 0.1% low on CapFrameX? I only see 0.2% low.

[@gamester0123]

Still using a 5800x3d. Idk if it's worth the upgrade yet to the 9800x3d

[@DannyzReviews]

If you're playing at 1440P or above you're still fine, even with a 4090.

[@gamester0123]

@ yes I am at 1440p. I should’ve mentioned it. Thank you for the reply since it does answer my question!

[@nahbro3240]

​@@DannyzReviewsso your saying upgrading to a 9800x3d is not worth it if i have a 4090 ( currently have an i7 13700k)

[@Seven7ails]

Nice ! 4090 is over my budget. 4070 to doable with this 9800x3d?

[@rbh2na]

No...Don't di that...get a 9700x and rest spend on better GPU...IF YOU DON'T GONNA USE 4090 AND PLAY ON LOW 1080P YOU WON'T NOTICE...JUST IN FUTURE GET A ZEN6 X3D OR SOMRTHING LIKE THAT....Don't just overspend on CPU and get a worse CPU...If you need productivity Intel is still better up to the i7

[@ZomgZomg007]

If your using mail afterburner then gpu power values is causing amdx3d CPUs to have very bad .1%

[@WindowCreep808]

great work!

[@just_OsaMa23]

brother for more latency improvement try disabling power down mode and gear down mode in bios.. also max tRefi you might also set tRfc to 450 or 400 but it might be unstable

[@awakenedsoul2638]

Can you make a comparison of the 14900KS versus the 9800X3D?

[@BusterFortnite]

Been waiting for weeks for my 9800x3d to arrive, after many delays should be arriving 10th january, for my £3k pc that I bought with my own money at 15 years old

[@leebutler9470]

Geez at 15 if you had invested that £3k....

[@ErikLetzner]

Is there any specific reason why the 9800X3D has comparetively slower memory? It is able to run 8000MT/s C36-40 which would mean even higher Bandwidth and lower latencies compared to 6200 C28

[@meowmeowmeowmeowmism]

Good shit bhai. Interesting info. Glad to see the to-the-point approach. Id love to see a quick video on navigating the msi bios for the 9800x3d. I feel like on my msi x870p wifi 7 that i see the same settings options listed multiple times throughout bios and I'm not sure at the end whats actually running or conflicting. Options under advanced and overclocking and amd overclocking (in advanced), like y'all could make this less redundant. Same with ddr settings - like why can I pick a memory profile or xmp/expo option then go and alter everything under it‽ Just ranting - the gigabyte x870 for my 9950x is much more user friendly. But good vid. Curious how long it took to put together

[@topelite666]

The real question here is why didnt you disable all but 1 P-core for the 285k. people online have shown doing this increases the average fps by an insane amount due to poor scheduling. Jayz2 cents did a short video on it but i wish more people did. i would buy a 285k if i saw enough benchmarks of it running 1p core only

[@friedmule5403]

I do really like your great comparison. How do you think the Ryzen 9950X3D will be compared to the Ryzen 9800X3D?

[@haukikannel]

Better in productivity because of more cores… in gaming… about the same sometimes slover when Windows scheduler put game to non cache chip!

[@friedmule5403]

@@haukikannel Thank you, I may be wrong but as I understand it, has the 9950X3D two cache chips?

[@haukikannel]

@ One.

[@johnscaramis2515]

@@haukikannel The 9950X3D can also be faster. Acc. rumours core frequencies are higher on the 9950X compared to the 9800X3D due to the higher TDP. So faster, IF Windows scheduler assigns tasks correctly (OK, better do it yourself...)

[@haukikannel]

@ definitely in productivity… in gaming… not so much. Lose sometimes wins sometimes. There is allways problem, if the program jump from one chip to another. Less likely now than ot was in the past because Windows scheduler has become better, but still It happened from time to time.

[@THEWILD-tv]

u have high end cpu why 1080p? why not run 2k or 4k?

