Kind of, yea. It sets the precedent for how good the new generation is compared to the older one and whether which generation is the better buy. @@GRIGGINS1
Everybody so shocked by Standing Steve's new results that they missed the most important news, HU got themselves a new device for measuring EPS 12V power instead of total power consumption! It's always good to see more testing equipment for even better data
Great summary. I fully missed the part about the missing cooler vs increased price and also the 65W vs 105W is something I didn't consider in my video at all. This kind of re-cap is really cool to see after multiple reviews are out!
Daniel Owen is the name, and numbers are his game! 😁 seriously though the guy is a math teacher. I've been following his channel for about year and a half now and he does like to repeat himself at times, BUT he is never wrong. so when Steve said "school is in session" it literally was. 😁😁😁
@@inkredebilchina9699 That's why I like Daniel. He's probably the most independent unbiased techtuber out there that neither takes cheap shots or overwhelmingly praises any company or product. It's all numbers with him and I like that he tests at tons of different settings to show people almost _NO ONE_ games at 1080p medium with a 4090.
still one of the best review, 5-10% better efficiency is hard to show even if real bro. Onlything, i wish some one try the stability of 7700x vs 9700x for several day tasks, i feel like my 5700x has freeze if i work more than 48 hours stray with no restart, i was wondering if its fixed on the new gen.
I can certainly see the attraction of playing basketball with a cricket bat, but for basketball, a liberal application of billiard cue chalk to the elbows is decently effective and slightly less obvious than wielding a big MF of a cricket bat. 😉
@-firefly7931 Haha, me too. I jumped on this opportunity because scared people without comprehension or skills to change a few bios settings, have been returning perfectly good 13 and 14 gen CPU's. I just got a 13700k open box from Amazon for $235. it's never been used. 😂 and now it has a five-year warranty. I have been laughing at broke gamers trying to lecture us about computers and what chips to buy. It's been rather entertaining.
@-firefly7931 RMA Inbound lmfao, these are old, voltages have been fixed but i9's all the way to server, and mobile haven't been rip to this man. Lmao y'all get a load of this guy xD
Intels really gone down the route of productivity only with its tiny core design. They could totally release a 14700k 3d but don't really care about such a small market. Xeon is most of their business at this point. Same with amd server cpus which is painfully obvious here.
@@lunawense6288Yes intel is playing in the big leagues while AMD is in the minors… that’s why intel has a $32 billion dollar loss, lowest stock prices in a decade, and has the CEO quoting bible scripture on twitter…
Humans evolved to be dynamic, to stand and move 🏃, not to sit on their widening rears and vegetate! 😄Chairs are mostly unnecessary unless you need to rest because you're tired, or are injured or disabled.
@@stennan did u guys know hardware unboxed was the one to actually help daniel owens when he had less that 50k subscribers more than a year ago? yep, HU helped daniel get in contact with corsair for gaming equipment includng a corsair 5000t case. I used to watch him when he didn't make 20+ minute videos everytime.
Kudos to the positive Call Out for Daniel Owen. He really made it clear to me how little efficiency gain there was on these by taking the lead on the 65W comparisons. With those 65W comparisons, and the street prices of the 7700 non x, as well as the 7600 non x and 7500f, AMD has managed a massive marketing fail. The price to performance and even performance per watt of the Zen 5(%) is abysmal. No other word to describe it, and that's without even the setup dramas many channels had.
This right here. I love all the pats on the back that Steve gave out, but this was such a crucial point that I was surprised nobody touched on other than Daniel Owen. In a sense, I almost wonder if his being "new to the scene" really has helped him keep the general consumer's mindset with purchasing these products, even if Daniel's now actually receiving review samples for things.
Can't wait for the 9800X3D to offer the same gaming performance at like 3 watts less power for a significantly higher price. You very likely did the right thing just getting a 7800X3D now. You also dodged issues like defective processors. Can't forget that despite AMD delaying the launch to retest for processors that aren't meeting spec that they managed to send Gamers Nexus a 9600X that didn't work.
"And now I'm rambling because I'm too tired." I'm pretty sure you were rambling for the last 5 minutes of the video, not that anything said was wrong mind you.
Why would anyone have some sort of stupid brand loyalty to a PC part company? I want the best performance and whoever gets that gets my money. I don't care what the logo on the box says. You fanboys are insane...
