(16core added)AMD 3000 specs! 4.7 GHZ, R9 3950x, R7 3700x, 3800x.

[Rune]

Eh, already have a 9900k pretty heavily OC'ed. Maybe next time, AMD.

[Drak3]

Just now, Quadriplegic said:

Which is almost irrelevant as Intel CPUs have higher ceiling of clock speeds, entirely negating difference in IPC. Zen 2 chips have to overclock well to be successful

If AMD's IPC claims hold true and can OC to their top base clock, the performance difference between Intel and AMD will be pathetically small. In the neighborhood of a 9900K @ 5GHs vs a 9900K @ 5.05-5.1GHz.

[Origami Cactus]

2 minutes ago, RejZoR said:

Not quite 5GHz, but this ain't bad I guess. 4.6GHz max turbo and 15% IPC uplift should give quite decent performance. I don't think these will be hitting 5GHz even on high end daily cooling, probably like 4.8GHz and 5GHz if really lucky if we are basing it on the existing OC capabilities.

Yeah but new process and new design, so the past OC capabilities shouldn't matter.

[xg32]

28 minutes ago, plsno said:

I really do hope these will OC to like 5GHz. 

it's most likely 4.6, look at the single core boost on zen + and the max all core oc, if it does go to 5.0 i think ppl would lose their shit, but slightly too good to be true.

[RejZoR]

1 minute ago, Origami Cactus said:

Yeah but new process and new design, so the past OC capabilities shouldn't matter.

It's still Zen and the clock step up is about the same as it was between Zen and Zen+. If process and design were so much better they'd be hitting 5GHz out of the box at least on the highest end CPU. But they aren't.

[LAwLz]

Quite frankly, anyone who has trusted AdoredTV in the last 1-2 years is either a moron or not researched enough. The guy (AdoredTV) has consistently been wrong on the stuff he says, only to then backpadel or cherry pick benchmarks to kind of get it right. 

 

Only 2 out of 20 benchmarks gets the results he wants? Let's only show those 2 benchmarks in the video! Nothing biased or misleading about that, right? 

 

Fuck that dude. 

 

 

Anyway, Ryzen 2 sounds really good. Not anywhere near as good as some of the leaks (some which even said the 6 core would be like 99 dollars) but still really good IPC improvement. 

Correct me if I'm wrong, but we aren't really getting any more cores per dollar, nor that much MHz increase per dollar compared to current prices. Correct? 

[Humbug]

2 hours ago, xAcid9 said:

How you like Adored now? 

LOL, at least it is entertaining to listen to him. So even though he is bad at predictions/leaks I will keep listening to him just cause I enjoy it.  And I don't tend to get over-hyped and disappointed. So listening to somebody wade through technical stuff and speculate about GPUs and CPUs and the industry is just fun. Even though he has lost his credibility for predictions.

 

But he actually could have done some damage here to AMD because he made people expect the 3800x to be 5Ghz boost. So the people who got overhyped based on that can be blamed on him.

 

And then people blame AMD for overhype and let down... when in reality they have given us a 200Mhz clock speed boost with 15% IPC gain and better efficiency! Which is actually really good, AMD is consistently executing on their roadmap faster than Intel  and with bigger improvements per gen than Intel.

1 hour ago, SenioRR said:

Will Intel even be recommended anymore in future builds? Doesn't look like they have much of an advantage left after the 15% IPC increase for AMD.

They IPC advantage is completely gone. But they still have a few 100Mhz clockspeed advantage, and they will throw power/efficiency/price concerns out of the window and try to sell 5Ghz CPUs. However they will hold on to the laptop market for now because AMD is not bringing Zen2 to mobile parts for now.

[S w a t s o n]

If they release 16 core by holiday 2019 I guess I can forgive them for being greedy. Still though, they could definitely be selling the 16 core on launch, they arent hard pressed for working 8 core chiplets.