[@yr6sport418]

When AM4 to AM5 migration was introduced, there was an uplift in performance in every aspect. But core 200 have the opposite story. Even now Intel must compete with their own predecessor CPU harder.

[@allxtend4005]

you mean with the CPU's that will fail in like a year ? Nha the real competition is against intel itself because nobody know yet how good the 285k will deliver because maybe after one year the CPU's will die as the 14700/14900/13700/13900

[@Garrett8817]

would there be a significant improvement for the 14900k with 32gb 8200ish mt/s a-dies? great testing overall though.

[@UNr34]

Good video but i wish you included 0.1% lows for all games.

[@Blitzz_69]

Have you tested if the 1% and 0.1% improof when you disable GPU Usage monitoring? Therese a view Games wher it has an Impact.

[@JDD_Tech_MODS]

Raptor Lakes IMC is not inferior. What are you comparing that statement to?

[@Antal46]

isnt the 13k+ series slowly cooking themselves due to some manufacturing bug?

[@drumtro11]

Can you do a 9950x vs 9950x3D for workstation and 4k stuff?

[@cszulu2000]

comparing an 8 core vs 24 core and the 8 core keeps winning with less power... wow!

[@pinkman6217]

Power 🦾

[@User-ys7cb]

24 cores are useless for gaming. E cores can help in some cases, but they can also ruin performance.

[@johnscaramis2515]

And how many cores are used for gaming? Effectively, we all know that Windows and scheduling is not optimal as it shifts the threads.

[@dastiffmeisterman]

1080p benchmarks are completely pointless for these CPU's. Anyone buying them will have a system running at 4K. You need to re-run at 4K and see what the difference is. I've seen this run on another site and the difference dropped to about 1%.

[@mikebruzzone9570]

Followed this very closely on the GPU VRAM usage and frames range to come to grips, some observations, concerning the CPU decode buffer (which is what they both are but 3D Large Cache [LC] in particular) and where and when look ahead matters most and matters less when performance is in parity this is about optimized encoding in relation non optimized, how slim the code is very interesting. Fat as in needs some weight reduction = 3D and Slim optimized all comers. So 3D hides fat code. On slim code 3D should provide some midrange advantage but when parity on highly optimized 3D matters less. This is more than just what is the byte size of the CPU playback requirement to feed the GPU. I'm thinking on the U distribution of applications and AMD does well everywhere BUT the middle. Arrow is an enterprise CPU that should have never been sent to game land for review. Arrow is for an entirely different market segment that is Office Business use Danny just say that please. Right, thank you "Intel should have completely avoided gaming with Arrow" and that was obvious in relation Raptor and I said that before Arrow launch. I do know why the snafu happen and its treacherous, nefarious and stupid all wrapped up into one BIG issue. mb

[@vipromeoyt]

Can you share your rtx 4090 overclock settings please 🙏 🎉

[@rkwjunior2298]

Its in the video

[@Vennux]

but new ultra core cpus with high ram speeds and overclock in ecores are beasts as jayz benchmarked them like 1 month ago before the fixes now i think they are way faster

[@3up3down.]

1. The 285k is an embarrassment. 2. Could you have tuned the 9800 a bit further since you did for the 14900… was that pbo -20? 3. Great review!

[@CaptainCave-t3n]

I wish the 7800x3d would have been included. Its all good nice video.

[@ALE79-79]

is crazy to buy intel cpu's right now... its not worth it

[@PAIN-ot4cj]

Now if you Switched the benchmark to only 3D rending and work application tests the 9800X3D and the 285k would flip and the 14900k would still be in the middle, it really is the best gaming and work CPU in one

[@DannyzReviews]

Rocking another 14900k for my personal build. I can't see myself switching off of it for a while, handles everything I throw at it.

