AMD reveals info on Zen base Opterons

[TheRandomness]

3 minutes ago, JamGorby said:

My non-stance on the issue does not equal your speculation. Do you even logic? I'm saying there's no reason to believe claims that this chip will have such 'potential' because there is no evidence to suggest it. You're pulling claims out of your ass because you want to believe them. Big difference. 

I don't do logic. Also, I was agreeing with you in the second halfish of my 3rd sentence. And about the temps and performance, I was just using an example as you said they could be on fire so I went for the opposite end

15 minutes ago, TheRandomness said:

You have no idea either. Maybe they're running a sub-ambient and have the performance of 5 E5-2699 v4s. We don't know and that guy is basing his statement on previous leaks, so I'd just ignore it.

 

[Suika]

1 hour ago, Djole123 said:

40% IPC increase

Man, I remember the last time AMD claimed their performance was going to be higher than actual.

 

Oh wait, that was last month.

[Energycore]

6 minutes ago, Suika said:

Man, I remember the last time AMD claimed their performance was going to be higher than actual.

 

Oh wait, that was last month.

Pretty sure AMD never claimed performance over 970

 

The only performance claim they made was the one CF benchmark in computex. Every other performance point we got was a leak, unofficial and actually against AMD's NDA

[2FA]

2 minutes ago, Energycore said:

Pretty sure AMD never claimed performance over 970

 

The only performance claim they made was the one CF benchmark in computex. Every other performance point we got was a leak, unofficial and actually against AMD's NDA

AMD honestly didn't hype Polaris at all, consumers did all of the hyping themselves.

[Suika]

3 minutes ago, Energycore said:

Pretty sure AMD never claimed performance over 970

Did they not claim performance equivalent to $500 GPUs?

 

I mean, I don't remember the GTX 970 ever being that expensive.

[Energycore]

2 minutes ago, Suika said:

Did they not claim performance equivalent to $500 GPUs?

 

I mean, I don't remember the GTX 970 ever being that expensive.

I think I saw that. Do you have the slide where they showed it?

[Curufinwe_wins]

1 hour ago, NumLock21 said:

32 core dies don't exist. Even Intel's 24 cores isn't a single die with 24 cores.

Technically it is a single die, but it is a three ring die. 

 

 

[JamGorby]

29 minutes ago, TheRandomness said:

so I'd just ignore it.

 

But you aren't when you're making claims like "it has such potential". That's not ignoring. 
What you're saying is, because you're wrong, I should just ignore it while you should be allowed to make claims like this chip will be a beast. 

Other than that, everything else has been word salad. 

[NumLock21]

22 minutes ago, DeadEyePsycho said:

AMD honestly didn't hype Polaris at all, consumers did all of the hyping themselves.

Well, Raja did started it.

[Qub3d]

22 minutes ago, Suika said:

Did they not claim performance equivalent to $500 GPUs?

 

I mean, I don't remember the GTX 970 ever being that expensive.

The claimed "performance equivalent to $500 GPUs ****"

 

*In VR

**In specific games

***at specific clocks

etc.

 

So technically, they didn't lie. 

[laminutederire]

We all know that AMD isn't offering the best cpu ever, but that it will be exciting since they're going to get up to speed with Intel with performances closer to 5 or 6th generation of intel. By the time it comes out, 7th generation will be out, but AMD won't get owned by 3rd generation i5 anymore or something like that. It's like with their gpu, it's impressive they arrive at that efficiency, knowing where they come from.

I think they will have those +40% to catch up a few generations where they fell behind. They know they can't lie, and the whole pcie slot fiasco is proving them they really can't. 

[Coaxialgamer]

1 hour ago, NumLock21 said:

AMD Zen Opteron is similar it's connected via a 100GB/s fabric. It's a server CPU, they're not that stupid to have them separated and running on its own. It has ultra high speed interconnect.

