AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***

[humbug]

There is still a fairly large divide from a quality standpoint. Zen is still extremely rough around the edges as far as being an enthusiast platform is concerned...

Are you serious? Intel CPU's are leaker than a colander, while AMD's CPU's stand up to all these security tests.

Intel are also still behind on architectural form factor while AMD are running rings round them.

Seriously how does a $210bn company get it so wrong while a $30bn company gets it so right?

 

[Silent_Scone]

It is? would be interested to hear your thoughts.

Just the nuances that come with overclocking the architecture remain, and then some additional ones on top. It's not deal breaking, but can be jarring nether the less. Point was, Intel is a lot easier to work with - that's just the way it's been in my experience.

It's going to be amazing if AMD with their Zen 2 16 core@~4.2GHz or so can match, or win, over the Intel i9 7960X at 3x the price - That is nuts!


That was a fair comment when Zen was first released, especially with RAM compatibility, but it's not the case now. Sure we could have a similar situation when Zen2 is first released but I don't think it'll be as bad as when Zen1 arrived.

God no, least not at launch in a short time.

Are you serious? Intel CPU's are leaker than a colander, while AMD's CPU's stand up to all these security tests.

Intel are also still behind on architectural form factor while AMD are running rings round them.

Seriously how does a $210bn company get it so wrong while a $30bn company gets it so right?

Those with hands-on experience with both the Z platforms and Zen will know precisely what I'm talking about. Nobody in this thread should honestly care about the security noise. Unless, of course, it suits their argument.

 

[beany_bot]

fanboi for saying the truth.

It's not the truth though is it? You are saying "Intel deliver". They haven't. At all. They only released 6 and 8 cores when AMD forced their hand. And they are riddled with bugs. That is NOT delivering. You don't talk the truth.

 

[pmc25]

Just the nuances that come with overclocking the architecture remain, and then some additional ones on top. It's not deal breaking, but can be jarring nether the less. Point was, Intel is a lot easier to work with - that's just the way it's been in my experience.



God no, least not at launch in a short time.



Those with hands-on experience with both the Z platforms and Zen will know precisely what I'm talking about. Nobody in this thread should honestly care about the security noise. Unless, of course, it suits their argument.

So, it's secret stuff that you can't specify. Nod, nod. Wink, wink. Right.

 

[Silent_Scone]

So, it's secret stuff that you can't specify. Nod, nod. Wink, wink. Right.

Doesn't really make any difference, it'll be apparent soon enough. Just speaking from experience, which tends to help in these situations. You don't get the same round the houses approach with Intel that you can often get on the other side of the fence.

That doesn't mean I'll still be using my 9900K for much longer, though lol

 

[Potatowithearsontheside]

I guess that you mean having to tinker with RAM timings in order to get the best out of Ryzen CPUs, as opposed to flicking on an XMP profile with Intel. Overclocking the CPU itself is mostly pointless with Ryzen, since it maxes itself out of the box.

 

[kitfit1]

I guess that you mean having to tinker with RAM timings in order to get the best out of Ryzen CPUs, as opposed to flicking on an XMP profile with Intel. Overclocking the CPU itself is mostly pointless with Ryzen, since it maxes itself out of the box.

Was thinking exactly the same thing myself. As there are little or no real challenges in clocking a CPU anymore, the only challenge left is ram. As someone that has always enjoyed clocking and tightening ram, the AM4 platform has been a delight. As was the X58, Opteron and Athlon before it. There really is more satisfaction getting ram right than anything else.

 

[Journey]

Was thinking exactly the same thing myself. As there are little or no real challenges in clocking a CPU anymore, the only challenge left is ram. As someone that has always enjoyed clocking and tightening ram, the AM4 platform has been a delight. As was the X58, Opteron and Athlon before it. There really is more satisfaction getting ram right than anything else.

This forum would not exist if overclocking used to be an easy exact science. I think we can all agree on that.

 

[~>Dg<~]

It's not the truth though is it? You are saying "Intel deliver". They haven't. At all. They only released 6 and 8 cores when AMD forced their hand. And they are riddled with bugs. That is NOT delivering. You don't talk the truth.

how have they not delivered ? they have had for the last 15 years straight the fastest cpus in nearly every sector. they dont say check out this cinebench score. they just give you what you need. without the hype. AMD is hyping you are many other pro amd are. what if they dont deliver ? another 15 years of AMD will do it next time ?

the funny thing is people who are pro amd crow on and on for the last 15 years. so when AMD does finally pull there finger out there backsides it will be told you so ! if you say it all the time sooner or later you might be right. take the bias out of your opinion and tell me has amd been better than intel in the last 15 years. the truth is no. they have a shot this time. its not guarenteed. so dont dive in double footed. just incase they do the same old trick.

then it will be but the price is good....LOL. AMD never beaten even when slower.

