DDR4 vs. DDR5, New Game Benchmark + Has 5800X3D Aged Worse Than 12900K?

Поділитися
Вставка
  • Опубліковано 3 лют 2025

КОМЕНТАРІ • 1,2 тис.

  • @ayuchanayuko
    @ayuchanayuko 7 місяців тому +949

    Context: DDR5 prices were sky high with the release of of the 12900K. That and the sky high prices of DDR5 boards too. Compared to DDR4 prices and the price of the 5800X3D at the time, it was truly good value for its performance.

    • @jonivilander5758
      @jonivilander5758 7 місяців тому +72

      im still thinking buying 5700x3d to replace my 5600x i play 1080p with rx 6700...

    • @mircomputers
      @mircomputers 7 місяців тому +2

      bards were cheaper after the release wave, as no one was purchasing them

    • @Gazer75
      @Gazer75 7 місяців тому +35

      @@jonivilander5758 Not worth it I think... I upgraded to the 5800X3D and got some big improvements in some of my builder and management games, but little to none in others.

    • @YordanHristooov
      @YordanHristooov 7 місяців тому +53

      @@Gazer75it’s important to know which games he is playing because in MMOs, sims and a few other online games the cache makes big difference and especially in the lows. I’ve experienced it myself! Also my friend replaced his 5950x for 5800x3d, because we were doing 50v50 wars and he couldn’t believe how much better the 5800x3d was in that

    • @High_On_Hope
      @High_On_Hope 7 місяців тому

      ​@@jonivilander5758save for a better gpu
      Go for x3d if u have the budget for 4070ti/7900xt level of gpu power

  • @marinipersonal
    @marinipersonal 7 місяців тому +507

    7800x3D at £314 in the UK, and can be cooled by a £30 air cooler IS the best bet for gaming. No arguing.

    • @Toufan1-gr4br
      @Toufan1-gr4br 7 місяців тому +42

      phantom spirit 120se my beloved

    • @AlexHusTech
      @AlexHusTech 7 місяців тому +95

      Also doesn't need an addition PSU just to power it like the 14900 does lol :yt:

    • @TechnoFreak-IN
      @TechnoFreak-IN 7 місяців тому +56

      7800x3d temps are ridiculously amazing for its gaming performance!

    • @ShadowsOfThePast
      @ShadowsOfThePast 7 місяців тому +6

      ​@@TechnoFreak-IN what kind of temps are you getting in gaming and what sort of cooler..? I'm upgrading and gonna go for 7800x3d most likely.

    • @Boofski
      @Boofski 7 місяців тому +32

      £314?
      Bro the 2nd hand market is being flooded with 7800x3D’s because of zen 5 being around the corner. I can pick one up for 260 euro here in the Netherlands that’s like £220

  • @FatetalityXI
    @FatetalityXI 7 місяців тому +968

    These "experts" on X 😂

    • @Hardwareunboxed
      @Hardwareunboxed  7 місяців тому +263

      Gotta love'em :)

    • @insideline158
      @insideline158 7 місяців тому

      @@Hardwareunboxed "Xperts", that platform really is full of them.

    • @saphyre-l4w
      @saphyre-l4w 7 місяців тому +109

      Ah, Twitter... the Digital mental asylum. 🙂

    • @christophermullins7163
      @christophermullins7163 7 місяців тому +84

      The fact that no one has said "X-perts" 😥

    • @Owen-np3wf
      @Owen-np3wf 7 місяців тому +8

      @@saphyre-l4wthought it would of been reddit

  • @DragonOfTheMortalKombat
    @DragonOfTheMortalKombat 7 місяців тому +694

    It's not 5800x3D aging worse than 12900K, it's DDR5 aging like a fine wine. More games are taking advantage of the high bandwidth.

    • @fVNzO
      @fVNzO 7 місяців тому +116

      Availability of lower CL 6000 and 6400 memory also went from being premium to the standard cheap option. So most people building today will have decent lower latency memory that'll just work through XMP with zero tweaking necessary. This will give a couple percent more performance in favor of a DDR5 config.

    • @FatetalityXI
      @FatetalityXI 7 місяців тому +74

      But but the "experts" on X will tell you it's the 5800x3D aging like milk. How dare you make sense!

    • @stennan
      @stennan 7 місяців тому +24

      ​@@FatetalityXIjokes on them, I like cheese 😋🧀

    • @SR-fi8ef
      @SR-fi8ef 7 місяців тому +3

      ​@@stennanit's made from milk!

    • @insideline158
      @insideline158 7 місяців тому +8

      I wonder if equally large gains can be had if/when AMD can make CPUs that can fully use DDR5 speeds.

  • @avi4796
    @avi4796 6 місяців тому +22

    Steve, the "lines" separating the graphs per CPU tier makes it far better to handle then without. Same for those % numbers. Much much easier. Good idea 🍻

  • @JarrodsTech
    @JarrodsTech 7 місяців тому +412

    Steve staring into my soul 👀

    • @fleurdewin7958
      @fleurdewin7958 7 місяців тому +21

      Your soul has been leaking data just like Zotac . Steve picked that up.

    • @MuhlisErtugrul
      @MuhlisErtugrul 7 місяців тому +1

      DDRy stare

    • @ibslice
      @ibslice 7 місяців тому +2

      yup.. he keeps it in a nexba bottle.

    • @EbonySaints
      @EbonySaints 7 місяців тому +1

      It's like an old 80s slasher flick poster where the killer is staring you down before he goes in for the kill. Steve channeled that Freddie Krueger energy hard for that thumbnail.

    • @Shadow_banned_by_YouTube
      @Shadow_banned_by_YouTube 7 місяців тому +2

      👁️👄👁️

  • @cracklingice
    @cracklingice 7 місяців тому +177

    Would have been nice to see the 5800x3d in the final charts. I know you guys do not recommend AM4 but for anyone that wants a board with two 8 lane PCIE slots, the price of AM5 board is literally double so the 220 dollar increase in platform cost does make AM4 a consideration, particularly with the 5700x3d at just 200 dollars. I mean that's the CPU and board for the cost of just the AM5 board.

    • @paulrmurrayful
      @paulrmurrayful 7 місяців тому +22

      Yeah and the saved cash could help afford a much better GPU to pair it with.

    • @cracklingice
      @cracklingice 7 місяців тому +2

      @@paulrmurrayful I get that, but it is not my concern as I already have a GPU. I just want to be able to upgrade from my X99 setup and still be able to have a GPU a network card and my SSDs.

    • @n3xus236
      @n3xus236 7 місяців тому +7

      Wrong and outdated, decent am5 motherboard can be found for as low as $150. Not $220. You 5800x3d fanboys really love inflating numbers for new gen, reminds me of 10series cards owners. We get it was a good investment, but times change

    • @Shadowninja1200
      @Shadowninja1200 6 місяців тому +1

      The issue is that you're buying into a dead platform. I tell anyone looking for a new system that if they want the best gaming performance for 2-3 years is go for am5/7800x3d. You get the benefit of potentially doing a cpu swap down the line (if amd does a similar support like the am4 line) and it's easier to swap a gpu out then a whole motherboard/cpu. A solid foundation makes for an easier time keeping a system updated for years imo.
      That and you're looking at pcie gen 5 which can be split easily because of the insane bandwidth that it run at. Highly doubt anyone will actually use up a gen 4 lane much less a gen 5 lane. I mean hell we're just now getting to the point where gen 3 is a bottleneck.

