AMD announces Ryzen 9 3900X flagship desktop CPU, Ryzen 7 3800X, more

Look in the mirror. I am using facts and charts. You are unable to accept those facts, nor take AMD's charts at face value.

You are the only person here, who doesn't understand why AMD is showing off the performance of the new chip, verse their old.

It is as if... you do not understand percentages.



(also: This is not an Intel related article, you are the one spamming AMD doom & gloom.)

You failed to provide me a single chart on Ryzen 3000 vs 9700k/9900k in gaming, because AMD didn´t provide one. They chose their comparasions wisely as every other brand does. JayzTwoCents said the same on his last video, if AMD had advantage over intel on the most popular games, they would have shown it. So I´m not alone with this opinion. Cinebench tells me nothing (again, Steven from GamerNexus said the same).

Unlike you, I will wait for 7th July to make my judgement. You are the one screaming AMD won on everything without any reason. Let´s wait and see, then we can talk with facts and charts from independent reviewers.

Never forget this:

maxresdefault.jpg
 
Cinebench is a crap benchmark and was never a reflection of the true performance. R5 1600 beats 8600k on cinebench and it gets destroyed on every other test.
 
Even 1500x beats i5 8400 on cinebench and then it gets dumped on every task. 7th July it is, until then, anyone claiming Amd destroyed Intel is a fanboy, sorry.
 
Lmao and they said Intel was done! Look at that garbo!! Cant even beat a 7700k that costs 150€ 2nd hand. Laughable really! But but cinebench.... Lel
 
Wow. You have two people on here which seem to be the ONLY ones listing and using logical facts, m3tavision and stevea. And instead of editing the parts where one of them went a little too far, you just deleted all of their posts, while with others, you only edited them to suit your tastes. This is blatant discrimination and censorship. Yet, you leave, unedited the posts of people posting absurd conjecture. I am losing respect for you, Techspot.
 
Last edited:
Some new benches appeared and new infos too: https://wccftech.com/amd-ryzen-3000-zen-2-cpu-5-ghz-overclock-4-5-ghz-all-core-boost/

Look at Ryzen 3000 memory latencies and look at max clocks they can possibly achieve... if these leaks are true (I have no way of knowing if they do), AMD will still have no chance vs a 5ghz 9700k or 9900k. No chance at all.
So I went to your link and read this article. Interestingly enough, I found a lot of information directly refuting your statements. For instance, the article clearly stated that the Zen2 processors could achieve 5ghz. Also, they make a point of saying that the Ryzen 3900x was much faster than the 9900k. And it managed to do this even at a much lower clock speed of 4.4ghz, while only using 1.35v. That is a major achievement. They later pointed out that the memory latencies were "simply amazing." So that you don't think I am exaggerating, here is the exact quote, "The CPU can be seen running at a core frequency of 4.6 GHz across all cores while the memory speed is configured at 4000 MHz (CL18 timing). You could see some beefy read/write/copy numbers being pumped out by the chip but the main thing to focus here is the latency which is simply amazing." So I would love to know what prompted you to post your erroneous statements?
 
I think its best for everyone to wait for reviews on July 7th...
The point of this article and comment section is to comment on THIS ARTICLE, and the information we have right now. We can intelligently do that, without speculation, as we were given some clear examples to look at. So as to what might or might not be in the future, yes, I agree we should wait to June 10th and July 7th. But for the information we actually have right now, commenting on it, is what this entire article and thread are for.
I wasn't even going to login or comment on this. I was plenty entertained by just reading all of the comments. But when I saw what I believe to be an injustice, I felt the need to chime in. Because nothing those two posters have written can be refuted by any real logic and reasoning. We've all witnessed what has been taking place in recent times in this industry.
 
Wow. You have two people on here which seem to be the ONLY ones listing and using logical facts, m3tavision and stevea. And instead of editing the parts where one of them went a little too far, you just deleted all of their posts, while with others, you only edited them to suit your tastes. This is blatant discrimination and censorship. Yet, you leave, unedited the posts of people posting absurd conjecture. I am losing respect for you, Techspot.
Posts were deleted for ad hominem comments and subsequent quoting and replying to them. I don't edit individual posts. That tends to encourage, "Let's throw a bunch of crap on the wall and see what sticks," behavior. If you don't follow forum guidelines, your post is subject to deletion regardless of any other content.
 
Posts were deleted for ad hominem comments and subsequent quoting and replying to them.
I completely understand. But the mods have edited out such things in the past on many occasions. There was a lot more information in those posts, than just the ad hominem part, which was quite small, actually. They could easily have deleted that part, and sent a warning to the poster. But they blatantly deleted all of his posts instead, even posts without attacks in them. That was the only point I was trying to make. I do not condone any personal attacks.
 