A lot of reasons to have brand loyalty. Previous history with no failing parts, good customer service, company ethics, open source/locked down, compatibility, what your used to.
@@benclimo461 I totally agree with this, but I guess I should have specified that I was referring to those people who shit on other brands because it's not THEIR brand. That kind of brand loyalty drives me nuts.
Welcome to the sad reality of tech. Just full of disgusting fanboys who will contradict them self when needed. Remember when 40 series released with a modest 10-15% performance increase but absolutely insane power efficiency? What did every amd fanboy say? "haha power efficiency means nothing its not 30+% more performance so trash cards hahaha" Now what are they saying when 9000 series has literally zero performance uplift to 7000 series? "well its 40% more power efficient so these cpus are actually amazing you just dont understand" If I can give you advice. Don't try to ever find logic or understand fanboys and brand loyalty in tech. It simply has no logic and the people who are hard loyal to one brand are the most brainless two faced people you will ever speak to. I was intel cpu for years but the 7800x3d is absolutely insane so i swapped to that. I have been nivida for years and I have looked at amd and even intel but the cards are just trash so I have stuck on nivida. I have no brand loyalty I just want the best gaming experience so I will swap as needed. Would be nice if everyone followed this because not only is it better for your self it also makes a more healthy market where companies can't release trash and get away with it because of "loyal customers" who will buy it regardless of how bad it is. 9000 series is literally just AMD trying to sell you an undervolted 7000 series simple as lol.
@@donkeymoo1581 "9000 series is literally just AMD trying to sell you an undervolted 7000 series" Different die size, new instructions, but this is more of an architecture refresh than anything.
Dude. You're missing the point. After the Intel D.E.I. self-destruct, we're left with AMD as the only game in town. And they now start resting on their laurels. Looks as if AMD has reasonably good hardware, but cheaped out on the software side, in particular Windows drivers for gaming. That's the simplest explanation for the gaming flop of Zen 5, and of the inconsistent results were some games randomly show a great improvement. Lack of competition will do that, makes companies lazy and cheap with their investments.
While I agree, people should be getting AMD chips right now to help even out the market. If we can get back to a 50/50 market share, it'll be better for consumers. We'll get better performance at lower prices because the competition will be there. Intel having the majority of the market for so long only allows them to do things that are worse for their customers and get away with it. I will admit that the 9700X was a bit disappointing and that a 7700X would be a better upgrade due to the cost alone.
@@franzlyonheart4362 No, that is not the 'simplest explanation' at all. Zen 5 isn't good for gaming because all the effort was put towards the kind of architectural changes that weren't gonna help gaming that much. There's no magic CPU software that would be limiting gaming here. You seem to think these are like GPU's and the important GPU driver, but it's not the same sort of situation. That said, there is a potential some years from now that game developers start optimizing more around super wide architectures like this, getting more from it. But that wont happen within the next year or anything.
its tempting to get it but we need a GPU that is not $1000 to pair with it. Nvidia is obviously overpriced but AMD really gave no option for us either and its really disappointing. I would have easily built a team RED computer with this CPU, if they had better GPU/ better price GPU to begin with.
@@happybuggy1582 Or you can have it (specially with that microcentee free Ram) and don't change it in the next 5 to 6 years, until the Zen6 X3D part lowers its price.
Aye. Like looking at your kid or your favourite player taking drugs and/or saying really stupid stuff in public, with an attitude. Cringe cringe cringe everywhere. I guess except Apple. Which I hate for completely other reasons.
@@Winnetou17 Apple is now slowly locking the MacOS more and more (read: boiling the frogs). They are much smarter at it than how Microsoft tried (and failed). Forced notarizations, certifications not just for malware but design, nagging screens for non-appstore programs, even Valve has enough and doesn't want to push it on their steam game devs. It's getting bad, but they are the best CPU designers right now... ehhhh
It’s the same nonsense they pulled with Graphics cards. “Over-naming” the new part to gain a favourable comparison on price but not a true comparison for performance.
@@Ttallonn They will obviously release a 9700 at 45W that people will vastly prefer over the 9700X because a 65-80W BPO will effectively make it a 9700X. Just like the 7600/7700 is a 7600X/7700X when you give it the same power, or just a little more to overcome the less good binning.
