Interview With AMD's Desktop Processor Marketing Manager, Don Woligroski

[Virtual01]

1 hour ago, Syntaxvgm said:

Thanks. 

Misleading graphs are OK cause you'd need to be retarded to fall for it, and if you are mentally retarded, you likely live in the US. Real classy.

Honestly dude, as a retail worker in an electronics store, he's 100% on the money. People ask me every day what something costs when every price tag is right next to or right below every item. People ask me what brand a computer is when it's in very obvious contrasting font right on the front. People walk into the store and ask me where the TVs are despite the fact that the entire back wall is plastered with huge, bright TVs at all times. I'm betting that a majority of normal consumers would be fooled by that graph.

[Syntaxvgm]

2 minutes ago, Virtual01 said:

Honestly dude, as a retail worker in an electronics store, he's 100% on the money. People ask me every day what something costs when every price tag is right next to or right below every item. People ask me what brand a computer is when it's in very obvious contrasting font right on the front. People walk into the store and ask me where the TVs are despite the fact that the entire back wall is plastered with huge, bright TVs at all times. I'm betting that a majority of normal consumers would be fooled by that graph.

I assume that would apply almost anywhere though, not just the US. About the graph, I'm not saying non-retards would fall for it, I'm just saying a graph should be...you know...a visual representation of the data...

[Virtual01]

7 minutes ago, Syntaxvgm said:

I assume that would apply almost anywhere though, not just the US. About the graph, I'm not saying non-retards would fall for it, I'm just saying a graph should be...you know...a visual representation of the data...

I agree but we all know that marketing is just another word for lying.

[spartaman64]

2 hours ago, Syntaxvgm said:

TIL I'm retarted for making fun of stuff like this 

and that doesnt mislead me because they labeled the axis. it only misleads people who cant read and dont know kindergarten math

and the amd one is even less misleading because they put a big 40% IPC improvement next to the line on the example above maybe some people dont read the axis but it is almost impossible to miss the 40% ipc improvement on the amd graph

[dalekphalm]

While I personally expect Zen to deliver, and I'm very excited for it....

 

 

This interview literally told us nothing. Nothing at all. The guy is head of Marketing. It's all PR speak. Not that I expected the PR guy to really give us anything too juicy in a random Interview. They'll save the juicy bits for some big reveal at a Press Release Conference at one of the big Computer Trade Shows, like E3 or something.

[79wjd]

1 hour ago, spartaman64 said:

and that doesnt mislead me because they labeled the axis. it only misleads people who cant read and dont know kindergarten math

and the amd one is even less misleading because they put a big 40% IPC improvement next to the line on the example above maybe some people dont read the axis but it is almost impossible to miss the 40% ipc improvement on the amd graph

All that means is that IPC imroved 40%. Total performance could still have increased more. So, no the graph is misleading. Not to mention:

1 hour ago, Syntaxvgm said:

.... a graph should be...you know...a visual representation of the data...

Graphs were created to make data easily understandable at an instantaneous glance. Marketing departments all just use misleading graphs because it's lying without technically lying.

[ApolloFury]

3 hours ago, Syntaxvgm said:

 

 

lol wut. Is it referring to vending machines being broken or people die during the use of vending machines?

[spartaman64]

1 hour ago, djdwosk97 said:

All that means is that IPC imroved 40%. Total performance could still have increased more. So, no the graph is misleading. Not to mention:

Graphs were created to make data easily understandable at an instantaneous glance. Marketing departments all just use misleading graphs because it's lying without technically lying.

the graph is only about ipc it doesnt say anything about performance

[Syntaxvgm]

1 hour ago, ApolloFury said:

lol wut. Is it referring to vending machines being broken or people die during the use of vending machines?

I think shaking them and having them fall on top of you. A few people a year die from that. 

[Syntaxvgm]

18 minutes ago, spartaman64 said:

the graph is only about ipc it doesnt say anything about performance

This is a good point to emphasize: I currently don't doubt the 40% ipc gain, but I love the graphs making 40% look bigger than it is. Marketing, NVIDIA and intel do it too 

[shdowhunt60]

2 minutes ago, Syntaxvgm said:

This is a good point to emphasize: I currently don't doubt the 40% ipc gain, but I love the graphs making 40% look bigger than it is. Marketing, NVIDIA and intel do it too 

You mean where they showed Summit Ridge being twice the performance of Orochi? Honestly given the 5960X's multi-threaded performance relative to a 8350, which is something like twice as fast on a conservative estimate, then that's probably not as misleading as one would think if Zen's going to be somewhere around that level of performance.

[Syntaxvgm]

11 minutes ago, shdowhunt60 said:

You mean where they showed Summit Ridge being twice the performance of Orochi? Honestly given the 5960X's multi-threaded performance relative to a 8350, which is something like twice as fast on a conservative estimate, then that's probably not as misleading as one would think if Zen's going to be somewhere around that level of performance.

I don't know. I'm skeptical that it will be at that level, but I'm still hopeful for the fresh start. I really don't want to judge their performance until it comes out, and if it's good enough I'll probably make an AMD system. I have a 5820k in my gaming system and zen doesn't have to necessarily beat it in IPC to win me over, as if the per core performance is good enough, 8 cores would still make for a good workstation part even if they still weren't caught up to intel, as the price will likely be much better. Not to mention SMT, they've already confirmed 16 threads, so I'm happy about that already. 

