Official? Ryzen leaked slides

[Drak3]

4 minutes ago, Sauron said:

Not to the point where you should spend 50 bucks more.

Sometimes the factor is performance, period. Not price to performance, not price.

When the best stock performance is required for the task, it doesn't matter if it's more expensive or has slightly worse price/performance.

[zMeul]

5 minutes ago, Drak3 said:

-

yes, yes .. distort the truth as much as you can

typical AMD tactic

[zMeul]

anyways ...

IPC of Zen seems to be lower or on par with that of Haswell - just as I and patrick said for many months now

[leadeater]

1 minute ago, LAwLz said:

This is the biggest cop-out argument I have ever seen...

 

I think it is pretty mind blowing that people actually think "official benchmarks" (that were also leaked) mean anything. Because no company has ever lied and/or inflated their own numbers right? On top of that, these are not even official from what we know. So there is a double layer of "you'd have to be an idiot to trust this" at play here.

This holds as much credit as someone over at Reddit going "yeah I got one, and it's super fast"... But knowing this forum and how delusional people are, it would probably be posted here as a news thread and people would go "I knew it! Ryzen confirmed for second coming of Jesus!".

How is it a cop-out? I literally don't listen to marketing from AMD, Intel or Nvidia. Why do I do this, for the reasons you just stated. Don't confuse an argument with a statement, I legit never heard anything about Fury X being an overclocker's dream.

[The Sloth]

7 minutes ago, Sauron said:

Not to the point where you should spend 50 bucks more.

more money= better right? i really don't know why people do it, just accept it and move on. 

[Techstorm970]

12 minutes ago, Sauron said:

but then they shouldn't buy a K skew cpu anyway.

Actually, being built for overclocking means being built like a tank.  Not only are K-skew CPUs overclockable, but they are also extremely reliable.

[VagabondWraith]

10 minutes ago, Drak3 said:

 

In reference to the 500W cooling dissipation capable loop and power delivery system, they call these components (actually, specifically the cooler) 'an overclocker's dream.' They never called the Fury X itself a dream. Even then, the Fury X overclocking was artificially held back via the BIOS available at launch. After a series of updates, cards like the reference Fury X and the Sapphire Nitro could see damn near a 50% OC, which is significantly better than the +100MHz that an RX 480 might see over it's 1266MHz boost, than the +130 MHz over what a GTX 1080 will hit before running into issues with the power draw limit.

 

It's jackasses taking what these companies are saying out of context, and then later claiming it is as AMD said (when, in fat it's not), that have been accelerating AMD's acquisition of the image of being the cut rate, budget solution that is only viable at the bottom end.

First party benchmarks hold weight when the companies are under constant scrutiny (especially after the creative interpretation publications and potential consumers are jumping to with what little REAL information AMD gave us). Especially as AMD's results, not necessarily the numerics of each component (FPS being the big one), but the % differences are being replicated, when tested as AMD stated.

"You'll be able to overclock this like there's no tomorrow." "This is an overclockers dream."

 

If you consider 50MHz an "enthusiasts dream" I don't know what else to tell you. 

 

At 49:30 you can hear what the engineer has to say.

 

 

[Taja]

7 minutes ago, Drak3 said:

Sometimes the factor is performance, period. Not price to performance, not price.

When the best stock performance is required for the task, it doesn't matter if it's more expensive or has slightly worse price/performance.

If that is the case, it is stupid af to not OC.

 

If you want performance, buy k and overclock.

If not, dont buy k.

Buying k and not overclocking just because of the boost does not make sense really.

[leadeater]

5 minutes ago, techstorm970 said:

Actually, being built for overclocking means being built like a tank.  Not only are K-skew CPUs overclockable, but they are also extremely reliable.

In what way are the more reliable, considering they are taken from the same silicon wafer and packaged in the same way? Other than microcode allowing us to change the multiplier and passing a higher verification they are manufactured in the same exact way.

[Drak3]

1 minute ago, zMeul said:

yes, yes .. distort the truth as much as you can

typical AMD tactic

You don't see me saying NVidia is a lying group of corporate assholes, who we should hang, when they were advertising '2X performance of the previous generation" when they released Pascal, do you?

Here's why: in the fucking "fine" print, in the discussion that was being made, it was about VR performance. And that was proven true. If one actually observed what was said, in it's entirety, without buying into the creative interpretations third party publications like to make, then there isn't a problem. These companies deliver on what THEY said. Not what WCCF said, not what PCPer said, not what Anandtech said, not what you said. What AMD themselves said.

 

AMD NEVER said that the Fury X was an overclocker's dream. They said the cooling solution was. Even if they did, a BIOS update made the statement fairly reliable.

AMD Never said that the RX 480 crossfire configuration was better than a single GTX 1080 100% of the time. They said that it was under a single scenario. That was proven true.

 

Most of AMD's statements, that are touted as 'lies,' have been true WHEN ONE ACTUALLY OBSERVES THE FULL STATEMENT. Not the cherry picked bullshit that people (including you) LOVE to do.

1 minute ago, Taja said:

If that is the case, it is stupid af to not OC.

 

If you want performance, buy k and overclock.

If not, dont buy k.

Buying k and not overclocking just because of the boost does not make sense really.

One can argue that overclocking is stupid as all fuck too. It wastes more power, requires a more powerful (and thus more expensive, even if by $5) cooling solution, a higher end (again, more $$$) chipset, and if not done correctly, can cause system instability/

 

Just because you use K SKU CPUs to overclock doesn't mean someone else just needs the stock raw performance for their task.