[Humbug]

26 minutes ago, RejZoR said:

Not quite 5GHz, but this ain't bad I guess. 4.6GHz max turbo and 15% IPC uplift should give quite decent performance. I don't think these will be hitting 5GHz even on high end daily cooling, probably like 4.8GHz and 5GHz if really lucky if we are basing it on the existing OC capabilities.

I expect 4.7Ghz to be the common max stable overclock on all cores without doing anything exotic.

 

I say this because the highest clocked part we got is 4.6Ghz and so far AMD clocks Ryzen close to the max reasonable frequency. 

[S w a t s o n]

Just now, Humbug said:

I say this because the highest clocked part we got is 4.6Ghz and so far AMD clocks Ryzen close to the max reasonable frequency. 

on 14nm soc process at the literal max the process can handle, meant for phone chips. It's not the arch as much as it was the node.

[Drak3]

1 minute ago, S w a t s o n said:

on 14nm soc process at the literal max the process can handle, meant for phone chips. It's not the arch as much as it was the node.

And this contradicts Humbug's statement....how?

[Origami Cactus]

5 minutes ago, S w a t s o n said:

If they release 16 core by holiday 2019 I guess I can forgive them for being greedy. Still though, they could definitely be selling the 16 core on launch, they arent hard pressed for working 8 core chiplets.

There is actually a big reason.

All the max binned 8 core chiplets are going to EPIC and ryzen 3800x. So the 3950x will come when they have enough chips.

[Zodiark1593]

1 hour ago, IAmAndre said:

So is it game over for Intel?

Losing the battle != losing the war. Cases in point, Intel came back from Netbust and AMD came back from Faildozer. At the very minimum, I wouldn't expect Intel to have been sitting on their thumbs these past couple years since the initial Ryzen launch (one would hope anyway). I want the war to heat up and benefit the consumer, not for either to cede entirely.

[S w a t s o n]

4 minutes ago, Drak3 said:

And this contradicts Humbug's statement....how?

AMD clocked ryzen near the max clock because that was what they were working with and needed to be competitive with intel. This node is a high performance node specifically designed for chips like this vs a phone soc at 12w. Along with this goes things such as clock speed increases. AKA the chips should be able to go much faster, and with the IPC lift, AMD doesnt have need to go into the poor scaling region of the voltage curves on stock chips. That's how you get an 8core at 65 watts to a 9700k at 95w.

[S w a t s o n]

2 minutes ago, Origami Cactus said:

There is actually a big reason.

All the max binned 8 core chiplets are going to EPIC and ryzen 3800x. So the 3950x will come when they have enough chips.

The epyc chips do not need the max binned chiplets that can do 5ghz. They want power efficient chips that will play nice around 2-4ghz. That is basically all working 8core chips most likely. Yield is supposed to be something like 70% for working 8 core chiplets.

[LAwLz]

13 minutes ago, valdyrgramr said:

So, where's the proof that AMD did the changes rather than him just speculating as usual?  He already admitted that he had speculated and was wrong about the CES portion of that.

 

Gets 1 thing right and 3 things wrong.  Much credible.

Gotta love what shitty defense he has as well.

If he says something which turns out t be wrong he can just blame his "leaker" (which I wouldn't be surprised is nobody, just him making shit up to get views) for being wrong, or AMD for "changing" things "last minute".

 

Even if we make the massive assumption that he has contacts with a leaker in the industry, that source has consistently been wrong about things and as a result, Adored should not be seen as a reliable source (since Adored is just a proxy for the leaker) by any rational human being.

 

 

Here is someone on Reddit who did a quick summary of AdoredTV's "leaks" (aka, something he just made up) and the official specs):

But hey, he got like, 3 things right so obviously he is very reliable! Never mind the other ~11 things he got wrong. Let's all focus on that he managed to leak the clock speed of the 3700X correctly (even though he got the price wrong, and name wrong).

[Drak3]

2 minutes ago, S w a t s o n said:

AMD clocked ryzen near the max clock because that was what they were working with and needed to be competitive with intel.

They could overtake Intel in terms of performance by doing the same here, if their IPC number is accurate.