 

[humbug]

how have they not delivered ? they have had for the last 15 years straight the fastest cpus in nearly every sector. they dont say check out this cinebench score. they just give you what you need. without the hype. AMD is hyping you are many other pro amd are. what if they dont deliver ? another 15 years of AMD will do it next time ?

the funny thing is people who are pro amd crow on and on for the last 15 years. so when AMD does finally pull there finger out there backsides it will be told you so ! if you say it all the time sooner or later you might be right. take the bias out of your opinion and tell me has amd been better than intel in the last 15 years. the truth is no. they have a shot this time. its not guarenteed. so dont dive in double footed. just incase they do the same old trick.

Literally the only one with hardware bias in this thread is you, just you, and only you are complaining about it, all your doing is trying to distract from your blatant bias by accusing a bunch of others of doing exactly what you do.

 

[Scougar]

We already know they will be (much) faster

This is a leak, not a fact. You should not present it as fact until confirmed.

 

[Scougar]

Just the nuances that come with overclocking the architecture remain, and then some additional ones on top. It's not deal breaking, but can be jarring nether the less. Point was, Intel is a lot easier to work with - that's just the way it's been in my experience.

God no, least not at launch in a short time.



Those with hands-on experience with both the Z platforms and Zen will know precisely what I'm talking about. Nobody in this thread should honestly care about the security noise. Unless, of course, it suits their argument.

We care if it IMPACTS PERFORMANCE, due to patching. I definitely agree on Intel being easier to OC though. However, if someone decided to script these exploits...then people will start caring more.

 

[humbug]

This is a leak, not a fact. You should not present it as fact until confirmed.

Yes, however, it doesn't seem at all unrealistic, given that we do know a Ryzen 3000 8 core scores the same as a 4.7Ghz 9900K in Cinebench R15, so it almost stands to reason that the same cores times 16 would be roughly around twice as fast as a 9900K in the same application, or as fast as a 16 core Skylake-X running at 4.8Ghz.

The only question is, what speed was that 8 core Ryzen 3000 running at?

 

[Potatowithearsontheside]

how have they not delivered ? they have had for the last 15 years straight the fastest cpus in nearly every sector. they dont say check out this cinebench score. they just give you what you need. without the hype. AMD is hyping you are many other pro amd are. what if they dont deliver ? another 15 years of AMD will do it next time ?

the funny thing is people who are pro amd crow on and on for the last 15 years. so when AMD does finally pull there finger out there backsides it will be told you so ! if you say it all the time sooner or later you might be right. take the bias out of your opinion and tell me has amd been better than intel in the last 15 years. the truth is no. they have a shot this time. its not guarenteed. so dont dive in double footed. just incase they do the same old trick.

then it will be but the price is good....LOL. AMD never beaten even when slower.

They've certainly delivered on 10nm...
Nope, 3 years overdue and still no sign of delivery.
Fact is, they were beaten to 8c+ on the consumer platform, beaten to PCIe4, beaten to 7nm. I say beaten...Intel have not delivered on either yet.

 

[Orange Nexus]

Doesn't really make any difference, it'll be apparent soon enough. Just speaking from experience, which tends to help in these situations. You don't get the same round the houses approach with Intel that you can often get on the other side of the fence.

That doesn't mean I'll still be using my 9900K for much longer, though lol

Will this be the first AMD that you'll keep? I mean it is a keeper?

Tell us more!!!

 

[Scougar]

Yes, however, its doesn't seem at all unrealistic, given that we do know a Ryzen 3000 8 core scores the same as a 4.7Ghz 9900K in Cinebench R15, so it almost stands to reason that the same cores times 16 would be roughly around twice as fast as a 9900K in the same application, or as fast as a 16 core Skylake-X running at 4.8Ghz.

The only question is, what speed was that 8 core Ryzen 3000 running at?

Seems realistic I agree. But wccftech has been posting so much stuff and now people are suggesting it is fact. No. Just no. Don't need people posting some thing and mislead the casual observer thinking it is true.

Hopefully the 16 core will be 5ghz capable, that would be very nice.

 

[humbug]

Seems realistic I agree. But wccftech has been posting so much stuff and now people are suggesting it is fact. No. Just no. Don't need people posting some thing and mislead the casual observer thinking it is true.

Hopefully the 16 core will be 5ghz capable, that would be very nice.

wccftech i agree, but i think the content of the slide is fine, tho as always pinch of salt, you're right, we don't know.

 

[Scougar]

P s. Thank you scone for sharing what you have been able. I'm involved in a sort after product release and trying to share things can be tricky. So thanks.

 

[4K8KW10]

But wccftech has been posting so much stuff

wccftech is wrong when they have to think or interpret its moles' informations. In this case, you have a figure which shows an exact metric - in this case the performance of an ES at 4.2 GHz.
I don't know why it's so difficult for some to understand simple things.

 

[pmc25]

They've certainly delivered on 10nm...
Nope, 3 years overdue and still no sign of delivery.
Fact is, they were beaten to 8c+ on the consumer platform, beaten to PCIe4, beaten to 7nm. I say beaten...Intel have not delivered on either yet.

10nm losing EUV, and their 7nm EUV scheduled to ship in ~late H1 '21 is the bigger story than the actual node size. Samsung are already in full production for their 7nm LPP EUV. 2 years later than a company they were considered to be 'years' ahead of until recently.