Always felt a bit guilty for upgrading to a 7700 at the end of last year, instead of waiting for the 9000 series, glad I did it. I'll wait a few years and, when this platform is at the end of its cycle, I'll get the best CPU deal to prolong it's life for a little while.
I'm looking at upgrading from a 3600. I was hoping the 9700x would be the best option, but I might just end up going with a 7700. Now if I can only get someone to do cooler testing of all the coolers I'm considering using.
As a user of R7 7700 (non-X), this is the review I really wanted to watch. I'm using 7700 becuase it has the 65W TDP, and it is very intersting to see 9700X showing similar power behavior, yet performing better than 7700. However, since it is clear that 7700X3D uses less power than any other CPUs on the charts, if I intend to upgrade my gaming rig, it would be better to wait for 9800X3D review, and choose after the review. Actually It was my original plan anyway, but happy to confrim it was a tight one. Thanks steve.
Lol.. from 7700 to 7800x3D is like 20% at best case scenario I imagine 9800x3D will be like 22%more or less Would be a wise choice to wait for 2026 release of 11700x (or whatever they will call it) Am sure it's will be faster, maybe even will be 10 Cores instead of 8, who knows Lol
Tom's Hardware is one of the outlets that found a bigger than typical gain for Zen 5 in games, and also a bigger than typical gain for PBO. The biggest factor that people are missing with that site's reviews is that they test with stock memory frequencies only for non-PBO results, and they only enable EXPO for their PBO results. This means the Zen 4 processors use DDR5-5200 and the Zen 5 processors use DDR5-5600, both with JEDEC timings, and then the PBO results used DDR5-6000 with EXPO timings. Steve has shown before that Ryzen chips tend to scale well with memory speed, so the boost to memory speed between architectures is pulling some of the weight in that review, while someone who tests with the same memory speed across all chips will eliminate that difference.
Well of course there are big gains when testing with PBO then. Memory speed is very important for AMD, testing at stock speeds is like tying one of its hands behind its back.
Why even make a not like-for-like comparison like that? We are comparing CPUs, not memory kits. Plus, it clearly misleads the audience who seemingly almost never get it right and always interpret these results incorrectly.
@@w7bUxhwRYUo8Lv Using EXPO or XMP officially voids your warranty (though unofficially they will never ask), so Tom's Hardware and some other reviewers such as Anandtech stick to the max officially supported speed profiles, and tom's only uses EXPO for PBO or OC results. This info is hidden in their test system table that most readers won't look at, so a lot of people get the wrong idea looking at their benchmarks. This maybe could've been mitigated if they showed the OC/PBO results for the Ryzen 7000 parts too, but they only showed them for the new 9000 parts which is extra misleading and not apples to apples at all.
So when AMD claimed that that the 9700X was e.g. 31% better in HZD than the 14700K, they must have been limiting both to 65W, right? That's all I can think of.
Thanks for doing this "Re-review" Steve! I was one of those that was left confused by the mixed coverage and this 2nd video was excellent in helping to clear up many of my questions. Very much appreciated!
Great video Steve, I was one who wanted to see PBO across your suite of games tested based on what Der Bauer saw, It was more for interest rather than copium 🙂 thanks for putting the time in to revisit it
With all the recent Intel woes, lets hope that they can deliver something compelling with Arrow Lake when it drops. Let's be honest, praying for the demise of one company because you like the other is pretty stupid, we need both Intel and AMD to be one upping each other so we as consumers can benefit.
I tend to go for the chip that is the best all around processor for all workloads. 5 years ago when I built my system, that was the 3900x (3950x wasn't out yet and it was too expensive anyway). Now, when I need to upgrade my system, it looks like Intel will probably take that spot with the 285k or 265k.
Arrow lake seems promising so far, TSMC 3nm and Intels new 20A node will be a massive improvement over their current node. Mainly power consumption should be a lot better. Single threaded performance are rumored to be 3% and multi threaded performance 14% better. And Intel have said they will copy 3d v cache as well, which is a good thing.
Excellent re-review! Hats of for the honesty on the relevant parts. The thoroughness on the parts people needed schooling in. And the positiv shoot outs. Love the work and approach you and Tim go with. It's Mega! 🙌
It's really funny to pay so much attention to typical marketing slogans. It's okay when it's written in the comments, but when the authors of the channel behave like this, all that's left is to put a dislike.