    • @markh4750
      @markh4750 6 місяців тому +6

      @@n3xus236 Yeah, I recently updated my son's system from a R5 3600 to a 5700x3D, which made a lot of sense since he was already on AM4. When I built new for my stepson, it was 7800x3D all the way, because it didn't make sense to go back a gen.

  • @auritro3903
    @auritro3903 7 місяців тому +305

    5800X3D may not be as powerful now, but it will always remain in our hearts...

    • @oogwaythesussyturtle
      @oogwaythesussyturtle 7 місяців тому +75

      tbh it's still one of the best value gaming cpus out there. u can grab a 5800x3d around 300 bucks these days and it's simply unmatched in gaming at that price point. especially since older mobos and ddr4 ram is dirt cheap.

    • @Arejen03
      @Arejen03 7 місяців тому +16

      Me who recently swapped 2600 to 5600x 😅😭

    • @DevouringKing
      @DevouringKing 7 місяців тому +7

      In Star Citizen, its still extreme good because X3D Cache Advantage is huge there. with 40 Watts CPU drain i got over 400fps inspace.

    • @mikem2253
      @mikem2253 7 місяців тому +14

      Its still a very powerful gaming cpu.

    • @angeltzepesh1
      @angeltzepesh1 7 місяців тому +38

      Ehh, i dont know why everyone has this mentality "it's not the biggest number, it is not good anymore". Most people are still gaming at 1080p/1440p with 120hz/144hz panels, a 5800x3D will still suffice for such a scenario.

  • @MuhlisErtugrul
    @MuhlisErtugrul 7 місяців тому +199

    The time I upgrade to DDR5, DDR6 will be the new trend so I'm gonna stay at DDR4 for now

    • @davewhitefield1030
      @davewhitefield1030 7 місяців тому +16

      same, bought and built my pc with the latest ddr4 and am4 availability while ddr5 and am5 were barely emerging on the market, can't be happy enough with the purchases

    • @brockysthoughts1662
      @brockysthoughts1662 7 місяців тому +4

      Nar its just going to change form first in to this standard lpcamm2 so we will all need a new mb again

    • @eliadbu
      @eliadbu 7 місяців тому +25

      new revisions of DDR release every 6-7 years, and it's adopted year or two after official release. If you are the type of guy that wait for the new tech to come down in price, I would say you will wait until 2030 at the very least. so upgrading to DDR 5 somewhere until then isn't unreasonable.

    • @endlessgrinding3764
      @endlessgrinding3764 7 місяців тому +2

      ​@@eliadbu a good assumption. I have heard same from some of my friends as well.

    • @christianmino3753
      @christianmino3753 7 місяців тому

      @@brockysthoughts1662 Oh, I'm sorry. I didn't know you could see the future.
      You understand the 2 at the end of Cam, would imply that Cam[1] came first, right? Cam has been around for a while now, and we have to wait to see what happens to know if it will replace DIMM. We've had plenty of different ram form factors after DIMM became a thing, and plenty of form factors within DIMM itself too. Sodim, Udim etc.

  • @simplyscholar21
    @simplyscholar21 7 місяців тому +64

    My 5800X3D was on a flash sale. So it cost me 290 CAD. I replaced a 3600X. 12900k was never going to be able to compete on price with a drop in solution that good.

    • @dieseIboy
      @dieseIboy 6 місяців тому +1

      That’s a great price! I got mine for $315 CAD. Where did you get yours at with that great price? Cheers 🇨🇦

    • @bufalong
      @bufalong 6 місяців тому

      amd cpu work well? any crash in gaming?

  • @RedundancyDept
    @RedundancyDept 7 місяців тому +101

    I'm glad you tested with DDR4-3600 instead of 4000. The former is more representative of what people actually use. Locked Alder Lake CPUs don't even have the option to use memory speeds much higher than 3200 in Gear 1 mode.

    • @CaptainScorpio24
      @CaptainScorpio24 7 місяців тому +2

      really ?? damn mine stuck at 3000mhz on i7 12700 non k

    • @MuhammadRidwan-pe7ny
      @MuhammadRidwan-pe7ny 7 місяців тому +2

      @@CaptainScorpio24 really? even my i3 12100 could manage 3200 on xmp. maybe try changing the ram itself.

    • @CaptainScorpio24
      @CaptainScorpio24 7 місяців тому

      @@MuhammadRidwan-pe7ny i hv old 32gb set of ram that i used to run on my i7 7700 non k..

    • @Kumoiwa
      @Kumoiwa 7 місяців тому +18

      Yes, no point on testing DDR4-4000 when it's usually even more expensive than regular DRR5-6400. Even 3600 cl14 is unrealistic as most people will buy 3600 cl18 or 3200 cl16

    • @Solrac-Siul
      @Solrac-Siul 7 місяців тому +2

      I understand your point but he is testing K versions so the results presented even with ddr4 would surpass what one would get with locked version. And quite bluntly if there are people running 13600k/14600k and 13900k/14900k with ddr4 3600 I do not know what to say except that they should not have purchased those cpus and use ddr4 . The point of K version is to tweak them, and 13600/14600, 13700/14700 and 13900/14900 can run 4000+ in 1 to 1 ratios. I run my 14700k at 4533 16-16-16-36, and that allows it to match ddr5 7600+ in many games. in fact in some titles , like arma , the ddr4 setup beats my other ddr5 system, a 14900k running ddr5 at 8400. Let me repeat 4533 1 to 1 beats ddr5 8200 in some games, and probably will beat 8600 or 8800 also since the increase in performance above 8200 is very limited to the point i prefer to run 8200 instead of 8600 in order to limit ram temperatures.

  • @hasnihossainsami8375
    @hasnihossainsami8375 7 місяців тому +168

    X3D's biggest selling point isn't that it's the fastest - honestly, the average % difference between all of the CPUs mentioned here mean nothing at the proper resolution, because 1080p on a 4090 is silly.
    No, the real selling point of X3D is the 1% lows. Smoother gameplay > highest fps any day. And that is where even the 5800X3D sometimes comes out ahead of the 14900k, while the 7800X3D is on another planet.

    • @magnusnilsson9792
      @magnusnilsson9792 7 місяців тому +37

      Indeed, cache is king, for the 1% low.

    • @Nib_Nob-t7x
      @Nib_Nob-t7x 7 місяців тому +9

      4:16 Maybe look at the charts. Its 1% low is worse.

    • @copperypuddle3858
      @copperypuddle3858 7 місяців тому +27

      @@Nib_Nob-t7xWhen compared to a faster CPU using faster RAM, yes. Compare that to the DDR4 results and it’s better. Now compare the 7800x3d to Intels equivalent, the 7800x3d is always superior in 1% lows.
      Cope’n’seethe

    • @copperypuddle3858
      @copperypuddle3858 7 місяців тому +10

      @@hasnihossainsami8375 How is it possible that people still don’t understand why games are tested at 1080p. How.

