AMD provides a glimpse of it's new ZEN processor : breakthrough performance !

[Prysin]

3 hours ago, othertomperson said:

I do own products though -- exclusively of AMD's competitors. Also an AMD fanboy whose comments about Zen have so far been overwhelmingly negative is astounding shit at shilling.

Your counter argument is rather weak. Considering you have been proven wrong on so many occasions.

 

i think, the forums would be better without your misinformed, toxic and trollish posts.

[ScratchCat]

AMD beats Intel Broadwell IPC!?!?!

AMD has better power efficiency than Broadwell-E!?!?!

*does dance of joy*

[dalekphalm]

Just now, ScratchCat said:

AMD beats Intel Broadwell IPC!?!?!

AMD has better power efficiency than Broadwell-E!?!?!

*does dance of joy*

Well, don't dance too hard now - these are only preliminary results.

 

I'm feeling quite confident about the Zen launch, but keep in mind, AMD has fucked up easier wins before.

[ScratchCat]

Just now, dalekphalm said:

Well, don't dance too hard now - these are only preliminary results.

 

I'm feeling quite confident about the Zen launch, but keep in mind, AMD has fucked up easier wins before.

Even then this is an engineering sample and is doing this well.

 

[Vode]

It's quite possible the IPC is not as good in gaming or other tasks. We'll have to wait and see.

 

But considering even the best FX would even lose against the quad core i7s, (and let's be honest rendering was one of the FXs strong suits), being on par at equal clockspeed with Broadwell-E is pretty good.

 

5960x performance at 95W TDP would be amazing. Fingers crossed. Let's get some real competition back in the high-end CPU space. It's only good for us consumers.

[Coaxialgamer]

20 minutes ago, Vode said:

It's quite possible the IPC is not as good in gaming or other tasks. We'll have to wait and see.

 

But considering even the best FX would even lose against the quad core i7s, (and let's be honest rendering was one of the FXs strong suits), being on par at equal clockspeed with Broadwell-E is pretty good.

 

5960x performance at 95W TDP would be amazing. Fingers crossed. Let's get some real competition back in the high-end CPU space. It's only good for us consumers.

the only thing i am worried about is what clock speeds it will reach . Samsung's 14nm lpp was never used on chips at ~4ghz . I don't know the specifics of the process , but i sure hope it can clock as well as intel's .

[Vode]

12 minutes ago, Coaxialgamer said:

the only thing i am worried about is what clock speeds it will reach . Samsung's 14nm lpp was never used on chips at ~4ghz . I don't know the specifics of the process , but i sure hope it can clock as well as intel's .

No idea. I have a feeling it will get to at least 3,5Ghz gracefully. But we don't know the final clockspeeds for Zen anyways.. (feel free to correct me)

 

But there is no way to tell, the process node is different than the comparable Intel 8-cores, so is the cache system and a whole lot of other things.

[patrickjp93]

55 minutes ago, ScratchCat said:

Even then this is an engineering sample and is doing this well.

 

And there's not a doubt in my mind they rigged the results by using fewer memory channels on Broadwell, slower RAM, looser timings, or custom code (Blender is O-S) to make the results tip in its favor. Blender is extremely cache sensitive, so it's not remotely possible for AMD to be winning without trickery that will not be reflected in real life.

[ScratchCat]

Just now, patrickjp93 said:

And there's not a doubt in my mind they rigged the results by using fewer memory channels on Broadwell, slower RAM, looser timings, or custom code (Blender is O-S) to make the results tip in its favor. Blender is extremely cache sensitive, so it's not remotely possible for AMD to be winning without trickery that will not be reflected in real life.

True , lets hope not for the sake of competition

[Coaxialgamer]

Just now, patrickjp93 said:

And there's not a doubt in my mind they rigged the results by using fewer memory channels on Broadwell, slower RAM, looser timings, or custom code (Blender is O-S) to make the results tip in its favor. Blender is extremely cache sensitive, so it's not remotely possible for AMD to be winning without trickery that will not be reflected in real life.

to be fair though , they did say they used similarly configured machines . Whatever that means.