[Techstorm970]

3 minutes ago, leadeater said:

In what way are the more reliable, considering they are taken from the same silicon wafer and packaged in the same way? Other than microcode allowing us to change the multiplier and passing a higher verification they are manufactured in the same exact way.

Each flagship mainstream i7 processor is tested before hitting the market.  The good overclockers become K-skews, the bad ones become non-K.

[Drak3]

6 minutes ago, VagabondWraith said:

"You'll be able to overclock this like there's no tomorrow." "This is an overclockers dream."

 

If you consider 50MHz an "enthusiasts dream" I don't know what else to tell you. 

 

At 49:30 you can hear what the engineer has to say.

 

 

Given that the conversation at that point is about the cooler (a brief mention of the power delivery system thrown in), and not the actual performance and potential of the card, all you're doing is taking AMD's statement out of context, and falsely accusing them of lying because of your refusal to acknowledge the full conversation.

[leadeater]

1 minute ago, techstorm970 said:

Each flagship mainstream i7 processor is tested before hitting the market.  The good overclockers become K-skews, the bad ones become non-K.

Yes which is what I said, however they are built exactly like a non-K so a K SKU cannot be "built like a tank" and the non-K not.

[Drak3]

1 minute ago, techstorm970 said:

Each flagship mainstream i7 processor is tested before hitting the market.  The good overclockers become K-skews, the bad ones become non-K.

Overclocking =/= reliability.

 

Every i7 has a minimum stability bar to meet. A K SKU can just barely pass that bar at stock speeds, and a non-K SKU can exceed that bar by a huge margin before hitting the stock speed of the K SKU.

[Droidbot]

Pack it up boys, Intel's killed Zen already. 

[VagabondWraith]

1 minute ago, Drak3 said:

Given that the conversation at that point is about the cooler (a brief mention of the power delivery system thrown in), and not the actual performance and potential of the card, all you're doing is taking AMD's statement out of context, and falsely accusing them of lying because of your refusal to acknowledge the full conversation.

I didn't say they were lying, I was saying it's not an "overclockers dream." When I think of overclocking, I think of pushing something way past it's stock limits, not how beefy a MOSFET is or that it has a water cooling loop. That means nothing to me.

 

"It can overclock like there's no tomorrow," "this is an overclockers dream." That applies to overclocking. If you consider 50MHz some wonderful feat, then Pascal must be something beyond extraordinary. (Which btw, is what I consider overclocking.)

[Taja]

1 minute ago, Drak3 said:

 

One can argue that overclocking is stupid as all fuck too. It wastes more power, requires a more powerful (and thus more expensive, even if by $5) cooling solution, a higher end (again, more $$$) chipset, and if not done correctly, can cause system instability/

 

Just because you use K SKU CPUs to overclock doesn't mean someone else just needs the stock raw performance for their task.

Well,one would be ignorant to say that.

About the points you made:

 

Power draw is irrelevant. A bath will waste more power than the extra watts of the oc will in a year.

 

The cooling solution is also irrelevant: a k cpu will not come with a cooler, so you will have to buy one anyway. If you get a bad cooler, you will get worse temps and noise, but thats it.

 

The chipset part is real, but your own argument was "performance, not price x perf", SO by YOUR condition, the person would get the better chipset.

 

REAL WORLD CONCLUSION:

Want to oc? Go for k.

Dont want to oc? Dont go for k.

 

Its that simple, people can defend going for k and not overclocking, but it is not the best and most logical option. If you WANT to do that, go ahead. You can also buy a titan black in 2017 and pair it with a fx 4300, that is YOUR choice, but dont go arround saying it is a good choice "because". 

 

 

 

[HelplmChoking]

I wouldn't water cool Ryzen, too leaky.

[The Sloth]

1 minute ago, Rangaman42 said:

I wouldn't water cool Ryzen, too leaky.

controlled leaks for marketing 

[NumLock21]

Powered by AMD Ryzen

 

 

Power by Intel

 

 

[dizmo]

Oh yay. More Cinebench benchmarks. Why is this the only one we see? Where are the results when it's teamed up with GPUs for gaming? Where are other benchmarks? It seriously makes me think that this is the only benchmark that AMD excels at.

I'm all for news and advances but even I'm starting to get sick of seeing the same thing over and over again.

Can't wait for this to be over.

 

1 hour ago, Jed M said:

Intel is (hopefully) fucked.

Why would you hope that?

1 hour ago, vorticalbox said:

They won't be close to fucked. Remember how chrome came out as this super light wright fast browser and people still think that enough they it eats so much ram and is so heavy?

What sense does that comparison make?

 

[HelplmChoking]

1 minute ago, nerdslayer1 said:

controlled leaks for marketing 

Yeah, I know how "leaks" work. It's just amusing how many there are.

[The Sloth]

1 minute ago, NumLock21 said:

Powered by AMD Ryzen

 

 

Power by Intel

 

 

 

this joke 

[NumLock21]

2 minutes ago, nerdslayer1 said:

this joke 

 

 

And here's Microsoft

[vorticalbox]

1 hour ago, Enderman said:

I don't think you understand why people use chrome.

Hate to break it to ya, but it's not because "oh its super light and fast".

maybe not now put a lot of people I speak to still think chrome is light weight.

 

I haven't used chrome from google in a while, I like opera better free unlimited VPN and ad block built in.