Insane Gaming Efficiency! AMD Ryzen 7 7800X3D Benchmark & Review 

AMD should have designed similar to Apple, same structures multiplied to add, dedicated cache per chip, double the 7800X3D to double the cores.

At this point you might as well just do one test for one game that includes 4k so people understand that it basically has no discernible/significant difference between CPUs at that resolution, so MAYBE they understand its a pointless metric for reviewing a CPU

Hey dude when listing power consumption per game it adds confusion when showing total system power. It's one part of a picture and not the rest. The rest of the picture is power consumption per frame. In the chart for SOTTR for instance a person could look at the bar for the 7800X3D and say it's less power efficient than the 7950X3D and that's not the case per frame. So, if you're going to show TOTAL power consumption which is in no way a reflection of the CPU alone, then on the same chart, maybe even the same bar, show power used divided by fps which would be frames per watt or watts per frame really. I guess you would divide the other way to give frames/watt. Either that or list the power consumption for the CPU only. Yes, it's harder to do, but it's better info. Basically your showing the jump in graphics power consumption since it's power consumption is much higher than the CPU's power consumption and I don't know why that's useful on a CPU review. The CPU will ALSO consume more power when it's pushed harder so it's consumption will fluctuate too, but not nearly the same. So, as the CPU can get better fps, in general it's going to look worse in the graphs your showing because GPU power goes up. You have cost per frame based on retail pricing for the CPUs and that's fine, but it's very different than power consumption data. To me it's useless data because wait 6 months and AMD CPUs will be less than retail.

​@@johndoh5182 no one cares about power consumption. I prefer total system power because it is tangible. Individually device power consumption is reduculous

It's unfortunate that you don't have access to CS2 to see the performance results. CS2 is built on the Source 2 engine and can utilize all 8 cores, which is a significant advantage over CS:GO, which is limited to a single core.

"13600K is better, myar!"

When did they stop screaming ?!?

​@@GewelReal NEANDERTHAL

They'll add a "doesn't produce enough room heating" category to level the field.

Damn, I've not cared about them in over a year.😊

Yup, cheaper power bills, cheaper PSU, and cheaper cooler.

Nailed it. Pair this with the ridiculously efficient 4080 and 4090, undervolt them and you'd have a machine that's a performance champ while coming in way below what you'd expect it to use.

Also means you don't have to buy as an expensive power supply. You can downspec and save some decent cash.

but this is only good when you look at gaming. Intel CPU's do more then just beeing good in gaming. They have proper Video Encoders/Decoders for H.264 and H.265 media / Video editing. The ones in AM5 CPU's don't work with video editing software.

Mora doesn’t know thermal throttling

A 7600 will save you about 150 Euros up front, and uses even less power. Do you really need the additional performance that the 7800X3D gives? The 7600 won't depreciate in value very much by the time the 8600 comes out, and then you can sell your 7600 and upgrade to an 8600 for a net upgrade cost of maybe 90 Euros, and then you'll probably have about the same level of performance as a 7800X3D in games. I would also suggest that the performance advantage of the 7800X3D is not likely to matter very much at all until AFTER the 8600 comes out. All that being said, the 7800X3D is an extremely impressive and competitive product, no doubt. It's just that CPU performance is so good these days, and there's so much competition between Intel and AMD with pricing, that almost any CPU will do a really excellent job for gaming. I definitely wouldn't look at the 7600 as a "low end" CPU. Even with an RTX 4090, a 7600 isn't really going to be holding it back that much when using graphics settings which make the most of what the graphics card can do. Obviously not many people would pair a 7600 with a 4090, but it would actually be fine. You'd be getting more frames per dollar than with any high-end CPU, that's for sure. I just assume that most people buying a 4090 are probably not caring about having the most frames per dollar :P A 7600 would make a lot of sense to pair with even a 7900 XTX, and would definitely be more than enough to pair with something like an RX 6800 XT or RTX 4070. If someone gets a 7800X3D to pair with a 6800 XT or RTX 4070, I would call that a bit of a waste of money on the CPU. Certainly not a big waste, but sub-optimal, in my view.