[@PAIN-ot4cj]

@ yeah I love the entire experience. Great video Danny and happy new year man 🎉🥳

[@eduardmc]

plus gaming at higher resoultion gaming you wont see a differences. who plays at 1080p lol. how about a 4k review benchmark

[@leebutler9470]

The same 14900k with microcode issues that burns out with random odds like a lottery...? Would never touch a 14th gen.

[@allxtend4005]

no the best gaming and work CPU in one is the 7950x3d and the 9950x3d are coming soon, why even try to say that the 285k is anywhere good when it comes to the lowest FPS even lower then it's older generation, And then we have Gaming is one thing and working another thing, if you work with your PC then get a PC for work and if you game then a PC for gaming, i think you never worked with your PC because otherwise you will understand how much troubles games and gaming software can cause on a work Computer.

[@swolypanda2976]

What is the 1%lows mean

[@awakenedsoul2638]

For a good example, the game The Witcher 3, a CPU can 300 Frames Per Second (F.P.S) the highest it can reach! Its 1% low is 190 FPS. This means under heavy load, the lowest fps it will go down to is 190 FPS. It is the lowest FPS it can run the game!

[@guille92h]

No, The 1% low means the game will be under those fps for 1% of the run, the absolute lowest would be mininimum fps, is a bit tricky to understand, if you do a graph with all the fps and organize it from higher to lower, then you take the 1% part (lowest percentile) and the highest fps on the piece of the graph would be the 1% low.

[@awakenedsoul2638]

@@guille92h Are you sure that is it? You mean 1% Low is lower than the lower frames that you get on the game? How about 0.1% lows what does it mean?

[@LostInJPN]

​@@awakenedsoul2638 Pretty sure 1% and 0.1% lows are just the average for the 99th and 99.9th percentile respectively. Monitoring software has something called a polling rate, which is basically how often it refreshes the data its monitoring. So if you have 1000ms polling rate, you will have your FPS update each second. Each of these FPS updates is a new data row. To get 1% lows, you average out the bottom 1% of these data rows. Same with 0.1% lows. There are other calculations that involve taking the highest value from the 1% or 0.1% rows as well. I usually just average it out.

[@awakenedsoul2638]

@@LostInJPN So you mean in my example the low 180 FPS, its 1% low is lower than 180 FPS? Am I correct? You said to average it!

[@kimhenrypc9655]

I get why 1080p is used for accurate FPS benchmarking, but I’m fairly certain most gamers buying these high-end CPUs will be gaming at 1440p or 4K. When paired with high-end GPUs, the FPS differences at those resolutions are minimal or within the margin of error specially at 4k

[@Aracii]

This is true for most but not all games. More and more games are becoming cpu limited at 4k. Space marine 2 hammers the CPU, stalker 2 and quite a few other games including cyberpunk. Going forward the differences could become more exasprated as CPU requirements increase. My 5800x3d averaged 80-90 eith 50-60 on 1% lows at 4k maxed with a 4080super whereas my 9800x3d sits at 100-120 with 70-80 on 1% lows

[@KOAP33]

Sure that's 100% correct and in that case most of those people do not need to upgrade their CPU's you just need some better GPU and on 4K you can have perfectly fine experience even with some oldish ddr3 platform

[@Bass.Player]

You should have waited for for the 114 update on the 285K, MSI released it days ago. Do some productivity comparisons...

[@reviewforthetube6485]

I'll say it again and again if you own a 14900k or even a 7800x3d you do not need an upgrade lol I also love how the big mainstream techtubers leave out the 0.1% and 1% lows unlike you. Those matter so much more for smooth consistent gameplay. They did a great job on the 9800x3d in that area compared to past amd cpus. You could get insane averages sure but it never felt as consistent as intel cpus to me. Nice to see this.

[@OGSuperDirty]

Huh? I have both 14900k/9800x3d and the AMD cpu is smoother in all my games with better lows.