I completely understand. But the mods have edited out such things in the past on many occasions. There was a lot more information in those posts, than just the ad hominem part, which was quite small, actually. They could easily have deleted that part, and sent a warning to the poster. But they blatantly deleted all of his posts instead, even posts without attacks in them. That was the only point I was trying to make. I do not condone any personal attacks.

There was no information on those posts, only personal attacks, when all I said since the beggining was: Wait for 7th July and then judge. We win nothing by being on every AMD/Intel article saying things like "Intel is done, AMD won, Intel has no place to go, Thanks AMD you won", without any benchs or proof about that. Plus, when you start attacking people that´s when you lose an argument. I didn´t attack anyone and that´s why my posts were not deleted.

I never said Zen2 is not going to be competitive but I also didn´t say Zen2 will beat Intel. From all the graphs and info we have right now there is absolutely no way to tell Zen2 will be superior to Intel current offering. The link that I posted above had a very high latency of 80ns, wich even a 2700x can beat at the same memory speeds (2666 C16). BUT, that´s all rumours still, we shall wait for 7th July and if AMD finally beats Intel on everything without any compromises I will be the first on this website to congratulate AMD and say "Intel is done".

Right now I have no indications of that will happen, so I´m chilled waiting for official results. Because Cinebench means nothing in the real world and we know that for 1 decade now. I want actual high refresh gaming (GTX1080/1080ti/RTX2070/2080/2080ti) benchmarks; I want actual video encoding , handbrake benchmarks etc.
 
There was no information on those posts, only personal attacks, when all I said since the beggining was: Wait for 7th July and then judge. We win nothing by being on every AMD/Intel article saying things like "Intel is done, AMD won, Intel has no place to go, Thanks AMD you won", without any benchs or proof about that. Plus, when you start attacking people that´s when you lose an argument. I didn´t attack anyone and that´s why my posts were not deleted...
I want actual high refresh gaming (GTX1080/1080ti/RTX2070/2080/2080ti) benchmarks; I want actual video encoding , handbrake benchmarks etc.
I agree with most of your post here, but not all of it. We already have plenty of proof where Ryzen2 is flat out better, and you seem to ignore these things, even though you say you will acknowledge when there is proof. There is proof, that Ryzen2 is on a 7nm finFET, while Intel has been stuck on 14nm for ages, and consistently missed their deadlines to get to even 10nm. Where is your congratulations to AMD for this? I haven't seen it. Also, Ryzen1,2&3 are all MUCH more efficient chips. The TDP on these chips are much lower than Intel's best efforts. Where's you congratulations on that? What about AMD pricing their processors well below those of Intel's, some of them more than half the price, while offering better architecture and more efficiency? That's better. So here are things we already know for sure. Yet, you are silent on them. What about AMD beating Intel to all of these accomplishments, when they were well behind for decades? How did that happen? What about AMD being the first with PCIe4? Again, silence on this as well. So, IF you are a person of your word, and you say you would be the first to congratulate AMD when we have proof, here are several things we know now. What say you?
 
I agree with most of your post here, but not all of it. We already have plenty of proof where Ryzen2 is flat out better, and you seem to ignore these things, even though you say you will acknowledge when there is proof. There is proof, that Ryzen2 is on a 7nm finFET, while Intel has been stuck on 14nm for ages, and consistently missed their deadlines to get to even 10nm. Where is your congratulations to AMD for this? I haven't seen it. Also, Ryzen1,2&3 are all MUCH more efficient chips. The TDP on these chips are much lower than Intel's best efforts. Where's you congratulations on that? What about AMD pricing their processors well below those of Intel's, some of them more than half the price, while offering better architecture and more efficiency? That's better. So here are things we already know for sure. Yet, you are silent on them. What about AMD beating Intel to all of these accomplishments, when they were well behind for decades? How did that happen? What about AMD being the first with PCIe4? Again, silence on this as well. So, IF you are a person of your word, and you say you would be the first to congratulate AMD when we have proof, here are several things we know now. What say you?

Well that has a reason. First of all I stopped caring about litology a long time ago when I realized that it doesn´t matter as much as I thought. AMD´s 7nm aren´t really 7nm, more like equivalent to an Intel 10nm. I do appreciate lower power consumption, but I care more about that on GPUs because the differences are massive.