@@syncmonism My use case would be gaming and code compilation on Linux. Unfortunately the latter is a workload that shines a lot with more cores, Intel's Big-Little strategy would be giving me a substantial performance uplift there. As an Intel alternative under consideration, the 13600K would provide me good enough gaming performance with great code compilation performance at the cost of higher power consumption and a less attractive platform. As I come from a used 18-Haswell-EP Xeon, giving up on multi-core performance is not an option for me, but as I am cost sensitive, too, I face a bit of a dilemma here and might wait until AMD uses Big-Little with Zen 5 or Zen 6.

@@syncmonism I play on lowest settings maximum fps... So no, i would definetly go for a 7800X3D together with a 4070... If I was ever stupid to buy that piece of junk the 4070 is.

Yeah that 40€ difference in 4 years span will hit you hard

Wrong! PCs are on idle or close 50-90% of time unless you shut it down after your game right away. Idle power is horrible on AMD especially ZEN5. Intel 13900k (so horribly inefficient under load) consumes only 2w on idle and it is 30+w for AMD and entire AMD platform (MB, ram) also draws even more. 3D cached cpus consume even more than regular ones because cache draws power even under no load. Do your own research!

And it's price in Australia is very high. In Germany you can get it for 350$ and sometimes on sell-outs for less than 300

And they’ve done their absolute best to try and make it not seem worthy..

I’m so happy I got a 5800X3D on release date. My 3600 was getting dusty and AMD releasing such a beast for the old and dying AM4 platform is so cool.

I returned my 5600x for a 5800x3D and it was a good hump

@@danielfichtner667 not worth the cost

I'm still on AM4 platform, and am considering putting a 5800X3D in it. The longevity of AM4 greatly pleases me

@@desertfish74 That's where I'm at, deciding between going for a 5800X3D or jumping to AM5.

It's not just the power cost either, it's heat. When your system draws 150 watts less or something wild like that, that's 150 watts worth of heat you don't have to deal with. Ok... not all of that is heat, but still, it's going to lead to a system that runs cooler, lasts longer and sounds way quieter.

Well obviously it has 16 less cores

@@Batman-bh6vw I went 5800X3d last week. Had a ryzen 1700 before on a cheap b350 board with 16 gb ram and it runs great. Much better price performance ratio than buying AM5

He have been sassy lately ....

But.... but they ARE exciting graphs. I'm specifically here FOR the graphs (and the casual underhanded pokes at the PC industry). There's nothing to call me out on. I became a nerd at 13 years old when I bought a Commodore 64 and I'm not apologizing for it. Being a nerd pays the bills. Being a nerd is interesting and exciting. And Steve is my benchmark dealer!

Yo! What up my favorite up and coming RU-vidr!