[@reviewforthetube6485]

@OGSuperDirty did you even read? I said that they did a great job on that part with the 9800x3d compared to past amd cpus. That its nice to see what they did with the 9800x3d as I mentioned they worked on that with it. Was talking about past amd products. I said it never felt as consistent as intel cpus to me meaning past tense. Then said nice to see this meaning improvements on the 9800x3d.

[@prototypep4]

Who leaves them out? Gamers nexus, LTT, Jayztwocents, Hardware unboxed ALL keep 1% lows and .1% lows in their tests. tf are you watching?

[@External2737]

Yeah, no need to upgrade every generation. I didn't even buy the ultimate gaming CPU or GPU and I'll be fine for a few more years (7900x and 7900xt in a nice bundle a while ago).

[@reviewforthetube6485]

@@prototypep4 please tell me the last time any of the mentioned which felt smoother or was more consistent feeling? All they talk about is oh look at the average fps! Omg! Never actually telling you how it feels. Its how they sell it.

[@JMNovak1011]

would have been nice to see a 9950X OC with 6400mhz or 8000mhz mem. added.

[@ALEX336X]

Grabbed a new 14900k a couple days ago on a really good sale at $285. That’ll be replacing my 12700k when it arrives. Not a massive upgrade but it’ll scratch the itch. Really appreciate these benchmarks on properly tuned systems.

[@DannyzReviews]

Absolute steal of a deal. Make sure to adjust those AC/DC Load-Lines

[@TsunaXZ]

Where did you get that deal? I can't even find 14700k for 300

[@SteveT-r6k]

@@TsunaXZ I'll have a Delided 13900KS and the EK water block made just for the delids for sale in a few weeks.

[@burakyilmaz7363]

@@DannyzReviewshi What do you mean with ac/dc load Lines?

[@peacem8574]

​@@burakyilmaz7363 The 14th gen have production faults. They oxidize and then become unusable. They also wear out faster.

[@Slay3rOne]

Don't use 24H2 for the 9800X3D, I mean at least do a comparison between 23H2 and 24H2 to see the performance difference. I have something like a good 20% more performance, and much better 1% and 0.1% low on Black Ops 6 with the 9800X3D (RTX 4080, at 1440p) when downgrading to 23H2 compared to a fresh install of 24H2. I did install a fresh 24H2 when I got the 9800X3D, the performance was so weird on Black Ops 6 in particular, it made no sense. Downgrading to 23H2 (fresh install again) finally showed the real performance. It jumped from like 250fps to above 300fps average. I also have a 14900K build with a RTX 3080 running 23H2, and I had more or less close to the same performance on this build compared to the RTX 4080 and 9800X3D running 24H2. That made zero sense. So I don't know what's up with the 24H2 build, but until these issues are fixed, I'm staying on 23H2.

[@jasonfch]

why do the14900k use more System RAM to achieve results close to 9800X3D?

[@kartikpintu]

Games are cache bound especially AAA. Intel doesn't have 3D cache so it has more memory interactions.

[@jasonfch]

@@kartikpintu So on top of the expensive overheating last gen 14900k, a buyer have to pay more for RAM just to get close to the performence of 9800X3D ? No thanks Intel lol.

[@incaseofemergency6051]

Why would I want to "get ready for the 5000 series"? I will never overpay for an nVidia GPU again.

[@adi6293]

HAPPY NEW YEAR TO ALL OF YOU , HOPEFULLY 2025 IS BETTER THAN 2024 !!

[@vencler]

Should have also included 0.1% lows in the graphs.

[@abtechgroup]

how about testing xplane and msfs..?

[@Exodus.1337]

Would love to see the BIOS settings for the ring overclock on the 14900k...

[@n1kobg]

You still havent OC the E-Cores. P-cores OC doesnt do much but E-Cores get to above 5Ghz which helps the performance a bit. PS: The CPU needs higher Core & Ring Clock but the new architecture (and TSMC node) doesnt allow more. If Intel released the Alder Lake architecture with the new cores they would get the needed performance to stay ahead in gaming.