[handymanshandle]

4 hours ago, TheRandomness said:

I am completely guessing. An overclockable 4C/4T would be very nice sub £200. 

Ah okay. Well yeah under US$200 which is about £153.32.

[ScratchCat]

Just now, patrickjp93 said:

And there's not a doubt in my mind they rigged the results by using fewer memory channels on Broadwell, slower RAM, looser timings, or custom code (Blender is O-S) to make the results tip in its favor. Blender is extremely cache sensitive, so it's not remotely possible for AMD to be winning without trickery that will not be reflected in real life.

Still if it has a lot of cache this will help performance just like i7 5775c with its L4 cache. It was very close to the i7 6700k at 0.7ghz less

[patrickjp93]

9 minutes ago, Coaxialgamer said:

to be fair though , they did say they used similarly configured machines . Whatever that means.

And Intel said its Xeon Phi outperformed Nvidia's Maxwell on AI deep learning only to have Nvidia come out and prove they used an outdated (though stable) version. AlexNet is so buggy though I would never use it.

 

AMD is in all likelihood lying.

[patrickjp93]

13 minutes ago, ScratchCat said:

Still if it has a lot of cache this will help performance just like i7 5775c with its L4 cache. It was very close to the i7 6700k at 0.7ghz less

AMD has less cache than Intel, 20% less L3, and having a bigger L2 means having a slower L2 (cache interaction algorithms scale proportionally with size). The only other slightly legitimate way this could be possible is if AMD picked an image which could be chunked to have most of it fit in the lump sum of L2 data caches, but then that definitely won't reflect real world use cases.

[othertomperson]

21 hours ago, Prysin said:

Your counter argument is rather weak. Considering you have been proven wrong on so many occasions.

 

i think, the forums would be better without your misinformed, toxic and trollish posts.

lol triggered. I don't think I've said a single positive thing about Zen in either of these threads, and owned a two Intel products and six Nvidia products in my time and not a single one from AMD but SUCH AMD FANBOY OH NO

 

Fucking block me if you think "Fury X retailed for less than 980 ti where I am" is "toxic". Jesus fucking Christ. Look maybe I'm mis-remembering. If it helps you feel better assume that I am but the amount of people fucking shitting themselves over that statement is fucking hilarious. Does it even matter what two cards that no one should be buying any more cost a year or so ago? Really?

 

I think it says a lot about this forum tbh that you can't even be perceived as saying something good about AMD in a topic about a big announcement from AMD -- even when you actually aren't -- without being jumped on and called a fanboy.

[Prysin]

1 hour ago, othertomperson said:

lol triggered. I don't think I've said a single positive thing about Zen in either of these threads, and owned a two Intel products and six Nvidia products in my time and not a single one from AMD but SUCH AMD FANBOY OH NO

 

Fucking block me if you think "Fury X retailed for less than 980 ti where I am" is "toxic". Jesus fucking Christ. Look maybe I'm mis-remembering. If it helps you feel better assume that I am but the amount of people fucking shitting themselves over that statement is fucking hilarious. Does it even matter what two cards that no one should be buying any more cost a year or so ago? Really?

 

I think it says a lot about this forum tbh that you can't even be perceived as saying something good about AMD in a topic about a big announcement from AMD -- even when you actually aren't -- without being jumped on and called a fanboy.

no, i consider you the opposite. I consider you a pure blooded Green + Blue fanboy. Which is equally useless and unhelpful as the other type

[OptimisticRealist]

The more important question...

 

 

how much does the intel retail for??? They didnt even mention it...

 

$$$??? Intel?

[ace_cheaply]

8 minutes ago, OptimisticRealist said:

The more important question...

 

 

how much does the intel retail for??? They didnt even mention it...

 

$$$??? Intel?

http://www.newegg.com/Product/Product.aspx?Item=N82E16819117645

 

$1100, or you can get the older, better overclocking 5960x for $1015.