Don't forget also cheaper 😅

That's why only no brain ppl buy Intel thinking Intel = best

Thanks Steve

On top of AMD (in general) being cheaper per unit of performance and less-power/less-heat per unit of performance, I would like to make one additional point that encourages (for certain people with certain types of application, anyway) the use of 7800X3D and 5800X3D over the AMD non-X3D parts. This excellent review and other good reviews do a sufficient job in arguing that the 7800X3D is typically much better on gaming than the equivalent non-X3D part (for example, 77% faster than similar 8-core 7700X on game Factorio - which might be near the top in such 3D V-Cache gaming speedup), and similarly the 5800X3D is typically much better on gaming than the non-X3D 5800X. The former has a cheap $450 price, and the latter has a dirt-cheap price right now for those who don't want to pay the price bump to upgrade to the AM5 platform. And I know that this and similar channels doing performance analysis are heavily focused on gaming, where we see the now-obvious (if it wasn't already obvious with the 5800X3D) advantage of 3D V-Cache for those gaming titles that choose to leverage that AMD advantage. But I would also encourage reviewers to hold back a bit before making the typical claim that (for instance) the 7800X3D is *only* useful for gaming, and not useful at all for productivity applications. The answer to that one is "it depends," and one of the things that the benefit (or not) of the extra 3D V-Cache depends on the most is the size of the physical memory used by the virtual-memory space of the productivity application. We can identify about 3 generic answers to the latter question: (a) quite small memory, (b) average memory, and (c) very large memory. Starting with option (b), which might be the typical situation, what we get for a typical productivity application with typical average memory size, is that there is a slightly lower clock rate on the 7800X3D (than the 7700X), so that's a slight negative, but that slight negative is typically completely offset by a slight speedup in the time it takes for a memory reference, because now a larger percentage of those memory references are hits on the L3 cache (that is larger for all cores on the 7800X3D). So for case (b), it's pretty much a wash. For case (c) on the other hand, we have a very large set of memory that the program might reference, and in that case, the larger set of memory words can certainly benefit more from the additional L3 cache. So in case (c) it is almost always true that the positive benefit of very large L3 cache more than offsets the slight detriment of a bit slower clock speed. In case (c) it's nearly always a win to have the 3D V-Cache over not having it. For the small price bump of $50 over the base price of $400 (for the 7700X) it is almost always worth the extra $50, especially when you take that relative to the entire cost of all the component parts. For case (a) we get a rather non-intuitive result with quite a small amount of memory being used by a productivity program. One might think, small memory, therefore small benefit. But that's not the case at all. What you actually get is the possibility, at least, of all of the memory completely fitting within the (now hugely expanded) L3 cache. If that's the case, then you're substituting an L3 cache hit (or maybe better if it's an L2 hit or an L1 hit) for each and every memory reference. This is a memory-reference speedup of what, maybe up to 5X? Depends on memory timing I guess. But even if only half your memory fits into L3 cache, that's still a huge speedup to the average memory reference. So all total, over the three possible extremes {a, b, c} of memory size (for a given program) we get {huge speedup, a wash, some speedup} as the overall impact on memory-access speed. There are exactly zero cases in which you get a slowdown as a result of having the 3D V-Cache. Furthermore, if you have some mix of {a, b, c} type programs running at once, you then get some weighted mix of {huge speedup, a wash, some speedup} to memory access. I argue that the 7800X3D is typically no worse and almost always faster than the 7700X thanks to its huge pile of 3D V-Cache. And for game programs it's typically even better yet, on average, and perhaps more so in the future as more game companies get their act together and take advantage of the extra L3 cache in any way they can. Therefore, from here on out, one should always prefer the 7800X3D over the 7700X, and prefer the 5800X3D over the 5800X, unless one is so broke that you literally can't afford the extra $50. Of course, the situation is a bit different for the 7900X3D and 7950X3D chips. For those chips, you get an oddball distribution of extra L3 cache (or not) across the two CCX complexes. Now for these chips, it's an even more complex decision as to whether the extra cores bump up your overall chip horsepower sufficiently much to offset the drain on overall chip horsepower associated with the (now) non-uniform availability of the extra L3 cache. The situation is sufficiently complex such that most people can't make a good decision without professional help. These two chips effectively turn the otherwise uniform AMD cores into what is effectively an AMD equivalent of the Intel hybrid architecture, excet that in this case, the bigger-L3-cache cores operate more-or-less like AMD Performance cores, and the regular-size-L3-cache cores operate more-or-less like AMD Efficiency cores. They need to take advantage of the revised Microsoft scheduler to get scheduled properly. But that new scheduler code was written for the Intel architecture, so end users may need to wait for AMD to help Microsoft tweak that code, before being good at such oddball scheduling. Therefore, in some use cases, it might be better to prefer the 7950X over the 7950X3D, and the 7900X over the 7900X3D, depending on workload characteristics. Neither the 7800X3D nor the 5800X3D suffer from that non-uniformity problem, such that their cores can be continue to be considered normal AMD uniform cores by the scheduler. AMD released the 7950X3D and 7900X3D a couple months early so as to sell as many of those that it could to consumers that just really needed that extra L3 cache along with the extra cores at the same time. Most gamers were savvy enough to wait for the better value-for-your-money, namely the 7800X3D. Let me finish up with a huge gripe. It seems like most OEM vendors refuse to sell you a *X3D part shipped in a factory build. HP, for instance, never offered a model with the 5800X3D (a shame, since at least their high-end cases are fairly decent). They forced you to buy a box with a 5800X, then replace that with a 5800X3D, and later sell your old chip on eBay, I guess. It is unclear whether the OEM computer vendors will continue this ridiculous obstinate behavior with the AMD 7800X3D. One wonders, do these OEMs want to continue selling pre-built towers, or not? If not, well that's sad, but if they don't want to make money, well that's their choice. If so, then why would they not offer for sale, what should be the most popular 7000 series CPU, namely the 7800X3D? I predict, whichever OEM vendor chooses not to sell the 7800X3D, short their stock, because that will be the first large OEM to go bankrupt.

@@Blafard666 500$ for 8 cores isnt cheap.

At the same fps?