    • @NotMalix
      @NotMalix 7 місяців тому +12

      and it's an upgrade. If someone is building an entire new computer I'd recommend AM5, but if they have an AM4 already and just want to upgrade, the x3d chips are a solid choice

  • @SoranPryde
    @SoranPryde 7 місяців тому +92

    Yes the 12900K + DDR5 combo today is faster in modern titles than the 5800x3D + DDR4 combo
    But you also have to remember that back when the 12900K was new, to get its full potential you needed a decent Z690 DDR5 MB as well as DDR5 DIMMs to go along with it. DDR5 back then was expensive
    The 5800x3D on the other hand ran perfectly fine on cheap potato B550 MBs and 3600Mhz XMP D4 was much lower in pricing than even cheap potato 4800Mhz JEDEC D5. And for the 12900K + DDR5 to beat the 5800x3D + DDR4 by a decent margin, it needed 6000Mhz and above D5 which was EXPENSIVE AF

    • @cameronbosch1213
      @cameronbosch1213 7 місяців тому +3

      Keep in mind that DDR5 prices have freefallen in the 12 months since Ryzen 7000 was released. It's now a no brainer for new system builds.

    • @SoranPryde
      @SoranPryde 7 місяців тому +13

      @cameronbosch1213
      I'm comparing in the context of how market conditions were when both the 5800x3d and 12900k were brand new. Of course we know DDR5 price nosedived 1 year after Ryzen 7000 launch but by then it was already half a year since 13900k was launched

    • @chomper720
      @chomper720 7 місяців тому +8

      You can even go lower than B550... B450!!!

    • @adlibconstitution1609
      @adlibconstitution1609 7 місяців тому +1

      Past is the past. Bottom line is I will choose 12900k + ddr5 over the 5800x3d when building a new PC.

    • @SoranPryde
      @SoranPryde 7 місяців тому

      @@adlibconstitution1609
      The video's title is "Has 5800X3D Aged Worse Than 12900K"
      Therefore we are comparing how they were when they were both brand new back then to buy, and how they age till today in terms of performances vs money spent back then to enjoy from then until now

  • @TheHoff6300
    @TheHoff6300 7 місяців тому +38

    The real value of the 5800x3d was (an sometimes stills is) the upgrade from older Ryzen CPUs. It really pushes your PC on a whole different level when you drop a 5800x3d in your 2700x system. For a relativ small amount of money you are able to breath enough CPU power to your system to last a fair bit longer.

    • @francoiscoupal7057
      @francoiscoupal7057 6 місяців тому +2

      I'm planning to do just that very soon to give my current desktop a few more years of use. Updating from a 3600.
      And since most of my games are CPU intensive sims (Stellaris, Oxygen Not Included, Rimworld, MOBAs,...) I don't even need a new GPU to max it.

    • @alexanderramos9121
      @alexanderramos9121 6 місяців тому +4

      Agreed 100%. 5700x3d has even better value. I find it comical AMD actually made a part for AM4 users right before their AM5 launch that literally has 78% of a 7800x3d's performance according to the 13 game average at 1:54 for less than half of the cost it would be to get a new motherboard, new ram, and the new cpu.

    • @MissingNumb
      @MissingNumb 6 місяців тому

      The upcoming 5900xt are gonna be an epic productivity drop in upgrade for anybody still on AM4... When they drop into $200 territory my old 2800x productivity rig can be productive again lol.

    • @tonicipriani
      @tonicipriani 6 місяців тому +2

      I just did a full renewal of my PC from late-2019, from 3700X and 2060 to 5700X3D and 4070 Super, doubled the RAM to 32GB, all on the same motherboard. I should be set for the next while, maybe even till AM6.

    • @nateo200
      @nateo200 6 місяців тому

      @@toniciprianiYeah I’ve got a 3900X with a 3060Ti. 5800X3D or 5900X not sure which. I’ve got 32GB of RAM but it’s 2400MHz so maybe I should upgrade that as well

  • @tarkz7763
    @tarkz7763 7 місяців тому +5

    Another great video as always. As soon as I saw the title for this video it made me wonder if you would consider adding "Escape from Tarkov" to the battery of tests. There seems to be either a lot of conflicting opinions or misinformation, but for a few years now lots of people say it doesn't really matter too much about your GPU or CPU, but rather your RAM. A large portion of the player base can't/wont play one map in particular, despite many having decent, if not high end PC's. Between this video and your other recent video covering capacity, it's made me realise that there is very little coverage on RAM and I am personally really appreciative of this content

  • @PookaBot
    @PookaBot 7 місяців тому +15

    I've seen people suggest that the 3D vcache makes memory bandwidth less important to those parts. It might be interesting to test something like the 7800x3d with slower memory to see if it's less affected than expected.

    • @rallyscoot
      @rallyscoot 6 місяців тому

      yes want to see that review too. It might give us some intresting data.

    • @ismaelsoto9507
      @ismaelsoto9507 6 місяців тому +1

      Funilly enough Hardware Unboxed did just that three days ago: ua-cam.com/video/aD-4ScpDSo8/v-deo.html
      And yes the Massive L3 Cache does make the R7 7800X3D less reliant of fast DDR5 compared to regular Zen 4 CPUs, also the same for Intel 13th Gen vs 12th Gen, the big increase in L2 Cache and extra L3 helped signficantly.

  • @MN12BIRD
    @MN12BIRD 7 місяців тому +11

    Damn some of these differences are mind blowing tbh I remember when DDR5 launched 4 years ago (Yes, it was 4 years ago now!) and in gaming benchmarks it was only like %2-5 faster than DDR4 at best if you were lucky (some games it was basically the same!) and now to see %20 and even %30 gains is crazy. I know DDR5 RAM has got a lot faster since it launched. I mean what was the best at launch 5600 at cl38 or something like that? I wonder how much of this improvement is just the DDR5 RAM being faster vs games actually using more memory bandwidth?

    • @Violet-ui
      @Violet-ui 6 місяців тому +1

      Now you can get 7000 CL32 for 110 dollars

    • @danitch6848
      @danitch6848 6 місяців тому

      I think its mistake.
      They used ddr in gear 2

  • @buckiesmalls
    @buckiesmalls 7 місяців тому +22

    I am glad your clocks are upside-down. Get your vids around 7am (US) right after i get off work. They are nice to wind-down to.

    • @Osprey850
      @Osprey850 6 місяців тому +1

      Funny that you say that *their* clocks are upside down when you're the one getting off work when most of us in the US are just waking up. 😆

    • @buckiesmalls
      @buckiesmalls 6 місяців тому +1

      @@Osprey850 Yeah. It works because MY life is upside-down with my working hours, as an Electrical Lineman. LOL

  • @ishshah8695
    @ishshah8695 7 місяців тому +15

    Thanks Steve!

  • @Phil-q7h
    @Phil-q7h 3 місяці тому +2

    It’s great seeing what CAN be achieved and i understand the point of the exercise, however real world usage would be much more interesting. Who runs games at 1080p with a 4090? Seeing what difference DDR4/5 makes at 1440p ultra or some configuration of 4k would mean a lot more to most people’s everyday use.
    I see very little difference as far as game experience goes running at 4k. Often 2-5%, 10% at most

  • @AdamWebb1982
    @AdamWebb1982 7 місяців тому +27

    Still rocking a 5800x3d on my day one purchase of a X370 board… madness that longevity.