 

http://www.newegg.com/Product/Product.aspx?Item=N82E16819117404

 

Knock $100 off if you live near a microcenter. 

[79wjd]

2 hours ago, othertomperson said:

lol triggered. I don't think I've said a single positive thing about Zen in either of these threads, and owned a two Intel products and six Nvidia products in my time and not a single one from AMD but SUCH AMD FANBOY OH NO

 

Fucking block me if you think "Fury X retailed for less than 980 ti where I am" is "toxic". Jesus fucking Christ. Look maybe I'm mis-remembering. If it helps you feel better assume that I am but the amount of people fucking shitting themselves over that statement is fucking hilarious. Does it even matter what two cards that no one should be buying any more cost a year or so ago? Really?

 

I think it says a lot about this forum tbh that you can't even be perceived as saying something good about AMD in a topic about a big announcement from AMD -- even when you actually aren't -- without being jumped on and called a fanboy.

It only matters because you used their price over the last year to validate your argument that AMD will undercut ZEN by a meaningful amount as they have in the past.

 

 

On 8/18/2016 at 11:31 PM, djdwosk97 said:

There literally isn't even the slightest chance in hell that a cpu that's competitive with a 6900k ($1100) will be priced at $300. 

On 8/18/2016 at 11:41 PM, othertomperson said:

They did a similar demo showing how the Fury X was competitive with a 980 Ti, and then priced it around £100 less. I wonder why...

 

And maybe you did mis-remember, and that's fine. It's perfectly reasonable to remember something incorrectly, however there comes a point when you have to accept the fact that you're wrong. When you're showed multiple cards with similar pricepoints over the course of the  year you have to accept that you must have misremembered, not just discredit those sources because you remember otherwise. If you want to prove me (or basically everyone else) wrong, then go right ahead, but until you do, I'm the only one who's provided any actual proof of pricing. 

[Nineshadow]

I won't trust anything until we hear more from other reliable sources. AMD's marketing department is just "Hype, hype, hype! Let the fanboys do the rest". I've seen it way too many times already.

[RxL]

From "barely" edging past an old i5 4670k jumping to on par with a 6900k is a massive jump as amd leaks have shown both of these kind of performance from ZEN i think i will wait till legit 3rd party does some testing and releases some data on ZEN

 

Until then ZEN is just a hype trainwreck waiting to happen in my eyes

[Coaxialgamer]

Just now, RxL said:

From "barely" edging past an old i5 4670k jumping to on par with a 6900k is a massive jump as amd leaks have shown both of these kind of performance from ZEN i think i will wait till legit 3rd party does some testing and releases some data on ZEN

 

Until then ZEN is just a hype trainwreck waiting to happen in my eyes

that "leak" was wccftech though ...

[RxL]

Just now, Coaxialgamer said:

that "leak" was wccftech though ...

and that actually matters when this info out now is deemed to be a handicapped 6900k like everyone is saying but each to there own opinion

[Coaxialgamer]

Just now, RxL said:

and that actually matters when this info out now is deemed to be a handicapped 6900k like everyone is saying but each to there own opinion

they were not comparing the engineering sample to a handicapped 6900k . They were simply doing clock for clock comparisons .

[79wjd]

20 minutes ago, RxL said:

From "barely" edging past an old i5 4670k jumping to on par with a 6900k is a massive jump as amd leaks have shown both of these kind of performance from ZEN i think i will wait till legit 3rd party does some testing and releases some data on ZEN

 

Until then ZEN is just a hype trainwreck waiting to happen in my eyes

They were comparing the IPC to that of the 4670k. So a 4 core Zen CPU with an IPC similar to a 4670k would perform similarly to a 4670k while an 8 core Zen CPU with an IPC similar to a 4670k would perform similarly to a 5820k (this most recent rumor is reporting broadwell-ish IPC -- so only a slight improvement compared to the older rumors).