TSMC nm cause

@@ajibsaladka Better FPS

@@-Ice_Cold- No, more cache is always more power efficient for memory heavy tasks like gaming. Intel 7 isn't that far behind TSMC 5nm.

You didn't see the 7950x being right along with it

I live in Florida so it is the opposite for me where as I have to use air conditioning to cool the room so electricity cost basically double when I play even in the winter.

as someone who borders you guys from the west, i can relate

@@rooster1012 oh no💀

Power efficiency is I would argue more important than raw performance. Especially being that all things considered, all these CPU’s can shoot out big numbers. Does it really matter if you get 150 vs 160 fps ? Don’t think so.

@@rooster1012 so then a cool efficient processor is good for you too. Keeps the room cool and minimizes the AC 😛

The 5800X3D is such a great chip. Worth it for any gamer on AM4

Do you have any sources for actually testing the game and not a command prompt benchmark? I've seen one person actually benchmark it and Intel was ahead... I want to find out if that's valid and / or if Intel drops off as it gets busier.

This benchmark is very misleading. On larger maps the gap is not that wide and if you run more than 60FPS, it's wasted. It will perform great, but not 50% better.

Totally bias review.

yeah it's been around that quite a lot on sale

@@Jinkle What are you on? Watch the video, in no way is it bias. Intel CPUs are better than Ryzen for productivity and the majority of other workloads outside of gaming. But if you want a CPU for pure gaming, the x3D CPUs are top dog, as proven in this video. Its not a hard concept to grasp. Intel is slower in gaming. Cope with it.

@@yeetus59 Ain't that weird ? Go back like 5 years ago to the first Ryzen, and read your comment again xD...

​​@@unholydonutsYeeeah.. I remember buying my 4870 for like 289 or something. Only the 4890 was better and if I remember right that was just 2 4870 chips on one board or something. Crazy considering I just paid 900 for a 7900 xtx.. Aside from the gpu priced it's kind of all relative. I just spent 2k building my new computer, 7800x3d and a 7900 xtx. My first computer in 2008 was an i7 920 and 4870, I spent around 1400 if I remember right. After inflation that's about equal. Also considering some things like ssds and monitors being cheaper is nice.

Of course it does well. In qhd and especially in 4k there is not a big difference between 5800x3d and the latest amd/intel processors.

@@dmytrosoboliev935 My i7-12700K still destroys to this day. I have it partially OC and doesn't go beyond 45 watts while gaming, rare watching 50 watts. +110fps worst case scenario on 2007 Crysis 1 at 720p, faster than the 5800X3D with the same framerate at 1080p, and i have 12 cores 😎. On winter, the CPU stays at 21-28watts with firefox or iddle. Just a Deepcool Assassin 3 and a Corsair 5000D, no liquid cooling needed. 5800X3D is good from someone upgrading from an older Ryzen 5, but it wasn't the best of the best when it was launched, just good platform value for only gaming.

@@saricubra2867 Yeah but intel was never mentioned by the OP, are you jelaous?

@@saricubra2867 Crysis is all about one powerful core…… you cherry picked the worst game to compare it too. lol Though not many people here are looking to max out crysis at 720P…… heck I was maxing out crysis 10 years ago on my 3570K @4.8 ghz @1080P. It not a hard game to run as long as you have a powerful cpu core and decent gpu. I’ll stick with my 5800X3D that only cost me $300 to upgrade with the same board and ram I bought 5 years ago. The 5800x3D is a beast and one of the cheapest upgrade path I ever had since I started pc gaming. It got old having to upgrade my board with Intel all the time. Heck I’ll probably be able to skip AM5 by the time I need to upgrade again. Saved all that money I would of spent on upgrading and throw it into a big GPU and monitor upgrade. :)

@@HifeMan "One powerful core" I have a 2008 patch and 64 bit, it uses 4 threads, Core 1 100%, core 2 around 60%, core 3 and core 4 around 33%. Not true, it's a hard game to run to this day, even the remaster. Draw distance, the physics, the AI (Delta difficulty is so hard and smart), the amount of stuff on the screen, all of these things make it a huge CPU hog many years latter.

To be fair it´s not quite as impressive as the 5800X3D was, but yeah, still great performance, especially effeciency wise.

@@NegerKim true, the 13700K still seems like a better, while offering much much better productivity performance

@@NegerKim well we're hitting the diminishing returns of cache. so of course we're not gonna see more improvement compared to the 5000series.