    • @Ferox63
      @Ferox63 6 місяців тому +3

      Me too. I'm still running a Preorder X370 C6H. The fact that we were able to run so many generations of CPU's on one board is wild.

    • @AdamWebb1982
      @AdamWebb1982 6 місяців тому +1

      @@Ferox63 yeh C6H here as well 💪

    • @W1ldSm1le
      @W1ldSm1le 6 місяців тому +2

      X370 taichi with a 5700x3d here. 3rd cpu in the same board. Probably end up using the same board for a decade, as someone who's been building computers since 2004 it's pretty mind blowing to me. Used to upgrade virtually everything every 2.5 years.
      Even case design has come so far that it'd be hard to justify upgrading a good case made in the last 8 years.
      The barrier to entry for a new builder has become more expensive over time but 10 years ago nothing you bought was ever worth keeping. The recurring costs have actually gone down. A 300$ hard drive in 2005 was functionally worthless by 2008 for instance. But if you bought a 300$ nvme drive in 2021 it wouldn't be a paper weight today.

    • @Third_Tower
      @Third_Tower 6 місяців тому +3

      5800X3D on a B350 lmaoooo

    • @dasboot5366
      @dasboot5366 6 місяців тому +1

      5800x3d on CH7 X470 +6800xt. Fabulous! I don't see any point in switching to AM5 in the coming years.😁😁😁

  • @Jasontyo
    @Jasontyo 7 місяців тому +53

    I just purchased a 5700x3D, upgrading from a 3900x and I feel it was worthwhile. Keeping my old system running for a reasonable price is more important than having the absolute fastest thing possible. I only have a 3070 after all. Honestly, I think the 1440p display was a mistake, staying at 1080 would yield such better performance.

    • @obeliskt1024
      @obeliskt1024 7 місяців тому +4

      what kept you from getting the 5700x3d instead? could've been the last upgrade you needed.

    • @imo098765
      @imo098765 7 місяців тому +3

      DLSS and FSR/XESS. This is the way

    • @jimdob6528
      @jimdob6528 7 місяців тому +7

      @@obeliskt1024he/she did get the 5700x3d. Unless you are asking why he/she didn’t get the 5800x3d? If so it’s 100-120 dollars more expensive fore less than a 5% performance increase over the 5700x3d.

    • @obeliskt1024
      @obeliskt1024 7 місяців тому +11

      @@jimdob6528 nah, they edited the post. They initially mentioned the 5700x non 3d. So I guess all is good

    • @Jasontyo
      @Jasontyo 7 місяців тому +6

      @@obeliskt1024 Typo. It was the X3D. for $192

  • @NEZLeader
    @NEZLeader 7 місяців тому +3

    These new style of thumbnails as of late have been hilarious & awesome

  • @TheZoenGaming
    @TheZoenGaming 7 місяців тому +39

    LOL My 5800X3D is still going strong. I've got it working with 32GB of DDR4 @3600 MHz CL14-16-16-36 and a 3090. There's no "degradation" to be seen.

    • @Resolute.j
      @Resolute.j 6 місяців тому +1

      This is my rig except on a 5900X. CL14 @3600Mhz is so good, I don't think I'll move up until there is an equivalent in the current or upcoming platforms.

    • @metradawn8071
      @metradawn8071 6 місяців тому

      What's the point of running b-die on a 5800x3d? Any shitty djr could have done the job. Also, if you're running your 3600mts cl14 b-die at stock, you might as well run at jdec spec then sell your cpu and replace it with an x3d chip.

    • @exoticspeedefy7916
      @exoticspeedefy7916 6 місяців тому

      @@metradawn8071 Latency makes a difference even on 5800x3d. I got down to 55ns with 1900fclk

    • @Great.Milenko
      @Great.Milenko 6 місяців тому

      ​@@metradawn8071sell his 5800x3D and replace it with an X3D variant? ... Hmm 🤔

    • @cannaroe1213
      @cannaroe1213 5 місяців тому

      @@metradawn8071 there isn't such point in b-die for any Ryzen as the infinity mesh or whatever it's called is capped at 4000Mhz. Still, I have 4 sticks in my 5950X and I won't be upgrading for at least 5 years.

  • @AtifSheikh
    @AtifSheikh 7 місяців тому +8

    This puts a spotlight on another thing... How minuscule 14th gen uplift actually is compared to 12th gen.
    Assuming if we put same ddr5 speeds on both 14900 and 12900 the difference shown here will only drop more.

    • @andersjjensen
      @andersjjensen 6 місяців тому

      The architectural difference between 12th and 13/14th is just an increase in L2 and L3. So it has, thus far, been Skylake all over again.

  • @amusetech
    @amusetech 7 місяців тому +26

    At the same time this video once again shows what a beast of gaming CPU the 7800X3D really is...

    • @aberkae
      @aberkae 7 місяців тому +4

      Yeah it's aging well against multiple generations vs i9 13900K/S and now i9 14900k/S. Intel decided a double smack down is needed renaming aside 😂.

    • @jimdob6528
      @jimdob6528 7 місяців тому +7

      I mean it also shows how well the 5800x3d and 5700x3d is aging as well. The fact I could put a 5700x3d on my b450 motherboard for 200usd and get a 60% increase on just the cpu performance is crazy. Basically a new pc and in 4-6 years I can platform jump to the am6 platform when that comes out and it’s fantastic all around. Love the amd/intel competition

    • @amusetech
      @amusetech 7 місяців тому

      @@jimdob6528 Sure! If someone is on the AM4 platform and just want to get the last boos out of it, it's an amazing deal and value.
      For people buying new gaming rigs, I can't see a valid reason to go with anything else than 7800X3D.

    • @donaldnemesis393
      @donaldnemesis393 6 місяців тому

      I'm curious about how it performs on video rendering to see if it also the budget option for video editing as well

  • @EnigmasityGG
    @EnigmasityGG 7 місяців тому +16

    I‘m really happy I went with the 7800X3D about a year ago. It’s an absolute monster, especially paired with a nice kit of 6000 MHz CL30 memory. The craziest thing is, that it runs at 40-70 W while gaming and at a max of 90 W in an all-core workload. Also gives you a nice upgrade path to AMD‘s new X3D chips down the line.

    • @andersjjensen
      @andersjjensen 6 місяців тому

      Unless you're running a stupid high end GPU at stupid low resolution you'll be well and truly set until Zen 6 X3D is being pushed down in price by Zen 7.

  • @umusachi
    @umusachi 7 місяців тому +1

    Great information! As a long term viewer, I like this multi-pronged up format!! What is great about keeping the older CPUs / hardware in the testing is buying advice in the sense that it helps you compare your current / older system. Even a CPU or two from those series. Like I'm on 10th get and would love to know how it compares, although I'm sure if I spent enough time digging all the information is there in your archive of videos. Excellent content. I really appreciate all the hard work you both do!!!

  • @ivofixzone6410
    @ivofixzone6410 6 місяців тому +42

    The video is about 5800X3D aging. Why game benchmarks are using 7800X3D , I'm confused.