This deserves just as much mockery as people give Intel for the same crap. Unfortunately, instead, you see a whole bunch of "these are not gaming CPUs!!!! only X3D are!!" and "do people not understand that data centers are a bigger customer for AMD than gamers??", "they made the right decisions from a financial PoV." Yes, these ARE meant to be gaming chips, actually. And no, they are making the wrong decision that could immensely harm their company long term, especially once they lose the AMD cult like Intel has largely lost theirs - through stagnation - and lose market share. If it's even a "decision" and not a simple failure to achieve performance targets, which it most likely is. You decide which one is worse.
@@nameless-user Curve Optimizer. In this case tuning the CO means adjusting the offset so that you lower the voltage while being able to maintain the same or higher clock speeds.
Yup in gpu segment it's because of drivers. AMD ages better everyone knows it. It's true. With each driver i got frame pacing improved and Smart portion of fps i got every time.
Not at all true. Lots of people cheer for something to be bad nowadays, if it's on the current 'bad team' side of things. Like, people are loving the Intel drama right now and I dont think most people who dont actually own a 13/14900k actually want to see Intel fix the issue, cuz then they wouldn't be the big villains anymore.
It's more like they brought a tennis ball to a cricket match. But since Intel left their bat at home, it's a moot point. ... I think this analogy might be a bit overstretched.
I think that the software development oriented performance doesn't really make that much of a difference for people targeting the price range of the 9700X and 9600X. I've been running a 5700X for quite a while, and I honestly have no reason to complain. Compile times don't really bother me, the performance is good enough to work on Android studios while having multiple apps and background tasks running. If anything I feel more compelled to use my money to buy a 4070Ti Super, or see what NVIDIA cooks up for the RTX5070. If you are in a point where you are working on large enough projects that a 5700X or even 7700X can't handle, you are probably making enough money to justify investing on a 9950X. This last 2 generations have proven to me that unless your focus is gaming, you don't need to keep upgrading CPUs every 4/5 years
Long live gaming unboxed...uh...oh...hardware unboxed!😁 Watching this at 04 a.m. while pedalling indoors on my road bike on turbo trainer. Good times! Thanks Steve!
I was in the camp that you painted Ryzen 9000 in an undeservedly uncharitable light with your initial review, but I gotta admit after watching this that you've backed up your original stance pretty much irrefutably. With how varied the takes on these initial two releases were, it was quite easy as a consumer to lean towards the more favorable assessments out there - so thank you for having the integrity to tell it like it really is. 👍 You guys are doing good work.
I wasn't, but did bemoan the lack of Curve shaper. But after seeing skatterbench's video on it, it seems quite difficult to grasp. So I understand now why it wasn't added.
I went to the AMD subreddit and read some threads about these CPUs. I didn't know Reddit perfected the technology of accessing alternative realities, but here we are.
So yes, zen 5 looks to be a flop for consumers at current pricing. It also puts a dapener on the X3D models. Suspect the 9800X3D might eek out 10% vs the 7800X3D at best in games baring some revolutionary v-cache enhancement. They do very well in database workloads, jit stuff, branchy JS code and other server workloads so I am sure they will be great for the data centre.
@@nextjinI would not be excited about it. Overclocking Ryzen CPUs is pretty boring. Usually you can gain 200-300MHz at best. It will not change anything, taking into account 9800X3D already clocked at 4.8GHz or even a bit higher. 500-600MHz would be a decent margin. But I doubt we will see this.
@@stangamer1151Scatterbencher already hit 5.75ghz single thread load and 5.45ghz multi thread load on a 9700X, at 170W with an AIO, hopefully they don't lock the 9800X3D down too much, it could possibly brute force out some impressive numbers.
Why would you look at server benchmarks anyway? These are desktop CPUs, not server CPUs. I'll care about server benchmarks when the video title is "Epyc Benchmarks/Review"
Some people do have home server setups and whatnot using mainstream hardware, but it's really a very small niche market that sometimes gets overrepresented within tech enthusiast circles.