A lot of us use AM4...sooo...nah. Not even worth the money to go AM5 just to get the CPU...I'll wait for 9800X3D instead.

I think it's performance is comparable to 5800X3D. What did 58X3D did, was showed us just how insane you can bump the performance with "just" slapping 3D-Vcache on the CPU. But perf. bump is comparable. I'm happy with 58X3D upgrading from 3600 non X.

Yeah, it's the reason 3D chips crank on EFT too

They benchmark of demand, and games take ages to benchmark. If you support them on float plane or patron they might be more likely to take game requests. I would be interested in seeing these kind of games benchmarked in a unique video.

Factorio is kinda like that, right?

5800x3d is killer CPU in HOI4, so I expect 7800x3d being even better! I can plau Hoi4 with 5800x3d really late and game still runs fast (64gb of system memory also helps…)

The factorio benchmark is a simulation speed benchmark, though stellaris or civ would be interesting as well. the x3d has been great for real time sim as well with ACC in this benchmark and I think a previous HUB vid benchmarks the 7950X3D in flight simulator and it also crushed.

Same

HUB for blue graphs and advice, GN for deep dive into technicalities.

@@beachslap7359 most definitely

Perfect for a SFF Gaming and HTPC. Go as small as your chosen GPU will let you. Good case fans and proper case ventilation make all the difference. =)

13900K runs just fine with Z690 and DDR4.

@@SixDasher while practically losing to 5800x3d - a 13600k will beat 13900k if the former has ddr5 while latter doesn't

Not buying lots of case fans? If you're shopping for a 7800x3D then why on Earth wouldn't you at least completely populate your case with fans? There are good fans for quite cheap.

I noticed that too. I got my 5800x3D for $320 with same day delivery from Amazon…. That’s the standard price not $450 listed. I won’t bash them though the reviews are always in-depth and fair

Yeah, 5800X3D is about 333 USD here as well. Meaning Romania.

380 USD in Denmark for a new 5800X3D currently.

$400 USD in Norway

299€ in Germany. Hopefully the price will sink for the rest soon as well!

It depends where you live. In here 5800x3d is still 450€…..

So are we going to price the 7800x3D at $300 also because in the future it's price will drop to that also when 8000 releases?

@@tomstech4390 So you just want to ignore that 5800x3d's price dropped $150?

@@tomstech4390 PC parts always drop in price eventually, so when comparing perf/$ with older gen parts HUB always uses current pricing. This is especially relevant for GPUs where MSRP has become completely meaningless. Your argument is dumb as rocks since you can't buy Ryzen 8000 today. This is a ---> Day 1 Review

@@tomstech4390 This is grossly inaccurate data. He said at the time of filming and AMD site lists at $350. $299 6 months ago for me. Everyone has different pricing, but he USD in the charts so he needs to use USD current pricing or the chart is useless.

Right, but it sounds like something which could be a bit dangerous if others try to do it, no?

That is a really good point

​@@ht3k agreed

It also means if you limited the framerate the power would drop even more. I also agree at FPS/W.

On the contrary, I'd say total system consumption _does_ show how efficient this thing is precisely for the reason you say. "This chip is so powerful that it beats Intel's offerings, driving up how much power the GPU sucks down compared to Intel." _"But,_ the chip is _so_ efficient that you still have a net reduction in power usage even with the 4090 going harder!" Truly incredible work, AMD.

Definitely. The units of such benchmark would be joules per frame (or calories per frame ).

To be fair...coal powered plant eats intel for efficiency.

Where have you been for the past 3 years lol (minus the time between 12th gen launch and 5800X3D launch)

@@GewelReal up intels you know what

All these tests are just bullshit. Reviewers don't bother to make elementary CPU voltage settings in BIOS and that's why they have such crazy power consumption. This is me now about 13900K. On my system, it consumes 246W at 100% load in Cinebench R23 and shows a result of 40400.

@@Macoved It's still using almost 250W? Damn

Just a clarification here, we don't use a replay for testing Assetto Corsa Competizione, it's in game at the start of the race, 50 AI drivers, starting position is back of the grind.

@@Hardwareunboxed Please could you include more SIMS, RTS, MOBA and ARPG games? Like Cities: Skylines, Age of Empires 4, Starcraft 2, Dota 2, League of Legends, Diablo 4 and Path of Exile? Thank you very much!