    • @MrDeathknight1
      @MrDeathknight1 6 місяців тому

      I thought the same and it would at least be on the chart but they don't have the 3800mhz stuff anymore

    • @UKKN516
      @UKKN516 6 місяців тому +2

      You skipped the video lol

    • @MrDeathknight1
      @MrDeathknight1 6 місяців тому

      @@UKKN516 na watched it all

    • @BananaDynastyX
      @BananaDynastyX 6 місяців тому +5

      @@MrDeathknight1 The video is about DDR4 vs DDR5 performance. The section about the 5800X3D is obviously just discussing why this testing is being done...

    • @MrDeathknight1
      @MrDeathknight1 6 місяців тому

      @BananaDynastyX fair I just thought it was about aging v cash and then it was mostly about ram speeds ddr4 vs ddr 5 are the cash latency s coming down on ddr 5 now are we even in the 20s yet cuz the band with s rnt making a huge difference I don't think

  • @ilijazafirov4192
    @ilijazafirov4192 6 місяців тому

    Liked this video, it is much much better to understand when comparing to older review and see the differences than just saying it how purely it is right now with top of the line GPU etc, really like these comparisons, gives us great information for newer builds, keep it up please, greatly appreciated!

  • @winterwolf9376
    @winterwolf9376 7 місяців тому +2

    We need merch with that image Steve!

  • @ianwiinamaki4770
    @ianwiinamaki4770 6 місяців тому

    This is the best thumbnail I have ever seen( just in case it changes) the one with the eye reflection on the ram.

  • @rockingfps5104
    @rockingfps5104 7 місяців тому +9

    There is a possibility to include memory latency in next vídeo? Lower latency is very important, maybe, more important than MHz / throughput.

    • @andersjjensen
      @andersjjensen 6 місяців тому

      In scientific compute it is, but in game benchmarks it doesn't account for all that much as most of it is masked by cache hits.

    • @rockingfps5104
      @rockingfps5104 6 місяців тому

      In gaming I see much a good difference in lower latency, most cases where CPU is bottleneck.

  • @SAFFY7411
    @SAFFY7411 6 місяців тому +1

    I have another video idea and this mainly originates from previous reviews of the 5800X3D, where when the settings are lowered, the X3D performs much better in v-cache sensitive titles than non X3D CPUs. This was first seen with ACC where the 5800X3D was stonkingly fast at medium settings but when the quality is maxed out, the X3D benefits diminish a little. What would be interesting to see is if the same applies to the 7800X3D at settings other than epic. Worth investigating as gamers, especially sim racers won't be using max settings when racing to keep FPS as high as possible, but I'd be very curious if other games respond the same.

  • @techwandoUS
    @techwandoUS 7 місяців тому +16

    Great thumbnail.

  • @evankirschenmann4653
    @evankirschenmann4653 6 місяців тому

    I just noticed and LOVE the kuai kuai in the background of your video

  • @TrackmaniaKaiser
    @TrackmaniaKaiser 7 місяців тому +14

    Isn't the idea behind the 3D V-Cach that you get all the performance uplift of a faster memory and thats why the 3D V-Cach CPUs have a terrible DRAM-Frequency scaling, because they get all the performance even with slower memory?

    • @FTWGame0N
      @FTWGame0N 7 місяців тому +15

      Thats exactly it. With more CPUs increasing its cache teh faster memory wont give better results.
      Its also whats makes 3dvcache so good on vlaue point. Get the cpu get a cheap/midrange RAM and you are set. You also dont need an expensive MB as well as the power draw is low on those x3d cpus.

    • @vx9330
      @vx9330 7 місяців тому +5

      Yes and no. x3D processors still scale with memory performance, they are just less sensitive to it but are bandwidth capped due to Ryzen's memory controller system's design. This means the addition bandwidth of higher frequency DDR5 (6400+) can't be effectively utilized, where as 13th/14th gen can scale with 7000+ DDR5.

    • @26Guenter
      @26Guenter 7 місяців тому +4

      No, the V-cache allows more data to be stored closer to the CPU cores for quicker access, and not having access system memory which has a higher latency. You would have to use very fast memory to notice a performance uplift because the CPU doesn't have to access it as much.

    • @FTWGame0N
      @FTWGame0N 6 місяців тому +2

      @@26Guenter why start with a No and then proceed to say the exact thing you are saying no to.

    • @26Guenter
      @26Guenter 6 місяців тому

      @@FTWGame0N Because that's not what op is saying. He's saying the cache gives you a performance uplift because of a bandwidth increase similar to using faster memory which isn't correct.

  • @imJGott
    @imJGott 6 місяців тому

    Thank you for test ACC which is a cpu heavy sim. It’s appreciated from the sim community.

  • @jayfangRSA
    @jayfangRSA 7 місяців тому +7

    I do love the "for science" videos 😎

  • @AdnanKhan-be9dr
    @AdnanKhan-be9dr 6 місяців тому +1

    Highly appreciate the thumbnail. You guys are on top of your game.

  • @jeroenvdw
    @jeroenvdw 7 місяців тому +4

    Which timings were used on the RAM? Can make a big difference in determining what to buy

  • @Helifax19
    @Helifax19 6 місяців тому

    Loved the DDR4 vs DDR5 digestion of data!!! Kudos!

  • @rangersmith4652
    @rangersmith4652 7 місяців тому +3

    LGA1700 is unique among current platforms in that we can use it to make direct DDR4 / DDR5 comparisons. Makes one wonder what we could achieve IF it were possible to modify a Zen3 CPU and AM4 board to run DDR5 memory.

    • @olnnn
      @olnnn 6 місяців тому +1

      They did make the mobile 6000 zen3+ cpus that support ddr5 and lpddr5 so I it does technically exist (and also some zen2 mobile ones that support lpddr5) - not sure if there is an easy way to compare them clock for clock to the zen3 ddr4 5000 counterparts tho
      I guess if they ever end up with a lot of leftover dies maybe they could make an ultra budget am5 APU based on them or smth

    • @ismaelsoto9507
      @ismaelsoto9507 6 місяців тому

      @@olnnn Sadly not as easy to compare considering that Zen 3+ is only exclusive to APUs and it was quite clear that 16 MB of L3 Cache reduced Zen 3 performance significantly compared to regular Zen 3 32 MB. But it was enough to make them quite competitive with mobile Alder Lake CPUs.

    • @olnnn
      @olnnn 6 місяців тому

      @@ismaelsoto9507 Yeah the cut in half cache really hurts the AMD APUs in gaming so at least the older ones basically ended up being about roughly equivalent to the previous gen non-APU cpu with the same core count once paired with a dGPU in performance.

  • @IslamGhonaym
    @IslamGhonaym 6 місяців тому

    I gotta be honest this is the first time I made it in steve videos to the end 😂 while that is always the case in tim videos 😅 so good job on this one. Thx steve

  • @kohiek
    @kohiek 7 місяців тому +4

    It's good to know that DDR5 is really getting used. DDR gens at launch have been underwhelming.

  • @clintonkinchen661
    @clintonkinchen661 6 місяців тому +1

    I feel like an opportunity was missed in this video. I would liked to have seen how the 5800x3d performed in newer games vs the Intel parts with ddr5. Otherwise, what have we learned here about how the 5800x3d with ddr4 has aged? What I mean is, I bet the 5800x3d still holds its own despite the older ram. For those that bought years ago... that's good value. Thanks for all your hard work Steve.

