[Techpowerup] AMD "Zen" CPU Prototypes Tested, "Meet all Expectations"

[frozentundra123456]

Not this again. Saying it was the "most important" isnt really hype at all. In the context of all the hyperbole from AMD, nVidia, and Intel with each new release, it actually seems a rather innocuous statement. A certain poster just seems obsessed with it for some reason. The "importance" comes not from a huge performance boost, but from all the competition in the market place. Personally,I think it is a decent improvement, but I was hoping for more from the cpu side. The problem with Skylake is that it is just an improvement in the market where Intel is already dominant. What they needed was for Atom and the other SOCs for phone use to be killer products, and they certainly are not so far.

The thing that *does* irritate me is the incessant ads from Intel saying Skylake is "our best chip ever". I should hope so. A company with the resources of intel in the current market certainly cant afford to bring out a chip that is worse than it predecessor.

 

[cytg111]

Core 2.

Edit: I'd also say SB was a larger step over Nehalem than Skylake was over Haswell. We'll have to wait and see whether SL-E over HSW-E is a bigger step than SB-E was over Gulftown.

And Nehalem over Core2 ?

From SB and onwards, hasnt SL been the LEAST incremental upgrade of them all if you take highspeed DDR4 out of the equation?

 

[Fjodor2001]

Saying it was the "most important" isnt really hype at all.

Are you for real? Anyone not being a complete radicalized Intel fanboi realizes that calling Skylake "the most important chip architecture in a decade" is a complete joke. If you do not agree, please let us know what you base that on.

 

[Phynaz]

Core 2.

Edit: I'd also say SB was a larger step over Nehalem than Skylake was over Haswell. We'll have to wait and see whether SL-E over HSW-E is a bigger step than SB-E was over Gulftown.

Depending upon how you look at things, Core 2 was about a decade ago, so Skylake being the most important product since could still stand.

Note though, importance has nothing to do with how big of a step performance takes.

 

[Arachnotronic]

The thing that *does* irritate me is the incessant ads from Intel saying Skylake is "our best chip ever". I should hope so. A company with the resources of intel in the current market certainly cant afford to bring out a chip that is worse than it predecessor.

ROFL, so true :thumbsup:

 

[cytg111]

Note though, importance has nothing to do with how big of a step performance takes.

What?

edit: Much like religion you can post-rationalize whatever of whatever to be the greatest thing ever and much like religion it is probably best served in the privacy of your own home. When you DO take these rationalization 'deductions' and make em official on the interweb forums.. Just dont expect them to fly - or walk on water.

 

[Phynaz]

What?

What makes a chip important?

For Core 2 I beleive it was the radical departure from Netburst that made it an important chip. It enabled Intel to have the performance lead for 10 years.

For Skylake it's the scaling across platforms. A single cpu architecture that scales from 4W all the way to 130W. No other cpu architecture does that. I believe that is what makes Skylake Intel's most important cpu in a decade.

People confuse what is important to them with what is important to Intel.

 

[Sweepr]

Note though, importance has nothing to do with how big of a step performance takes.

Exactly. And Intel is most likely talking about the whole lineup of Skylake products. No previous product will scale from tablets to 8P high-end servers like this new family. Also, I'm not 100% sure Intel said that. Yes, the article mentions it, but who at Intel said this and when? Find me a quote.

“When I look at the range of what Skylake’s able to deliver from the Core M level all up to the i7 and Xeon, it’s just going to be a fantastic product,”

 

[Fjodor2001]

Also, I'm not 100% sure Intel said that.

https://en.wikipedia.org/wiki/Skylake_(microarchitecture)

No previous product will scale from tablets to 8P high-end servers like this new family.

Is that what defines what is the "most important architecture"? By what commonly agreed definition? And how much does it improve on those aspects over previous generations anyway?

Isn't it kind of like saying "this is the fastest CPU we've released so far"? Well, no sh*t Sherlock, otherwise why would you release it in the first place, instead of just keeping the old one?

 

[Arachnotronic]

Exactly. And Intel is most likely talking about the whole lineup of Skylake products. No previous product will scale from tablets to 8P high-end servers like this new family. Also, I'm not 100% sure Intel said that. Yes, the article mentions it, but who at Intel said this and when? Find me a quote.

You know Intel never actually said what the PC World article said and what Fjodor2001 keeps going on about.

What Skaugen said was that Skylake + Windows 10 represent a once-in-decade opportunity because there is a new OS rolling out in tandem with a new and improved chip architecture.

Just as an FYI, of course.

 

[moonbogg]

True that Sandy was, from a consumer standpoint, way more exciting. That 2600K was ridiculous and OC'd like a damn dream. Everyone freaked out and bought the crap out of those chips. Skylake doesn't have that kind of excitement behind it and nothing likely will for who knows how long.
People are recommending 4790K's over Skylake in some cases or they think its just about as good. No one would have said the same about 2600K vs any other CPU, period. The only other chip people were recommending at the time was the 2500K.

 

[Fjodor2001]

You know Intel never actually said what the PC World article said and what Fjodor2001 keeps going on about.

Then you should edit the Wikipedia article that disagrees with you.

 

[Sweepr]

You know Intel never actually said what the PC World article said and what Fjodor2001 keeps going on about.

What Skaugen said was that Skylake + Windows 10 represent a once-in-decade opportunity because there is a new OS rolling out in tandem with a new and improved chip architecture.

Just as an FYI, of course.

