The Kyle Experience: Texas BBQ *spawn*

Yes, I'm sure you know just as much as anyone who purchases semiaccurate's $1000 yearly subscription.


People who actually know things tend to be unbiased and able to read press releases.

 

LOL, have you been around when I grilled Charlie, here? I guess you don't know because you are fairly new here.....

He asked me since I registered at his forums, why I didn't active my account, you know what I told him, it wasn't worth the time......

yes you can look back and find these posts here. And on top of this I don't even think Semiaccurate was around when the g80 launched.......

Good you can use a small font, PS where is that midrange statement from AMD again, Prior to March?

Too lazy to do the work or can't find it? Its one or the other?

 

You know that Polaris is an "underpowerd" GPU?

Please explain me again how you arrived at that conclusion , I'm "to lazy".

 

well get off your ass and do the math then. I'll give you question, lets use the Socratic method.

Why do you think, AMD is marketing at Midrange after Pascal was released, why not before?

Refute the information that Kyle stated, and take the focus off of the emotion in the article.

I not only could not which I tried, the part that stuck to me, when he talked about Intel and Raja, I heard that before, from a person inside AMD a close friend.

 

Yes yes, a "close friend" who's giving away critical and sensitive information about his company, so you can brag about it in forums and hurt his job by defaming said company.

Makes total sense. If he's an idiot.
(Or if you're doing it behind his back, making it two idiots in the whole scene.)

 

Razor there are 4 chips being release in a 6-12 month period by AMD, how are polaris 10 and 11 are anything but entry mainstream products. Your saying AMD is that incompetent that they are going to release a replacement for Polaris 10 after 6 months......

use you head.... jesus
and the talk of bring VR to the MAINSTREAM has been there from AMD from the very first time they talked about Polaris. Given that HARDocp has the worst possible color scheme for a dyslexic like me (litterly headache inducing) im not going to read his piece.

 

Where did you see mainstream before April of P10?

Link it, I have not been able to FIND it. Prior to that, they talked about TAM, in January, but never talked about which Polaris chips.

if we were to go by that, then Polaris should have been 970 competitor at a lower wattage? So you are thinking they wanted to replace the 390 at 50-100 bucks less? Wouldn't you think that would be fairly melodramatic, to talk about awesome perf/watt increase then that's all they can show?

You think that was the message AMD was going for?

I didn't see that either.

http://venturebeat.com/2016/01/15/a...-to-full-graphics-immersion-with-16k-screens/

This was the original article that people started talking about Polaris's performance.

P11 is the one they wanted for low end and mainstream

That was the system they were testing against the 950ti

So now this has changed, P10 for mainstream

 

February introduced the goal of increasing the breadth of VR by bringing the min-spec for VR down in price, which AMD defined as the 290X.
That seems unlikely to have been applied to Polaris 11, and if Polaris 10 in any iteration touches that lower bound, it becomes a challenge to see it going too far beyond AMD's high-end.
March brought the graphics roadmap and the infamous 2.5X power efficiency claim.

That number, the rumored die size of Polaris 10, similar graphics IP, and the VR goal all seem to fall decently in line. There's only so much those bounds can allow even within the scope of AMD's own product lines, so I am not sure what other set of expectations AMD would have had for Polaris.

 

when I first read the articles in January, it seemed like P10 were 390 to 390x replacement parts, I think many people took them that way. That could be the blame of the journalists..... but no one stated otherwise. There were many discussion about it being Fury X level performance, after April, that expectation went down to somewhere between 390x to nano, and sometimes reaching Fury X levels in very specific circumstances once AMD clearly defined their goals after April there should be no confusion about it now.

I agree to expect it to go beyond the performance level segment the size of these chips can't support that.

 

look in your own link!
http://venturebeat.com/2016/01/15/a...o-full-graphics-immersion-with-16k-screens/2/

So even if we are talking about Polaris 11, Polaris 10 is max 2.5 times Polaris 11. That puts it exactly where everyone except you and kyle are saying. So Razor how about you answer my question now!

edit: heres a good one:
http://arstechnica.com/gaming/2016/01/amd-confirms-high-end-polaris-gpu-in-development-for-2016/

 

Console level performance in thin and light note books,P11 low end, where does that put P11 that went against the 950ti? 950ti is a low end mid range part. So then you have 2 P10 still not being talked about there.

edit: heres a good one:
http://arstechnica.com/gaming/2016/01/amd-confirms-high-end-polaris-gpu-in-development-for-2016/[/QUOTE]

They are again talking about the P11 being in the mainstream segment. Where are the 2 P10's then, one to take on the gtx 960-970 (380, 390), and one to take on the 980 (390x)range. which puts them high end mid range to high end performance range.

yes both of those links are great, they support what I stated earlier, how expectations were higher before April.

They thought they were many months ahead with P11, they were never talking about P10, because they were only showing off P11. Which chip is in the AI for cars, the GP 106, they were comparing P11 to GP106.

 

To me everything confirms that AMD always intended for Polaris to the mainstream market (the largest TAM). I remember that Roy got some frak because he emphasized that Polaris would cheap. Vega will address the high-end.

I don't keep track of everything AMD said about Polaris, the only hint I got is this https://mobile.twitter.com/killyourfm/status/736448736256090112 and I only got that because he posted some ladies dancing

 

Why do we have a thread like this at Beyond3D in the first place?

 

Because it was stupid of me to not stop feeding the shills.
I'm sorry.


Please close the thread...

 

Yeah after you can't find anything of course you will say what you just said, fairly typical of you TotenTranz.

 

The Problem with that is, JPR's latest numbers the TAM



Here is your TAM, Performance is around the same as Mainstream last quarter..... So, how are they going to "increase" TAM when Performance seems to be growing and Mainstream is dropping? These increases are associated with certain performance cards were released or price cuts.

So your TAM, actually stays the same, just that AMD will take less margins and nV's margins are higher.

If we have these general numbers, AMD has had numbers for months before they know there TAM strategy is a deflection.

Edit: I shouldn't say stays the same, its a natural progression of what we have always seen, high end mainstream cards with performance level performance.

The main reason for AMD's strategy is not to increase TAM, its to try to stop what the gtx 970 did to them which the 1070 is now about to do, taking mainstream marketshare. Give that the price of VR systems and vr gear is still high ~2k for a total system, a 8% reduction in cost for end users is not enough to sway people to purchase a lower performing card for a niche market.

 

How are these segments even defined in the first place?

 

~150-~300 mainstream, ~300-~600 performance ~600 up enthusiast.

 

Is that how you define it, or how Jon defines it? I can't see him leaving out 0-150...

 

if mainstream also means 0-150, man Performance bracket is racking up major money.

The % of performance segment is equal to mainstream + value segments....

I just looked it up yes, he does put the Value segment in the Mainstream, that means, AMD missing the Performance segment, they absolutely screwed themselves if Jon's segments are same as AMD's/nV's segments.

Edit:

Yeah, what Jon considers, mainstream, is AIB mainstream + AIB Value segments. Performance, is AIB Performance, as Enthusiast is AIB Enthusiast.

That is quite surprising to see the performance segment as so big though.....