  • @AlexHusTech
    @AlexHusTech 7 місяців тому +8

    *Wow didn't think DDR4 > 5 would makes that make performance*

    • @kesamek8537
      @kesamek8537 7 місяців тому

      On appallingly 'optimized' console ports it certainly seems to.

    • @andersjjensen
      @andersjjensen 6 місяців тому +2

      @@kesamek8537 People have been saying what you just said for the three decades I've been in the PC space. Every time games feature more realistic physics, NPC logic and environment interactivity people expect it to come at no computational cost, and hence it must be "badly optimized".

    • @THU31
      @THU31 6 місяців тому

      @@andersjjensen Starfield features none of those things and it runs like garbage. 😂

    • @andersjjensen
      @andersjjensen 6 місяців тому

      @@THU31 I didn't say that people were wrongly claiming that games that have what we already had, but runs slower, are badly optimized. Because those obviously are. If game A has all the newest stuff in abundance, and is still faster than game B, which has yester-years stuff, then clearly there is something unusually wrong with game B. And yes, Starfield is one of those.
      That said: Starfield has really shitty textures wrapped over pretty high polygon objects, which makes it look meh compared to what it actually is. And that's just plain stupid. My main sour grape with it, though, is that it is utterly brutal on the CPU for no apparent reason.

  • @mrcnorth7149
    @mrcnorth7149 7 місяців тому +1

    Thx for this video. Lots of great info from this channel

  • @dangingerich2559
    @dangingerich2559 7 місяців тому +3

    To me, this is telling me "more cache > more clocks" in most cases. As a 7800X3D owner, I'm fine with that.

    • @seibstershots9527
      @seibstershots9527 7 місяців тому

      I don’t think it ever was about clocks as two CPU’s with same core count and speed could be completely different. Instructions per a clock is a better way to look at it as this helps you understand how much can go through the cpu per a given clock. It’s like two different size rivers going the same speed if that helps.

  • @MinishMan
    @MinishMan 6 місяців тому

    Best thumbnail yet! Showcasing Steve's slow descent into madness through benchmarks

  • @eliadbu
    @eliadbu 7 місяців тому +6

    The big reason I spent quite a lot and went the DDR 5 kit (6000MHZ CL 36) back in early 2022 for my 12900K - it makes very large impact on performance for gaming,
    especially after I upgraded to RTX 4090.

    • @BenState
      @BenState 7 місяців тому

      did you watch the vid? 15-20% improvement max

    • @eliadbu
      @eliadbu 7 місяців тому +3

      ​@@BenStatemaybe you didn't, on average of 13 games 22% improvement, some games went high as 33%. In my opinion as I already allocated high budget for PC, it was worth it - 22% is what I expect gen on gen improvement for CPUs. Was it critical? Of course not, but as I'm building highend PC it was worth the high price back then.

    • @BenState
      @BenState 7 місяців тому

      @@eliadbu$/frame for you extra 25 frames was?

    • @eliadbu
      @eliadbu 7 місяців тому +1

      @@BenState if you compare just delta of the RAM it was around 200-250$ more than fast DDR 4 which would make it around 10$/frame.
      if you compare how much it added to the total system cost it would be probably less than 10% back then (as GPUs prices were high , add case + fans , psu cables, CPU + cooler, ssd etc), as I said for highend system I was building it made sense for budget system it wouldn't.

  • @upforellie
    @upforellie 6 місяців тому

    These thumbnails have been ART as of late. I almost look forward to them more than the reviews themselves

  • @nipa5961
    @nipa5961 7 місяців тому +48

    The 7800X3D is untouchable.

    • @poocruncha3266
      @poocruncha3266 6 місяців тому +4

      until the 9000x3d chips come out 😎....

    • @jjglaser
      @jjglaser 6 місяців тому +3

      @@poocruncha3266 for double the price of 7000 series.

    • @nipa5961
      @nipa5961 6 місяців тому +2

      ​@@poocruncha3266 I fear Intel will come short again. Maybe barely competing with the 7800X3D but 2 generations behind 9000X3D, again.

    • @poocruncha3266
      @poocruncha3266 6 місяців тому

      @@nipa5961 Intel knows that their fanboys will buy anything they put out

  • @deviousnate7238
    @deviousnate7238 7 місяців тому +1

    Whatever 5800x3d content you want to produce, I want to see. Even if it is just five minutes at the start and you move to a different topic, I'll be here for the whole thing!

  • @Anomnomnomaly
    @Anomnomnomaly 7 місяців тому +14

    I see no reason to upgrade my 5800X3D system for another 1-2yrs, paired with a 6900XT and 32GB 3600Mzh DDR4 it can play anything I throw at it @ 1440p with a more than acceptable framerate... throw in FSR and RSR and it's a more than capable system... Once I'm seeing new games that I play in a constant sub 100fps avg. I'll start looking at a rebuild... I could always throw in a GPU upgrade to push that out until late 2026 if I see a bargain 2nd hand card for sale.
    But when I build a new system these days, it's so I can get at least a 100% performance improvement over my previous one.
    Value for money has always been my primary concern over bragging rights... I'll buy top tier stuff... just not at top tier prices. My 6900XT was 60% of the retail price in 2022, my 5800X3D was actually less than the 3800X it replaced.
    I'll stick with a platform that has proven longevity... intel don't know the meaning of the word. AM4 had 6+yrs, as did AM3+ before it.

    • @jimdob6528
      @jimdob6528 7 місяців тому +2

      This is exactly what I am doing. I got a used 7900xt for 580 on fbmp but otherwise I don’t plan on building a new pc till am6 comes out in 4-5 years. I rock a 5700x3d but my brothers 5800x3d only seems to show a 4% increase to performance so I am happy with my build.

    • @Spentalei
      @Spentalei 7 місяців тому +3

      As long as you stick with a 6900XT upgrading the 5800X3D isn't really necessary unless you suddenly fancy 1080p 480fps gaming or something. The 6900XT isn't that high-end anymore nowadays. Mid-end is a bit insulting, so somewhere in between. And that's where the 5800X3D is still adequate.

    • @t1e6x12
      @t1e6x12 7 місяців тому +2

      Quite content with my 5700X and 7900XT :)

    • @Ferox63
      @Ferox63 6 місяців тому +1

      I'm running a similar 5800X3D/6800XT setup on X370. I may opt for one more GPU upgrade but lacking pci-e 4/5 is starting to limit my storage options. I may be tempted to upgrade to AM5 depending on what X870 brings to the table.

    • @Spentalei
      @Spentalei 6 місяців тому +1

      @@Ferox63 you can use gen 4 and 5 on a gen 3 slot/bus, seriously doubt youre going to notice a major difference in practice unless you bench it. Being stuck on a gen 3 storage is one of the worst reasons to upgrade imo.

  • @Noobsaucer
    @Noobsaucer 6 місяців тому

    Great video, as always! It's very interesting that in some cases ddr4 is enough of a bottleneck to effectively reduce the 12900k to a 12600k. In spider-man remastered the 2 chips are almost identical with ddr4, while there is a massive delta between them with ddr5. I wonder if we would have seen similar results for ddr4 vs ddr3 if we had the platform compatibility at the time to compare them head to head like this. Fascinating stuff.