Thanks for clarifying. PC World must have misinterpreted it, so this false statement is just fuel for trolls.

 

[Fjodor2001]

Thanks for clarifying. PC World must have misinterpreted it, so this false statement is just fuel for trolls.

Yes, you know better than PC World how to interpret PC World's sources. Of course.

You should look in the mirror to find the troll.

 

[cytg111]

What makes a chip important?

For Core 2 I beleive it was the radical departure from Netburst that made it an important chip. It enabled Intel to have the performance lead for 10 years.

For Skylake it's the scaling across platforms. A single cpu architecture that scales from 4W all the way to 130W. No other cpu architecture does that. I believe that is what makes Skylake Intel's most important cpu in a decade.

People confuse what is important to them with what is important to Intel.

You mean 4.5 watts, not 2 watts right? 4.5 same as broadwell and 130watts vs haswell at 140.

So what we have here, in skylake, is a product that spans the capabilities of the prior and the prior-prior core product? Haswell was 2-3 years ago, - and today they have, with skylake - the same. When was broadwell, 1 year ago? - and today they have, with skylake - the same.

I get how they might be able to push margins higher on a single die - but i cant see them moving all that many dies alltogether .. cause its the same as yesterday and 3 years before that.

Of course, either way you put it, getting nitty gritty with the semantics, skylake IS the greatest CPU intel has ever produced.. its just not the greatest incremental upgrade they have ever produced.. If kabylake comes out at 0.001% ipc increase and nothing else then kabylake will be the greatest CPU intel has ever produced... Of course!

 

[Arachnotronic]

Then you should edit the Wikipedia article that disagrees with you.

I tried to follow the source links that Wikipedia quotes and the links are dead.

 

[Sweepr]

Yes, you know better than PC World how to interpret PC World's sources. Of course.

Can you provide the quote from someone at Intel where they said this? Otherwise I will believe that they misinterpreted what Skaugen said and you're basically repeating bullshit. Is Zen so uninteresting that you have to mention Intel at every opportunity?

 

[Arachnotronic]

Of course, either way you put it, getting nitty gritty with the semantics, skylake IS the greatest CPU intel has ever produced.. its just not the greatest incremental upgrade they have ever produced.. If kabylake comes out at 0.001% ipc increase and nothing else then kabylake will be the greatest CPU intel has ever produced... Of course!

Kaby Lake is going to be awesome because the media/display engines are getting some pretty sweet upgrades. 2x128mb eDRAM cache also points to some nice improvement in the GT4e graphics performance.

I am not 100% sure what to expect on the CPU side.

 

[Fjodor2001]

Can you provide the quote from someone at Intel where they said this?

Go edit the Wikipedia article if you disagree. I'm sure the other article contributors will love it.

 

[Fjodor2001]

I tried to follow the source links that Wikipedia quotes and the links are dead.

Then go ahead and edit the article if you disagree.

 

[cytg111]

Thanks for clarifying. PC World must have misinterpreted it, so this false statement is just fuel for trolls.

Hivemind much? have fun hun.

 

[Fjodor2001]

Kaby Lake is going to be awesome because the media/display engines are getting some pretty sweet upgrades.

Is that what you call "Awesome"!?

2x128mb eDRAM cache also points to some nice improvement in the GT4e graphics performance.

Just how much extra performance do you think that'll bring? Intel previously said anything above 64 MB did not bring any improvement:

http://www.anandtech.com/show/6993/intel-iris-pro-5200-graphics-review-core-i74950hq-tested/3

Regardless, it's only for niche products using GT4e. Most most people it's N/A.

I am not 100% sure what to expect on the CPU side.

Likely you can expect same uArch with minor 100 MHz frequency bumps. Disagree?

 

[Arachnotronic]

Go edit the Wikipedia article if you disagree. I'm sure the other article contributors will love it.

Here's the closest you're going to get to actually finding that quote, IMO:

"6th Gen Intel Core processors deliver some of the most significant advancements in computing that we've ever seen"

http://www.techtimes.com/articles/8...ing-its-core-m-processors-for-smartphones.htm

It's really hard to disagree with this statement. Skylake brought a big boost in 3D graphics, media encode/decode (Haswell/Broadwell sucked at this), integrated an ISP into the main die (first for a Core-class PC processor), and obviously CPU performance went up nicely. Oh and the packaging job they did on Core M Skylake was awesome (very small package, dies very tightly packed together) particularly for tablet-like form factors.

Skylake, when taken as a total package, was a pretty big jump gen/gen. Just because all you focus on is CPU performance in high power envelopes doesn't mean that this is what Intel execs are referring to.

 

[cytg111]

Kaby Lake is going to be awesome because the media/display engines are getting some pretty sweet upgrades. 2x128mb eDRAM cache also points to some nice improvement in the GT4e graphics performance..

Is this enough to make it "the greatest chip" intel ever produced - in your interpretation of the words?

 

[Arachnotronic]

Is that what you call "Awesome"!?

For a lot of people, these improvements will be a big deal. For others, not so much.

Only for niche products using GT4e. Most most people it's N/A.

It's going to be great for a 2017 15-inch MacBook Pro, 21.5-inch iMac (and maybe even entry level 27-inch iMac), and it could even enable some pretty thin-and-light gaming capable PCs that don't require low-end discrete GPU. As these iGPUs get better, they will become less niche

Likely you can expect same uArch with minor 100 MHz frequency bumps. Disagree?

I'm not so sure.