 

Plus, AMD has a history of offering a few SKUs near the max clock.

 

 

Basically, if they aren't clocking higher, it's highly likely that they can't.

[Humbug]

6 minutes ago, Zodiark1593 said:

Losing the battle != losing the war. Cases in point, Intel came back from Netbust and AMD came back from Faildozer. At the very minimum, I wouldn't expect Intel to have been sitting on their thumbs these past couple years since the initial Ryzen launch (one would hope anyway). I want the war to heat up and benefit the consumer, not for either to cede entirely.

As far as consumer segments go...

-Desktop workstation / PC gaming / enthusiast CPU market is clearly swinging towards AMD

-Intel is going to rely a lot on the laptop market. AMD still will not touch them in that market. Because AMD is giving us old tech in laptops. Wish they had given us new mobile APUs with Zen2 cores+Navi graphics. The 7nm efficiency and performance combo would have been great on laptops.

 

Intel will obviously come back in time. AMD has to make as much money as possible in the next two years. The good thing is that this time they are executing very well on their roadmap. Their plan of leapfrogging design teams is working and they are consistently moving forward every year from Zen to Zen+ to Zen 2. This is a good sign for them longterm.

[Drak3]

2 minutes ago, VegetableStu said:

 

 

also the ramifications of Threadripper ._. the 3900X is a solid alternative to TR2950X (and possibly a straight-up replacement to TR2920X). however much the memory topology affects their respective performance have to wait untill reviewers get their hands on them

Bear in mind, with Threadripper you get more memory channels and PCIe lanes, which coincides with the prosumer demographic those are aimed towards.

[S w a t s o n]

Just now, Drak3 said:

They could overtake Intel in terms of performance by doing the same here, if their IPC number is accurate.

 

Plus, AMD has a history of offering a few SKUs near the max clock.

 

 

Basically, if they aren't clocking higher, it's highly likely that they can't.

No. They are sandbagging, same reason for only releasing 12 core.

That history is not correct, 8350 vs 9590, go back further and see I'm still right.

They have overtaken intel, did you look at the charts?

[Drak3]

Just now, S w a t s o n said:

No. They are sandbagging, same reason for only releasing 12 core.

You have a source for that?

 

1 minute ago, S w a t s o n said:

They have overtaken intel, did you look at the charts?

Going with their IPC increase number and clock speeds, they're tied/right behind Intel. That's making the absurdly stupid assumption AMD's IPC numbers are accurate or even apply to a wide range of workloads.

[jake9000]

man, buying one of these puppies on day one is gonna be a trial. 

[Zodiark1593]

Just now, Humbug said:

As far as consumer segments go...

-Desktop workstation / PC gaming / enthusiast CPU market is clearly swinging towards AMD

-Intel is going to rely a lot on the laptop market. AMD still will not touch them in that market. Because AMD is giving us old tech in laptops. Wish they had given us new mobile APUs with Zen2 cores and Navi graphics they would. The 7nm efficiency and performance combo would have been great on laptops.

Not sure why AMD is staying away from laptops. Quad Zen 2 and Navi would be pretty amazing for laptops. If it's a branding issue, would probably be good to get that marketing department some midnight oil to burn.

 

I really want to see AMD getting into single board computers. A high clocked, dual core Zen 2 and modest Navi GPU would be amazing for emulation, though admittedly, this is kind of a niche within a niche.

[S w a t s o n]

1 minute ago, Drak3 said:

You have a source for that?

 

Going with their IPC increase number and clock speeds, they're tied/right behind Intel. That's making the absurdly stupid assumption AMD's IPC numbers are accurate or even apply to a wide range of workloads.

They are not behind. They are matched for IPS minimum. My source is for amd sandbagging is them releasing a 12 core and not a 16 core and winning single and multihreaded perf

[Drak3]

Just now, S w a t s o n said:

My source is for amd sandbagging is them releasing a 12 core and not a 16 core and winning single and multihreaded perf

So, no source.