[shdowhunt60]

2 minutes ago, Syntaxvgm said:

I don't know. I'm skeptical that it will be at that level, but I'm still hopeful for the fresh start. I really don't want to judge their performance until it comes out, and if it's good enough I'll probably make an AMD system. I have a 5820k in my gaming system and zen doesn't have to necessarily beat it in IPC to win me over, as if the per core performance is good enough, 8 cores would still make for a good workstation part even if they still weren't caught up to intel, as the price will likely be much better. Not to mention SMT, they've already confirmed 16 threads, so I'm happy about that already. 

I don't see why it couldn't. Even the most conservative estimates puts Zen right at Sandy Bridge performance, and lord knows Haswell isn't much faster than that. My chief concern is memory performance rather than IPC, there's no mention what kind of memory controller Zen is going to have other than DDR4. 

[Syntaxvgm]

9 minutes ago, shdowhunt60 said:

I don't see why it couldn't. Even the most conservative estimates puts Zen right at Sandy Bridge performance, and lord knows Haswell isn't much faster than that. My chief concern is memory performance rather than IPC, there's no mention what kind of memory controller Zen is going to have other than DDR4. 

my worry about the performance estimate putting it right at sandy will really hurt it in headlines, when I don't think winning strait performance out of the gate is the measure of success here. Given that bulldozer was a pile of shit and they still improved on that somewhat, Zen will likely get better. Not to mention I gamed on 2600 non-k until recently and still use it as a workstation. Still good performance today with sandy. If AMD can offer their new chip at significantly less than whatever 1151 quad core intel is offering at the time of Zen's launch, or the 2011-3 6 cores around the same price,  I think it will be a serious choice, especially for those not going for purely gaming.

And god I hope they go with typical AMD tradition and release a super power hungry balls to the wall version, like the 9590....but good. Would love an excuse to make a custom loop again. Unfortunately that would mean another situation with insufficient power delivery on motherboards with AMD sockets.

http://www.overclock.net/t/946407/amd-motherboards-vrm-info-database

[Sakkura]

3 hours ago, Energycore said:

I'm glad there's someone actually talking about the interview.

 

The head of marketing you say? That's the guy who needs to step it up or be replaced by Nvidia's Marketing Director.

 

Read the Interview, really did contain nothing new that we didn't know from elsewhere. A shame.

He's marketing manager for desktop processors. Their desktop processors are all pretty bad, so what's he supposed to do?

[Syntaxvgm]

5 minutes ago, Sakkura said:

He's marketing manager for desktop processors. Their desktop processors are all pretty bad, so what's he supposed to do?

ITLL BE BETTER I PROMISE BUT I CANT TELL YOU ANYTHING YET ITS ALL UNDER NDA BUT I PROMISE JUST BELIEVEEEE

[Energycore]

1 hour ago, Sakkura said:

He's marketing manager for desktop processors. Their desktop processors are all pretty bad, so what's he supposed to do?

Fair enough.

 

But we need some talent on the marketing side. And hopefully Zen comes out better marketed than Polaris has been.

[deviant88]

this thread man :))) its hilarious

[Doobeedoo]

Just want actual official info already!

[BOT Zee]

That graph is so retarded.... wow hahahaha who starts a graph at 0.9 makes it look like 4x the performance if youre far away from the stage and the font is size 5 ffs... lmao bs marketing hype. Obvs if you look at the numbers is 40% better but HONESTLY start your fucking graph at 0 for us to tell how fast it is based on the WHOLE BARS NOT SOME OF THE BARS... thats like this

 

.........

......

 

vs.

 

........

..

 

ITS MISLEADING.

[Atmos]

19 hours ago, Syntaxvgm said:

I'm so excited I made a misleading marketing graph. 

Oh wait they beat me to it damn

I wonder if they realize that its now half way through 2016, their little graph there appears to show zen launching in the beginning of 2016, and hell... I love amd, but they are not that- er- skilled with graphs. 

[patrickjp93]

15 hours ago, djdwosk97 said:

All that means is that IPC imroved 40%. Total performance could still have increased more. So, no the graph is misleading. Not to mention:

Graphs were created to make data easily understandable at an instantaneous glance. Marketing departments all just use misleading graphs because it's lying without technically lying.

Total performance of some workloads may not have improved at all if we're just discussing General IPC. That won't be the case because AMD is adding a micro-op cache finally, but let's stop assuming things until the chip reviews come out.

[Tomsen]

40 minutes ago, patrickjp93 said:

That won't be the case because AMD is adding a micro-op cache finally

They had that since steamroller.

[Eroda]

its all simple wait for Zen to drop stop speculating on fresh air grain of salt BS and then all will become clear

[79wjd]

2 hours ago, patrickjp93 said:

Total performance of some workloads may not have improved at all if we're just discussing General IPC. That won't be the case because AMD is adding a micro-op cache finally, but let's stop assuming things until the chip reviews come out.

I'm not assuming anything. I was merely pointing out how the graph is technically misleading.