To be fair, the benchmarks most people were pointing to was Phoronix (and maybe Wendell). Phoronix does Linux benchmarks, and yeah there are a lot of server workloads in those tests, because that is obviously a use case linux dominates in. But the benchmarks are the usual "evaluation performance in linux" set of benchmarks, not specifically targeting server-only benchmarks. Similar with Wendel, he had a bit more focus on workstation stuff as well. Which I think is totally reasonable to test zen5 for, though it will be more interesting with the higher core count parts, as the Ryzen5/7s aren't really what you're looking for for that sort of thing.
Not only server, but workstation actually. I am developer on Linux. And what I saw from Phoronix benchmarks is that I gonna buy one, since it will make a difference.
@@haukionkannel Sure but how does it compare to the 9700X with PBO? That's what we want to see. Only makes sense to compare that too since many people seem to grasp at the PBO straw to justify going for the 9700X over the 7700.
I currently have a 3600 system as well, would you say it's worth it to upgrade to 7700 directly? because I need to replace the motherboard and RAM as well
@@Nick_R_ By the time I sold my AM4 gear the net spend was €150 more than a 5800X3D. I now have an AM5 system with double the ram at 6000mHz instead of 3600mHz.
Guys, seriously appreciate the re-review 🙏 great work. Loved your first reviews, but did think you were light on efficiency & PBO performance. Again, much appreciated!!!
I love how you nitwits are already say Zen5 is a flop...based on the two lowest models. Last time I checked, there are no real reviews showing the performance of the 9950X or 9900X. Not to say that they're going to be amazing, but it's downright hilarious that anyone has the audacity to call an entire generation a flop when half of the cards aren't even out, let alone the most exciting ones.
@@rustler08 Couldn't it be that the 9700X is actually a 9700 non-X or a 9600X and that they've rebranded the CPUs just to make them look better? A marketing trick? Or something wrong with the drivers and/or BIOS (immature? )that will be corrected in due time?
The one thing the Zen 5 launch has confirmed for me is that you MUST never believe the marketing that CPU manufacturers release. Beyond that, I am glad that I have yet to upgrade from my processor from 5800X3D.
Unless you're willing to dump an absurd amount of money on a GPU the 5800X3D will be good for another couple of generations. Whether it will see you though to AM6 is still up in the air, but I actually wouldn't be surprised if it did.
The Ryzen 7700 was what the most reviews missed and i am happy you tested this: the 7700 can be bought for just 210€ in Germany while the 9700x is 380€... So again AMD named the products wrong, should be the 9700(non x) at best and priced around 210€ - the did the same bad naming and pricing they did with their last GPU series.
Thank you for this second review - it felt a lot calmer and pointed out the important parts which remained somewhat unspoken previously. It really cleared up the situation.
All that fast cache keeps the pipeline fed, meaning stalling and flushing isn't needed to be compensated for with a much higher clock speed and faster RAM.
Well and I will change if their change like 7700x vs 9700x and WHY? SImply... if you have ZEN4 wich is good architekture and products.. you will definitively see some idiotic hunt about "maximum" or GHz numbers. But i dont like 85° CPU and i dont pay for good CPU and go ECO.. thats nonsence. I have complete watercooled computer. But you dont change this like.. maximum freq but also maximum heat and tempature. You will mark this like "It can handle this.. " YE YE, but what about FAN behaivors? I dont see changes about bios, what i see.. or maybe a can say hear.. is noise :P So yes... definitively I will change... if their made 9800x3d 95w or 105w - almost same power (i hope for less difference between x3D or non-x and x about IPC, produktivity), but better thermals. Because if i spend some money like a 1k about watercooled PC.. i dont want hear fans because CPU goes 85°!
IIRC there's some other stuff (encrypting/decrypting large amounts of data, etc.) that are bottlenecked by the same instructions that got a lot faster here, but it seems like overall it's pretty niche improvements for too much price.
Good response, although i have a small rebuttal which is that in a post Steam Deck world, the faction of gamers that care about efficiency broadly is growing, even if this is not a mobile chip. We want the architecture to be able to move forward supporting cooler temps and longer battery life in mobile gaming handhelds while still giving good gaming performance. Otherwise innovation on pc gaming handhelds will eventually stall and we'd all have to switch to ARM lol
With the inclusion of full AVX512 in the chiplets, it is now quite obvious, that AMD is tuning the chiplets for the enterprise/server market (Epyc line), where density and power efficiency are increasingly more important. Power-perf scaling is not great beyond a certain point and the desktop parts quickly hit a ceiling (desktop Intel seems to scale better with more power). Unfortunately, dedicated desktops are becoming a niche market with low margins, and AMD is right to make these trade-offs (sacrifices) from the economic PoV. As a company, they live off their high margin high volume products.