For example, Even with a Cheaper Vega VII than a 2080, I would prefer the nvidia GPU because it uses way less power, even when overclocked. We are talking about 100w differences when both are max overclocked, and that´s a lot. However on the CPU department I don´t care about 30w or 40w difference and that´s on stress testing, because when you use your CPU in real world scenarios, you won´t be using that much power.

9700k can use as much as 130w/140w at 5ghz on a benchmark, but then you start playing games and you will be lucky to see it going above 80w and most likely it will be around 60w, depending on the game.

AMD will use less power than Intel but can already see it suposely needs 1,35v to reach 4,6ghz across all cores (according to recent leaks). 1,35v is already on the high side and I suspect the 7nm advantages go away once you start pushing these Chips, because is a new node and is not 100% optimized yet.

All in all, at the end of the day, unless there are 80w/100w differences between 2 different CPUs, what matters to me is raw performance. I don´t care if it is 5nm or 20nm, as long as it doesn´t use 100w more but has better performance. Same goes for pricing. I seen MRSP for 3700x being 360 dollars. That will be 400€ here in Europe. 9900k costs 480€. Alright, 80€ is money but is not enough money to bother me or to opt for the "weaker" cpu (in case it its worse than 9900k, this is just an example). Now if it was 300€... that would start to be different story.

3600 MRSP 230€ on Europe, but I7 8700 non K with a 4,4ghz turbo boost all cores costs 250€ here. Again, small difference.

The big thing on this launch for now, to me, is the 3900x because it is 3 times cheaper than intel equivalent, and that intel equivalent uses a way worse arch that is even worse than Coffee lake, minus the quad channel memory. The other models, for now, didn´t impress me, BUT as I said, 7th July will tell me if they are impressive or not, after seeing benchmarks. And if they are as good as 9700k/9900k at 5ghz in games and handbrake etc, or even only 5% slower, but cheaper and using way less power, then AMD wins, no argument. And if they are even better than Intel? Then I will turn into a troll and spam "RIP Intel" everywhere, with pride :D
 
Well that has a reason. First of all I stopped caring about litology a long time ago when I realized that it doesn´t matter as much as I thought...
AMD wins, no argument. And if they are even better than Intel? Then I will turn into a troll and spam "RIP Intel" everywhere, with pride :D

First 7nm IS 7nm, not Intel's 10nm. And btw, since you are continually beating the drum on not speculating, and waiting until July to see for sure, why are you even mentioning 10nm, because Intel doesn't even have a 10nm chip yet... Second, litology matters plenty to 99.99% of all enthusiasts! If it doesn't matter to you, so be it. But to this industry, trust me, it matters big. Also, any computer geek who bothers to overclock knows, you want every single bit of efficiency out of your chips, no matter how big or small, to MOST enthusiasts every unit counts, so you still sound like you're making excuses and extremely Intel bias to me. But I will take you at your word and be looking for you in early July, so that you can tell me how crow actually tastes. Lol. And although I agree, the concrete information will probably not come out until July, I am so completely sure of what AMD has done here, just looking at the architecture, I would be willing to place a rather large wager on AMD coming out on top across the board this year, and for several years following it. Their server, consumer and enthusiast lines will wipe the floor with Intel's offerings. And btw,, while I might agree with you on the 3800x, what you get with the 3700x for the price is nothing short of spectacular!!!! How can you NOT be impressed with that?
 
Last edited:
First 7nm IS 7nm, not Intel's 10nm. And btw, since you are continually beating the drum on not speculating, and waiting until July to see for sure, why are you even mentioning 10nm, because Intel doesn't even have a 10nm chip yet... Second, litology matters plenty to 99.99% of all enthusiasts! If it doesn't matter to you, so be it. But to this industry, trust me, it matters big. Also, any computer geek who bothers to overclock knows, you want every single bit of efficiency out of your chips, no matter how big or small, to MOST enthusiasts every unit counts, so you still sound like you're making excuses and extremely Intel bias to me. But I will take you at your word and be looking for you in early July, so that you can tell me how crow actually tastes. Lol. And although I agree, the concrete information will probably not come out until July, I am so completely sure of what AMD has done here, just looking at the architecture, I would be willing to place a rather large wager on AMD coming out on top across the board this year, and for several years following it. Their server, consumer and enthusiast lines will wipe the floor with Intel's offerings. And btw,, while I might agree with you on the 3800x, what you get with the 3700x for the price is nothing short of spectacular!!!! How can you NOT be impressed with that?

I´m an enthusiast and 7nm or 20nm has no interest to me, because as an enthusiast what I want is raw performance with acceptable power consumption. 7nm doesn´t use that much less power than 14nm so I don´t care.