  • @razorblazor571
    @razorblazor571 6 місяців тому +3

    I don't understand why people buy intel at all.

  • @SAFFY7411
    @SAFFY7411 6 місяців тому

    Once in a while this channel is able to come up with some legit top tier thumbnails, this is definitely one of those. 😁

  • @j3m638
    @j3m638 7 місяців тому +6

    My poor fast aging 9900K feels so abandoned already :( Looks like I had better finally upgrade to the 7800X3D at the end of this month then!

    • @Winnetou17
      @Winnetou17 7 місяців тому +1

      Weeps in 6700HQ. With 64 GB of 2133 CL 17!! RAM

    • @aberkae
      @aberkae 7 місяців тому +1

      I upgraded from 9900ks @ 5.1 ghz all cores to 7700x 5.65 all cores ghz to now 7800x3d stock. In terms of power/ throttling, long game sessions, minimum ambient temperature delta changes and small form factor build the 7800X3D is awesome especially with 50 watts of power use during average gaming sessions.

    • @BeatmasterAC
      @BeatmasterAC 7 місяців тому

      my Ryzen 3700X: "come again?"

    • @ShadowsOfThePast
      @ShadowsOfThePast 7 місяців тому +2

      Well..rocking an i5-6600k and a 6gb 1060... 😢

    • @DiogoSantos-qn8ev
      @DiogoSantos-qn8ev 7 місяців тому +2

      I5 4690K GTX 970 G1 Gaming 😢

  • @nicholasgiobbi279
    @nicholasgiobbi279 6 місяців тому

    Thanks for the work on this guys

  • @draxrdax7321
    @draxrdax7321 7 місяців тому +34

    DDR FOUR EVER! Not gonna download different RAM even when DDR FOURTY comes out!

    • @TheGreatTave
      @TheGreatTave 7 місяців тому +3

      But what about DDR Four Twenty??

    • @RecklessFables
      @RecklessFables 7 місяців тому +1

      ​@TheGreatTave reasonable. Very reasonable.

  • @isoluc11
    @isoluc11 6 місяців тому

    I am always interested in the X3D testing. my son has 5800X3D and i have a 7800X3D. Love the vid.

  • @manoftherainshorts9075
    @manoftherainshorts9075 7 місяців тому +26

    I've lost count of people telling me that DRAM speeds don't matter in CPU and overclocking memory is stupid.

    • @MaxIronsThird
      @MaxIronsThird 7 місяців тому +7

      For Intel sure, but Ryzen sees almost no improvement and the X3D parts see virtually no improvement at all.
      PS: this is also DOUBLING the frequency and getting 10% to 20% for 13th/14th gen and 15% to 25% for 12th Gen.

    • @alrizo1115
      @alrizo1115 7 місяців тому +4

      ​@@MaxIronsThird I remembered the time where AMD needing high freq RAM compared to Intel to work at best. It was about 5 years ago

    • @gizmo104drives7
      @gizmo104drives7 7 місяців тому +2

      The thing is.. in the right context that advice is actually good.. however it doesn't apply for every application... and THIS should be considered for all tech advice... there are very few universal truths

    • @M3N04
      @M3N04 7 місяців тому +2

      There is Difference but not that much here the difference is big because they are comparing DDR5 with DDR4 which is huge but if you overclocked like 200mhz you won't see big difference.

    • @yarost12
      @yarost12 7 місяців тому +2

      ​@@MaxIronsThirdthat's just false. 5200 to 6000 gives you around 15-20% of performance and tuning timings gives you another ~10%

  • @domneta
    @domneta 6 місяців тому

    Finally a nice and creative thumbnail!

  • @pseudonim1
    @pseudonim1 6 місяців тому +3

    Put 5800x3d on tests plese

  • @devonmoreau
    @devonmoreau 7 місяців тому

    awesome stuff as always guys, this is computer science at its finest ❤

  • @tarif7541
    @tarif7541 7 місяців тому +4

    what about productivity?

    • @Hardwareunboxed
      @Hardwareunboxed  7 місяців тому +23

      No time for work, we're gaming!

    • @tiay6269
      @tiay6269 7 місяців тому +7

      x3d is a gaming cpu not made for productivity

  • @root330i
    @root330i 6 місяців тому

    great video, just for the record on the CB2077 graph the 12600k delta is 22% not 29%

  • @RealLifeTech187
    @RealLifeTech187 7 місяців тому +24

    13th gen parts are missing in this comparison 🥁 LOL
    how are we supposed to know how they'd perform? 😂😂

    • @stennan
      @stennan 7 місяців тому +3

      Assume that they are 1-2% slower than 14th Gen? Doesn't change the conclusion of the video which was mainly about memory scaling on intel boards and 5800X3D

    • @Winnetou17
      @Winnetou17 7 місяців тому +4

      @@stennan He was joking :))

  • @thepunisher3160
    @thepunisher3160 6 місяців тому

    Best Pc content for this year easily and its only July

  • @TetraBF
    @TetraBF 7 місяців тому +5

    useless test we already know that stock ddr5 is faster than ddr4
    show us oc'd ddr4 v oc'd ddr5

  • @johntotten4872
    @johntotten4872 6 місяців тому

    IMHO I wish more focus would be put on the XX700k model Cpus. Usually they are within 5% of the XX900k and generally cost $125 to $150 less. 12th gen currently is an amazing price 12700kf is roughly $190 brand new, 12400f is under $120. Yes they are EOL but someone wanting a strong budget gaming PC these will still rock the house. Great video Steve.

  • @toddsimone7182
    @toddsimone7182 7 місяців тому +6

    A moment of silence for everyone who bought a DDR4 Intel board to save a few dollars.

    • @mattBLACKpunk
      @mattBLACKpunk 7 місяців тому

      Mehr, considering they would've had to buy a second kit of ddr5 a few years in to get any of those performance gains I'd say it was still the right decision.

    • @andersjjensen
      @andersjjensen 6 місяців тому

      @@mattBLACKpunk And a new board... and the DDR5 boards were also stupid expensive at the time.

  • @kenshirogenjuro873
    @kenshirogenjuro873 6 місяців тому

    Wow the gains with DDR5 are quite a bit larger than I would have guessed. Excellent testing.

  • @TrivialTax
    @TrivialTax 6 місяців тому

    Nice preview shot!

  • @Neopulse00
    @Neopulse00 6 місяців тому +1

    8:45 - Surprised seeing Homeworld 3 on there used as a benchmark. It is currently still brutal to run well.

    • @motmontheinternet
      @motmontheinternet 6 місяців тому

      Right? My 7600X can only get 44 fps on lowest settings including turning upscaling on while at minimum resolution.

  • @rx58000
    @rx58000 7 місяців тому +1

    Just purchased a 5700x3d and was searching for a video like this for a few weeks now. I haven't installed the CPU yet but happy to see the results

    • @LiLBitsDK
      @LiLBitsDK 7 місяців тому

      but why would you need that info? that chip ONLY runs DDR4...