They'll have them communicate through the FSB,  just like in the pentium d days /jk

[NumLock21]

1 minute ago, Coaxialgamer said:

They'll have them communicate through the FSB,  just like in the pentium d days /jk

More like HT

[Suika]

47 minutes ago, Energycore said:

I think I saw that. Do you have the slide where they showed it?

It wasn't a slide, more so a claim from Raja Koduri.

 

http://www.videogamer.com/news/amd_reveals_199_radeon_rx_480_gpu_claiming_500_performance.html

[Energycore]

2 minutes ago, Suika said:

It wasn't a slide, more so a claim from Raja Koduri.

 

http://www.videogamer.com/news/amd_reveals_199_radeon_rx_480_gpu_claiming_500_performance.html

Ah, they claimed on-par with last-gen $500 cards (980, Fury, Nano) in VR. I'd be surprised if AMD didn't provide some kind of VR optimization like what Nvidia did with SMP

[Suika]

1 minute ago, Energycore said:

Ah, they claimed on-par with last-gen $500 cards (980, Fury, Nano) in VR. I'd be surprised if AMD didn't provide some kind of VR optimization like what Nvidia did with SMP

I was hoping they would have but haven't released any specific information, and am only assuming they were just trying to brute force it with raw performance.

[Energycore]

2 minutes ago, Suika said:

I was hoping they would have but haven't released any specific information, and am only assuming they were just trying to brute force it with raw performance.

Hmm.

 

 

I dug that up from videocardz. It's videocardz though.

[CTR640]

AMD's history is not making it any better untill it happens actual.

For now:

 

[Energycore]

2 minutes ago, CTR640 said:

AMD's history is not making it any better untill it happens actual.

For now:

 

[patrickjp93]

2 hours ago, NumLock21 said:

32 core dies don't exist. Even Intel's 24 cores isn't a single die with 24 cores.

Intel's 24-core Broadwell E7s and 22-core E5s are single-die chips.

 

The KNL Xeon Phi is also single-die at nearly 700mm sq.

[Doobeedoo]

32 cores yo that's a lot of cores  on paper it looks amazing really.

Though for Summit Ridge as far as IPC performance potentially being twice of FX8350 or more is seriously attractive. Will be exciting to see the reveal.

[Humbug]

6 hours ago, Suika said:

It wasn't a slide, more so a claim from Raja Koduri.

 

http://www.videogamer.com/news/amd_reveals_199_radeon_rx_480_gpu_claiming_500_performance.html

In that link there is no quote from AMD or Raja claiming $500 performance. I also googled it and couldn't find a direct quote, just something that a few websites kept repeating probably after one of them started it. Feel free to correct me if you have a direct quote, I could be wrong.

But otherwise this may once again be a case of the community + media over-hyping something rather than the vendor itself.

[DarkBlade2117]

8 hours ago, TheRandomness said:

For a budget workstation, the 8 cores of utter crap are better than you think.

No. just saying, I'm not an AMD fanboy. I would rather go Intel any day.

When my i3 can render faster than my friends FX-6300

[Suika]

39 minutes ago, Humbug said:

In that link there is no quote from AMD or Raja claiming $500 performance. I also googled it and couldn't find a direct quote, just something that a few websites kept repeating probably after one of them started it. Feel free to correct me if you have a direct quote, I could be wrong.

But otherwise this may once again be a case of the community + media over-hyping something rather than the vendor itself.

Looking back through multiple sources, you may be right in assuming the community and journalists over hyped it, as the only direct quote from Koduri I can find through multiple articles is him mentioning that the GPU is built like a $500 premium video card, which may be referring strictly to the reference cooler being similar in aesthetic to the R9 Nano and Fury X rather than performance. Either a deliberately poor choice of words, or a statement taken completely out of context.

[SLAYR]

35 minutes ago, DarkBlade2117 said:

When my i3 can render faster than my friends FX-6300

That all depends on how many threads you can/are rendering with.