Anyway I decided to stop replying to you when the admin told me you are clone that created the account yesterday.... Nice try tho. At least you tried a 2nd attempt with a civilized way, I appreciate that wich shows you know you were wrong talking that way you did at first. At the same time, for you to bother to create clones to keep replying here, I can sense you are a bit nervous, so chillax and let´s wait for 7th July when we all will know what Zen2 is capable of.

Cheers
 
I currently run an older i5 3570K based system and have been considering an upgrade for a while. I genuinely like the sound of the 3700X but my real problem is thinking of a use case to justify such a powerful CPU. In gaming, my current CPU seems more than adequate and I always have the option of overclocking. I write software for a living but memory is more restricting than CPU speed for me. I just have no idea what I'd do with 16 threads, let alone 24. I do like the idea of 65W though as that would be quieter. It might be amusing having a small graphic running on my desktop that shows 1 core running at 25% and the other 7/15 flat lining.
 
I agree with most of your post here, but not all of it. We already have plenty of proof where Ryzen2 is flat out better, and you seem to ignore these things, even though you say you will acknowledge when there is proof. There is proof, that Ryzen2 is on a 7nm finFET, while Intel has been stuck on 14nm for ages, and consistently missed their deadlines to get to even 10nm. Where is your congratulations to AMD for this? I haven't seen it. Also, Ryzen1,2&3 are all MUCH more efficient chips. The TDP on these chips are much lower than Intel's best efforts. Where's you congratulations on that? What about AMD pricing their processors well below those of Intel's, some of them more than half the price, while offering better architecture and more efficiency? That's better. So here are things we already know for sure. Yet, you are silent on them. What about AMD beating Intel to all of these accomplishments, when they were well behind for decades? How did that happen? What about AMD being the first with PCIe4? Again, silence on this as well. So, IF you are a person of your word, and you say you would be the first to congratulate AMD when we have proof, here are several things we know now. What say you?
We have zero proof that Ryzen 2 is better. In fact if anything it’s all a bit unclear. I’m surprised that AMD haven’t demonstrated that Ryzen 2 is stronger than Intels 14nm competition in any games benchmarks, it makes me feel that maybe it’s not. We won’t know anything about it until it’s released and has been tested by unbiased hardware reviewers like anandtech etc. I get the impression that you really want Ryzen 2 to be good but don’t jump the gun, we have been here before with AMD and the product has eventually failed to deliver. Wait until the benchmarks come in mate.

Oh also, lithography is important but it’s not that important. Radeon VII is 7nm and the 2080 is 12nm yet it uses less power and performs better than the Radeon VII, as a tech enthusiast I would pick the 2080 every time. I’m a gamer and if 7nm Ryzen is slower in games than 14nm Intel I would prefer to own Intel. Performance is more important than lithography and at this point it’s quite clear that a die shrink doesn’t guarantee much, well, except for higher profits for the chipmakers.
 
I´m an enthusiast and 7nm or 20nm has no interest to me... 7nm doesn´t use that much less power than 14nm so I don´t care.
Anyway I decided to stop replying to you when the admin told me you are clone that created the account yesterday....
Cheers
First, your statement that 7nm doesn't use less power than 14nm is ridiculous. It's definitely a large difference by all research, physics and logic. Second, how would you know what the difference is, since AMD is the first to do it? And I have no idea what you are talking about calling me a clone or a bot. I'm not. I'm a real person. And the only reason I created my account yesterday was so I could respond, since I normally only read, as I mentioned yesterday. For a person who says he doesn't want to speculate, and only waits for the proof, you sure do assume a lot. Finally, just to show how erraddic your responses are, you said you decided not to respond to me, in a response to me. That's very amusing.
 