@@lagarttemido only Diablo 4, Cities: Skylines, Age of Empires 4 are the reasonable games that would be included in the benchmark, most cpus can run those other old ass games.

@@Hardwareunboxed Then I misunderstood you when you added ACC years ago. But even so, AI had their own racing line, nothing special, and CPU is not hammered like when you're online. Thanks for clarification.

That was the thing I was most impressed with as well, with the efficiency being a close second. V-cache is becoming less and less of a compromise in certain workloads, which makes X3D chips all the more compelling.

Comparing the x3D vs the 7700x and the 5800x3d is very impressive

@@benjaminoechsli1941 Yes and the efficiency in power is amazing too

@@tonydalimata Best hardware review guys on the internet.

7950x3d would be faster slightly as it has the higher FJMAX speed of 5.25 vs 5.05 of the 7800xd.

Yes. 7950x3d is sligthly faster, because it uses higher binned CPU chips that 7800x3d. But not worth the price though if you only play games. If you have too much money. Buy 7950x3d and switch of the second cpu chip to get the best gaming CPU in the world at this moment :)

@@haukikannel Regrettably, I have too much money but will probably simply use Process Lasso to do the same thing...

You can just go back and watch the 7950X3D review. He shows the results for both configurations and it's the same suite of benchmarks as here (except the chance to ACCs settings).

I'd like to see a video on this, too. I see a lot of arguments saying that it's not advantageous to disable CCD1 on a 7950x3D. Instead, the arguments say it's better to use setting the preferred CCD in bios, and then using Process Lasso to assure that non-cache-benefiting apps go to CCD1 by setting affinity. Me? I'm an owner of a 7950x3d and I'm not clear on what route is best. I understand optimizations are per game, but I don't want to have to be resetting my machine between app runs. I'd rather maximize the machine for gaming, but I'm not clear if these arguments of leaving the CCD1 on are valid or not. Seems easier to disable it, but I think we might be losing out on the opportunity for background services and utility apps to run on the second CCD, or setting a game that doesn't care about cache (like CS:GO) to CCD1 to take advantage of the higher clock speed of that CCD.

if you already have the 5800x3d there's no reason to upgrade really. Unless you just want the fastest gaming CPU.

@@PhilippinesLife101 Very true! But I plan to build a new PC soon, and with the 7800X3D out now it makes no sense for me to still build a AM4 system.

@@nipa5961 Very nice plus you can upgrade to 8000 or 9000 series in the future. Very good upgrade path for the future.

@@PhilippinesLife101 yes! I wonder what Intel is gonna do now? Their platform has no future upgrade, their CPUs suck 3x more power and have lower performance. Intel probably should drop their CPU prices to half to compensate for all the bad things.

@@nipa5961 Depends on how soon you plan to upgrade. Honestly with a 5800X3D you could probably get away with staying on that until AM6 rolls around, or at least until AM5 becomes dirt cheap.

I have 5800x3d in HOI4 and it is fast! No worries when the time goes to 1945 or later. The games still is fast!

It's been sitting at $300 at microcenter for months. Kinda waiting for it to go down further but already a great deal.

It's time for intel to start offering some 3DVCACHE cache variants of their KS cpu lineup

@Donuts LMAO, that's pretty funny. Can you at least try to be realistic and say $250 or below?

Right here in est of europe is 320 euro also, yes that change a lot the final cost, in case of whole platform it will be 2.44$/frame(not 3,74$ like in his graph) for whole 5800x3d platform versus 3,84$/frame for 7800x3d platform. Also in the case of cpu only will be 1,69$ /frame (not 2,52$ like in his graph)versus 2,03$ for 7800x3d (also i think i heard in europe it's already listed at 500 euro or more not at 450$ like in US:D so that change even more Also is become more interesting if you already have the am4 platform you can only switch the cpu to 5800x3d versus change the whole platform to am5 and 7800x3. And in this case if we calculate the ratio "Price/Performance gained" it,s look even worse: so 850$ whole platform vs 320$ for only 5800x3d cpu, which in ratio Price/Performance gained is translated like this: You will have to pay 165.63% more money while you will gain only 19,45 % more performance on average 12 games at 1080p??? that look like a very bad ratio if u ask me lol. Mabe it's should worth more to change the whole platform if the 7800x3d should be something like 80% more better than 5800x3d, so i can pay 165.63% more money for 80% more performance?:D but not for only 19,45 % more performance (cause conform steve graph 7800x3d is 19,45% more better than 5800x3d 12 games average at 1080p) So from the point of view of only change the cpu to 5800x3d versus change the whole platform to 7800x3d is not worth at all look very bad ratio price/performance gained:D But....On the other side ...this ratio..it will look more better if you sell your older platform with let say half of the price of new platform and then add only difference?, so now will be 425$ vs 320$, which in ratio Price/Performance gained is translated like this: You will have to pay 32,81% more money while you will gain 19,45 % more performance. witch look more better