They still really F***k up the marketing, the setting of expectations. And the price is ... not that great too. So, it is also right for the customers to boo AMD for the underwhelming product line.
AVX512 and AVX10 will be used one day in games, but it will be a while before it becomes standard on hardware. nothing that comes new in CPUs and GPUs is exclusive to servers and work, everything can be used in programs and games, but it depends on the developer. If there is a game that requires AVX512 and NPU, it will require more current hardware, right.
"Dedicated desktops are becoming a niche market" That claim doesn't sound right to me. If people are not using desktops anymore, what are they migrating to? Laptops? Or phones maybe?
Based on the new naming convention, I am almost wondering if they are going to be releasing 9700XT and 9600XT models that will show more performance. I cannot fathom them dropping a 9700 or 9600 model, because they already sent these out with 65W TDP. It will be interesting to see how their naming plays out now that they've released a product that essentially invalidated their entire Non-X lineup.
Thy could also go back to how they named the 8-core chips with zen2, i.e. release a 9800X as a higher TDP variant. Though I kinda doubt it, they probably just going to give the 9800X3D a bit more TDP, so that they can establish the X3D variants as the enthusiast class gaming king CPU, while the 9700X is simply the efficient desktop SKU as the non-Xes used to be. And we all know more Xes sell better.
9000 series just is not better with higher power… Same as with 3700x. It was 65w part. Later they did release 3700 that was little slover. They also did release 3800x 105w that was tiny bit faster with a lot worse power usage! 9000 series seems to be same! They most likely will keep same architecture in 10000 series and boost speed and give it more power.
I think its possible to get a 9800X with 95 or 105W TDP but I dont think we'd get anything for the 9600X. Pretty sure AMD aimed at increasing their margins this generation and I just dont think we'll be getting anything cheaper.
I needed to replace my cpu with one that has a igpu, and what I'm seeing from Zen 5 inclusing both Steves' retesting has convinced me to just pick up a discounted Zen 4. Thanks for all the testing Steves
OK, so straight in at the 5:00, I know you're trying to do apples to apples on wattage specs, agree totally on 7700 vs 9700X stock, but why 9700X Max PBO against stock 7700X?
First of all, Thanks for the review and the analysis! I don't think these new chips are for anyone who has a new CPU from 18 months ago, or 1 generation old. The way I see it is these chips have great value if you have not upgraded your PC for about 4-6 years minimum, for example someone like me who has a 3800x. I think going from that to a 9700x would be a big deal, especially if I am doing a SFF build, where the lesser amount of heat being output is taken into consideration. I know it's the norm to compare new generation CPU to ones of the previous generation, but I have not yet seen a tech channel do comparisons going all the way back to gen 1 Ryzen. I think for someone like me, that information would be very beneficial, being 3+ generations behind.
Well said!!! I do think that comparing new gen to previous gen is still needed, I do agree with you that it would be great to see the 9600x and 9700x would be a great option to upgrade to if you're using a R5 1600x all the way up to a R5 7700 and everything in between. I myself have a 2700x and a 5600x and would definitely buy a 9700x for a new platform build.
I'm not so sure. Even if you haven't upgraded in 6 years and are considering 9700x, then it's much better to buy 7700, which is only 3% slower, has the same TDP, same platform (AM5), but is 20% cheaper.
29:51 just like the rest 99% of PC Hardware reviewers on this platform. Edit: like literally, what I meant by this comment was that you can't find any channels on RU-vid that are primarily focused on those types of workloads, but you can find countless other channels that are focused on gaming related benchmarks. I don't even mean server or AI workloads, I mean even productivity workloads.
Thanks Steve! I'd like to have seen the parts running at 105w TDP (configured in BIOS) to match the old parts and see where they land performance wise rather than 65w.
Overall on the internet, I'm still predominantly seeing praise for Zen 5 for its efficiency and misleading claims about how they're actually great for gaming if you just up the power draw some more. AMD has really hoodwinked the audience and even many review outlets. Gonna take a good while before the truth can correct this wide-scale misinformation.