    • @rx58000
      @rx58000 7 місяців тому

      @@LiLBitsDK I was looking to upgrade from ddr4 but the upgrade wasn’t feeling significant enough as I already have decent mobo and RAM so for me it was between spending 190 bucks to get a 5700x3d or to spend 450 bucks to get am5 + 7800x3d + ddr5 mem . If the 5700x3d wasn’t available I would have upgraded to am5 as the 5800x3d feels too expensive for around 280-300 $

    • @LiLBitsDK
      @LiLBitsDK 7 місяців тому +1

      @@rx58000 well the 7800X3D would run loops around the 5700X3D in more than just games/FPS... and 20+% upgrade is quite huge.

    • @rx58000
      @rx58000 7 місяців тому

      @@LiLBitsDK yes I agree I would have gone with 7800x3d if I had a faster gpu , but I only have a 4070 and I noticed low cpu utilization especially when using ray tracing. I really wanted the x3d cache more than the IPC gains

  • @projectc1rca048
    @projectc1rca048 6 місяців тому

    Great Work as always guys!

  • @joshman5217
    @joshman5217 7 місяців тому

    Fantastic video and content, thanks! In my mind, I knew DDR5 was going to be faster but this video helped better visualize just how much faster it is depending on th scenario.

  • @123Suffering456
    @123Suffering456 6 місяців тому

    Thanks! Didn't realise DDR5 made such a noticeable difference. I'm upgrading to a DDR5 platform with a new CPU this year anyway, so that's another reason to look forward to it.

  • @Novdid
    @Novdid 7 місяців тому

    HUB is the hardware channel that does more interesting stuff (the science type tests as Steve calls them) when hardware news are at a low. Well done!

  • @Kiyuja
    @Kiyuja 6 місяців тому

    I got a 5800X3D and I actually experimented with my DDR4 4400 Cl26 kit today and found that reducing the frequency even below the "sweet spot" 3600 is worth it if you get the cas lower. Just for kicks I left my MB to "auto" on all settings and it defaulted to 2666 CL20 and it improved my frametimes and fluidity a lot. I guess latency is crucial for X3D

  • @thatzaliasguy
    @thatzaliasguy 6 місяців тому +1

    So was this test done solely on Windows 11? Since Windows 11 utilizes a different CPU kernel scheduler, this _does_ affect CPU performance on last gen processors. Windows 11 also has a new memory allocator, which could be adversely affecting DDR4. I'm interested to see this same test done on a normal Windows 10 install vs Windows 11 install (and possibly Linux, if you guys feel like it).

  • @-Nicky
    @-Nicky 7 місяців тому

    That thumbnail... Very cool and funny. 😂

  • @ShadowKstr
    @ShadowKstr 7 місяців тому

    I did enjoy that. Thank you!

  • @amirrad3873
    @amirrad3873 7 місяців тому

    Thank you Steve, you make the most honest benchmarks 🙂

  • @MarioCRO
    @MarioCRO 7 місяців тому

    Interesting video, thanks for this.

  • @gravidar
    @gravidar 7 місяців тому +1

    cheers. I'm about due to update my 5900X so this is really helpful. will check out the 9800X3D when that's out but wont rule out Intel just yet

  • @anasevi9456
    @anasevi9456 6 місяців тому

    I grew up with old powermacs my father updated, a big fat L3 cache, as fast as you could get it was utterly pivotal in getting newer software to cope well with an old aging G4 system. Things certainly have changed, just generally the only actual drawback to a large L3 cache is manufacturing costs... and in Ryzen's case; suppressed clocks due to thermals. (most older cpus with large L3 and L4 caches had them placed adjacent, not on top of the cpu core/s.

  • @cs_mns
    @cs_mns 6 місяців тому

    One of my rigs has an a750 and it's given me no issues. Point was to deliver 4k 120hz which it does with no issue. Even had it in a razer egpu box using thunderbolt with no issues

  • @LOLHoneybadger
    @LOLHoneybadger 6 місяців тому +1

    Call it Twitter, don't feed Elon Musk's stupidity.
    It does make sense that the 5800X3D should be retested, because of the fact that a lot of people that will be looking at the 9000-series may be on AM4 still.
    Also, the price difference in said setups, to me, makes the 5800X3D the clear winner.

  • @Viewable11
    @Viewable11 6 місяців тому

    Thank you very much for your performance and quality evaluations.
    From your _FPS per price_ comparison of many CPUs used for gaming, I sorely missed the Ryzen 7500F and 7700. These two should be the two best _value_ picks among AMD CPUs. I also missed the the Intel 12400F since it is the best value pick from Intel CPUs.

  • @steve579
    @steve579 6 місяців тому

    Good to see a modern refresh of the ddr4 vs ddr5 comparison, but testing only 1080p is unhelpful with upgrade decision-making for real world use cases. I built a 12700k ddr4 system 2.5 years ago and game at 4k with a 3080ti. Been thinking to upgrade to ddr5 for a while but unsure if the benefits will be worth it. Please add higher resolution results in future content

  • @SpinningSandwich
    @SpinningSandwich 6 місяців тому

    Homeworld 3 helps highlight the real advantage of X3D, which is how much it improves the 1% lows. This is especially noticeable in the smoothness of esports.

  • @chrisbeauchamp5563
    @chrisbeauchamp5563 7 місяців тому

    Great video Steve. I'm 12th gen ddr 4 and really been wondering how much that has been affecting my system but estimating 10%. It's Alot more apparently.

  • @dianaalyssa8726
    @dianaalyssa8726 6 місяців тому +1

    I do wish I had the 3D vcache. 5800X3D totally makes sense if it's affordable and those are on the platform. 12900KF has been great with DDR5. I can't speak to DDR4 but it's an option for budget builds at least. It depended when you bought into the platform, 399 CPU, 170 Z board, 90 DDR5 replaced with a 48gb 115 kit in cost. I do wish I had done a 3700X build then upgraded to a 5800X3D but oh well.

  • @e21big
    @e21big 6 місяців тому

    Hey Steve, seeing your data, I think it's kind of interesting now that if you own a 12th Gen Intel you now have two paths to upgrade your system, either go all the way to 14th Gen (if you own 12600K or 12700K) with the same DDR4 system and 12th gen mobo or just upgrade your motherboard and memory to DDR5 and keep your CPU. I wonder if you would be interesting to making a video about that.
    I think most 12700K or 12600K owner would be on DDR4 platform (DDR5 was unreasonably expensive by the time these parts were out and your channel also heavily reccomended it yourself), and if upgrading your system can net you 10-30 percent performance increase it's definitely a worthwhile investment.

  • @blarzgh
    @blarzgh 6 місяців тому

    Big fan of those % increase figures in the graphs, they can be a little bit difficult to unpack sometimes lol

  • @shlooky
    @shlooky 6 місяців тому +1

    I bought a 12900K with a DDR 4 motherboard and ram, because DDR5 and DDR5 MB's were expensive. It's good to know that there is still room for more performance if I buy a Z790 + DDR5 memory.
    Thanks Steve for the insight

  • @slothnium
    @slothnium 6 місяців тому

    I bought a 5800X3D when it was on a massive sale. I couldn't be happier with it, even though I have a separate 13700K gaming PC.
    And now that the 13/14gen Intel i9/i7/i5 seem to have issues...