@@lldjslim im really surprised Intel is not offering V cache .. after the 5800x3d crushing there 12900k ( while only in certain games ) coupled with there 2 year socket life span are they actively trying to push themselves out of the market .. these 7800x3d are going to fly off the shelves watch AM5 boards now start selling like hot cakes .. Its like AMD is saying and proving yeah v cache is the future and Intel is saying NAH

In real large maps, performance gap isn't that big. It's a very misleading test.

I got the very similar B650 Elite AX from them in my rig. Sweet spot for gaming indeed. No need to blow chunks on more if just gaming.

This CPU will run completely fine and at the same speed even on cheapest b620 boards because it has so low power draw and can't be overclocked anyway

For ITX build, we have no choice but to spend more than 200 for B or X board.😅 Oh well.

@@damara2268 I thought the b620 is not compatible with any chips over 60w, read an article yesterday about it. Bit of a shock as I thought they would be able to run all the chips but with cut down motherboard features, but apparently not.

You also don't need X570/B550 mobo for 5800X3D. It works fine on something like B450 Tomahawk which is $110.

Agreed, but looking around at different regions around the world those prices don't quite line up. Given that, I can see why they used the more conservative price point listed for the 5800x3d

HAHAHAHA

Who is what guy again? Is he still making videos?

H€CK YES :P I'll take your entire stock! Already lined up pcpartpicker with Flare X5 sticks under NH-C14S with 12V HPWR supply and a GPU later on when buying makes sense.

This sounds out of topic. We are talking about Desktop CPUs here, not Mobile. Though like console stations, handheld devices like Steamdeck use custom chips.

@@PainterVierax The progress that AMD makes towards power efficiency will translate to APU's. AMD's chips are the go-to when it comes to handheld PC gaming.

@@molsen7549 again, you're comparing desktop designs with embedded/mobile/custom ones. The overall architecture might be the same but the power/dissipation targets are totally opposite. And x86 in general is clearly not designed for low power efficiency. This ISA has many pros but in this subject it's heavily beaten by ARM and RiscV.

With these CPU results I think we should go away with 6 core 3D cache if a lower priced CPU has an 8 core 3D cache. The 7900x3D would be a wonderful CPU if it had 8 core 3D cach with a 4 core boost CPU.

Getting to 4.8GHz this gen (AMD seems to have implied in interviews that the 5800X3D had basically a prototype of V-cache, while this is the true Gen 1) is very encouraging. Assuming the 8000 series is mobile-only as they've done the past few years, the 9800X3D might be the chip to get me upgrading my AM4 rig.

AM5 is 8 core @t 4nm TSMC AM5c is 12 core @t 3nm TSMC both will come out next year one at the start at the other at the end.

As long as they stick to a single CCD (or have the 3D cache on BOTH CCDs) instead of the hybrid model they use now, I'll buy it. I hate schedulers. I want it to work out of the box, no fuss, no janky software implementation, and no stupid process lasso...

That's the one thing that could tip the scale back to the 13700K for me...

@@metacob Not a bad option, i went with the 13700K, no regrets, no insane power usage in gaming like some people claim.

@@ZoneXV The best way to not get stressed out over power usage is to simply not measure it. This is true for just about every aspect of your PCs performance. Do the research, spec out your build according to how reported performance meets your needs, and use your PC. Unless while using it you sense a performance issue, stop measuring. Constant measurement of PC performance and seeking to incrementally improve it leads to never being content. There's more than enough in life to have discontent about already.

But it's not upgradable, PPL on 2600x and 2700x had option of 5800x3d, so if anyone buy this today they will save alot in 25-26 or even later but if you are richi rich then don't matter

@@RahulJha-mc6nz True, Meteor Lake will be LGA1851 afaik, so while you can keep your mainboard for at least another upgrade with the 7800X3D, for the 13700 that's not happening