We have zero proof that Ryzen 2 is better. In fact if anything it’s all a bit unclear. I’m surprised that AMD haven’t demonstrated that Ryzen 2 is stronger than Intels 14nm competition in any games benchmarks...
Oh also, lithography is important but it’s not that important. Radeon VII is 7nm and the 2080 is 12nm yet it uses less power and performs better than the Radeon VII...
I listed several reasons in my previous post, where Zen2 is better. So I'm not going through all of that again. And if lithography weren't "that important", than why do the manufacturers spend billions on it. That alone, by all logic and reasoning, shows that it is vitally important. And the case of the Radeon VII is a poor choice to use as an example of an entire platform. It had many problems other than being 7nm. But there is always kinks to workout when using a brand new technology. So your example here is a bad one, and does not make the point. As to your statement here, "I get the impression that you really want Ryzen 2 to be good". We should all want Ryzen 2 to be great, as it benefits everyone for it to be. Not only does it give us a great product at a lower cost, but it also stirs competition which will get Intel off their butts, and start innovating again. And again, we all benefit from that. A strong competition not only pushes these companies to innovate, but it also drives prices down. A win win for all of us.
Maybe the problem is that you kids' entire world revolves around playing games, and the rest of us also use our computers for CAD software, rendering videos, scientific calculations and even simulators which need that power. There's a lot more to computing than just single mindedly playing games. And as for your surprise about not having a demonstration against Intel in games yet? There hasn't even been an real product release yet. That was just a Keynote at a conference. But there's another opportunity on June 10th, and it will be interesting to see what comes out then. Cheers.
 
Last edited:
I currently run an older i5 3570K based system and have been considering an upgrade for a while. I genuinely like the sound of the 3700X but my real problem is thinking of a use case to justify such a powerful CPU. In gaming, my current CPU seems more than adequate and I always have the option of overclocking. I write software for a living but memory is more restricting than CPU speed for me. I just have no idea what I'd do with 16 threads, let alone 24. I do like the idea of 65W though as that would be quieter. It might be amusing having a small graphic running on my desktop that shows 1 core running at 25% and the other 7/15 flat lining.
The 3700x looks like an excellent choice. The price is great, and as you mentioned, the TDP can't be beat. On top of that, it fixes your problem with the ability to use much higher speed memory, and with PCIe4 your entire system would run faster and smoother. Since you write software for a living, I would assume that more and more the requirements for such software would be to take advantage of all the new cores and threads of the latest hardware? Am I wrong in that assumption?
 
We have zero proof that Ryzen 2 is better. In fact if anything it’s all a bit unclear.
What's unclear to me, is why so many people seem to be so against AMD's efforts, instead of cheering them on? And why anyone would defend Intel, after Intel has screwed over millions of enthusiasts by refusing to innovate and milking their monopoly for decades, making minute upgrades, changing the socket every time to force sales and charging ten times the price for things, because they could. Wanting Zen 2 to be good is what every enthusiast should feel. Because it affects all enthusiasts in a positive way if it is good. So why wouldn't you want it to be?
And yes, we all want to see the benchmarks in July. But that's still several weeks away, so we get to play this game in the meantime. I think it's very interesting.
 
The 3700x looks like an excellent choice. The price is great, and as you mentioned, the TDP can't be beat. On top of that, it fixes your problem with the ability to use much higher speed memory, and with PCIe4 your entire system would run faster and smoother. Since you write software for a living, I would assume that more and more the requirements for such software would be to take advantage of all the new cores and threads of the latest hardware? Am I wrong in that assumption?
It's less the speed of the memory and more the quantity of it :) I haven't noticed much difference with memory speed - haven't done any conclusive tests though. You'd think more cores etc would help in software dev but I find my existing processor is, if anything, underused. I suspect if you play (certain) games and want to do video at the same time then these things might be useful. I'd like to upgrade, I feel more comfortable with Intel but have no real affiliation with them. I can build a fairly high spec PC and money isn't an issue. I'm just finding it hard justifying getting a 16 thread CPU - is there anything that needs it?
 
but the icebreaking is that Ryzen Zen2 has better IPC than Intel now.

The new Ryzens that will come out this summer have 15% more IPC than the current ones, so, yes, they'll be finally ahead of Intel.

But then Intel announced that their new Core chips, to be available this fall, will have 18% more IPC. So AMD is really further behind? Maybe not.

The graph accompanying that announcement seems to imply that it's 18% more IPC than some Intel chips from a while ago, not 18% more than what they're selling now that AMD will get ahead of with 15%. But I may not have understood it right.
 
I recently looked up some facts about Intel and AMD processors which corrected a false assumption I had.

When Ryzen first came out, they doubled the floating-point processor power per core - which was good, because Bulldozer had half as much as what the Intel chips it was competing with had. But Intel had doubled their floating point power per core after that - so Ryzen was still at half of what Intel had.

Now, AMD has doubled Ryzen's floating-point power per core. But I vaguely remembered something about Intel doubling theirs - yet again. Well, checking into it, I found out that the doubling I remembered was only on the Xeon chips in the Skylake generation, so AMD is not back where it was, it really is fully caught up with Intel as far as consumer chips are concerned. So I think we can expect the new Ryzens to be very respectable gaming chips, leaving very little benefit to be obtained by spending more on Intel.
 
Back