Another 6 months before VEGA?

Remember when people were sure Vega was coming in September last year? Pepperidge Farms remembers.

If I wasn't so lazy I'd dig up my posts and a few other veterans here who know which way the wind is blowing and called this early last year when Polaris got pushed back from the rumored date. But I'm lazy, so I won't.
 
Creig said:
You need to take a step back and realize that the spending of the members of hardware junkie sites like [H] do not reflect in any way whatsoever the spending of the general public. Nearly every person here has significantly higher system specs than the average gamer. You may as well read through a Ferrari forum and conclude that the average consumer spends at least $200,000 on their car.

The cost of R&D certainly does change depending on the product. Do you honestly believe it costs just as much to design the 601 mm², 12 billion transistor Titan X as the 232 mm², 5.7 billion RX 480? Come on.

As far as "want to see AMD just go belly up", somehow AMD stock rose 280% in 2016 even without releasing their own version of a "Titan". They're certainly doing a lot better than they have in quite a while. Right now I'd say they're doing what they should be doing. And that's concentrating on getting out products that the majority of the population wants to buy.
Click to expand...

I need to take a step back?

Sorry but I'm not getting my numbers from here, I'm getting them from JPR and Mercury lol

The cost of R&D doesn't change, because do you know what the difference is between feature R&D and manufacturing cost?

I don't give a shit about their stock price. Stock price is based on speculation of AMD doing better, not about its bottom line. And its bottom line is still in the red where AMD CEO's stated they will be going to be in the black, which looks highly unlikely. Their balance sheet is still quite unbalanced just less unbalanced than before. Which they can't spend more money in things like R&D until they go back in black.
 
harmattan said:
Remember when people were sure Vega was coming in September last year? Pepperidge Farms remembers.

If I wasn't so lazy I'd dig up my posts and a few other veterans here who know which way the wind is blowing and called this early last year when Polaris got pushed back from the rumored date. But I'm lazy, so I won't.
Click to expand...


That is the problem people believe AMD is in a good spot because of their stock price and what they say, or believe in crazy ass rumors because it makes them feel good about AMD, the truth is AMD is in such a bad position, its like a grape in the middle of the road and herd of elephants coming down that road. AMD has some reprieve with Zen but don't expect that to hold on if they can't get their shit together.
 
razor1 said:
I need to take a step back?

Sorry but I'm not getting my numbers from here, I'm getting them from JPR and Mercury lol
Click to expand...
Good for you. And what do your numbers say? I'm willing to bet they say there are more mainstream/low end cards sold than high end/bleeding edge cards.

razor1 said:
The cost of R&D doesn't change, because do you know what the difference is between feature R&D and manufacturing cost?
Click to expand...
I also know the difference between feature R&D and die/wafer R&D. And I know it costs more to design a large die, high transistor count Titan X than a small die, lower resistor count RX480.

razor1 said:
I don't give a shit about their stock price. Stock price is based on speculation of AMD doing better, not about its bottom line. And its bottom line is still in the red where AMD CEO's stated they will be going to be in the black, which looks highly unlikely. Their balance sheet is still quite unbalanced just less unbalanced than before. Which they can't spend more money in things like R&D until they go back in black.
Click to expand...
It isn't just speculation. AMD is doing better.

In 2014 and 2015, AMD’s revenue took a downward path as it lost discrete GPU market share to NVIDIA (NVDA) and PC processor market share to Intel (INTC). In 2016, AMD aims to regain the market share it lost to NVIDIA by targeting mainstream customers. It has so far succeeded in increasing its share from 21% in 4Q15 to 30% in 2Q16, according to Jon Peddie Research.
Click to expand...

And as far as AMD's revenue growth goes:

7-12.png
 
Creig said:
Good for you. And what do your numbers say? I'm willing to bet they say there are more mainstream/low end cards sold than high end/bleeding edge cards.


I also know the difference between feature R&D and die/wafer R&D. And I know it costs more to design a large die, high transistor count Titan X than a small die, lower resistor count RX480.


It isn't just speculation. AMD is doing better.



And as far as AMD's revenue growth goes:

7-12.png
Click to expand...

That is AMD's expectation that is not what financial institutes are saying. 2 and 1 Q's ago, AMD CEO stated in their Financial call, Q1 they will be in the black, and now they are still in the red by their own estimates!

What does that tell us, they don't know how things are going to shape up just a Quarter away, how can you even believe what their projections are?
 
razor1 said:
That is the problem people believe AMD is in a good spot because of their stock price and what they say, or believe in crazy ass rumors because it makes them feel good about AMD,
Click to expand...
You're the only person I've seen who is saying that. Everybody with even half a brain knows AMD is struggling and has been struggling for years. They are doing better now than they have been in the past, but they still have a long way to go.
 
razor1 said:
That is AMD's expectation that is not what financial institutes are saying. 2 and 1 Q's ago, AMD CEO stated in their Financial call, Q1 they will be in the black, and now they are still in the red by their own estimates!

What does that tell us, they don't know how things are going to shape up just a Quarter away, how can you even believe what their projections are?
Click to expand...
Did you not just see the JPR numbers I posted? I know you love JPR numbers so I thought that would appease you.

Just face facts. AMD did a heck of a lot better in 2016 than it has for a number of years now. And they did it without releasing an uber-expensive Halo effect super-card. Hopefully they will continue to do better in 2017 as well.
 
Creig said:
You're the only person I've seen who is saying that. Everybody with even half a brain knows AMD is struggling and has been struggling for years. They are doing better now than they have been in the past, but they still have a long way to go.
Click to expand...


LOL ok lets see you want links to everything they have stated, how much crap comes out of their mouth?

I'm not the only person saying these things, AMD has been saying everything and I'm going to be reiterating them. I'll ask you again, do you want me to link them? They are easy to find.....

If they don't take into account what Intel is doing, what nV is doing into their financial calls, that is the only way their numbers would turn the way they will.
 
Creig said:
Did you not just see the JPR numbers I posted? I know you love JPR numbers so I thought that would appease you.

Just face facts. AMD did a heck of a lot better in 2016 than it has for a number of years now. And they did it without releasing an uber-expensive Halo effect super-card. Hopefully they will continue to do better in 2017 as well.
Click to expand...


What numbers, you took an excerpt of their report, which has nothing to do with reality, it was AMD was hoping for, which didn't HAPPEN.

AIB-graph2.PNG


Anyone with any common sense could see AMD was talking out of their ass.

Performance is bigger than mainstream and value COMBINED! (JPR mainstream includes both mainstream and value segments)
 
NKD said:
thats actually a good thing. Lot of monitors were coming out with shitty ranges! I think controlling LFC is a good thing. I would rather pay a little more for that.
Click to expand...


Yes it is, better components for better features, just increases cost, how much don't know yet, but I don't expect it to be lower then 50 bucks.
 
Just for you Creig

http://www.pcworld.com/article/2919...s-to-return-to-profitability-by-year-end.html

http://www.reuters.com/article/us-amd-outlook-idUSKCN0YN36Y

http://finance.yahoo.com/news/u-chipmaker-amd-expects-return-093044241.html

How many sites reported this when Lisa stated it in Q2 of last year, they expected to return to profitability in 2nd half of 2016! Then it shifted to Q1 of 2017, now its Q2 of 2017. Just keeps getting pushed out.

Why, something going on with their product line up or something else happening in the market they aren't talking about? Like what Intel and nV has been doing?
 
razor1 said:
Just for you Creig

http://www.pcworld.com/article/2919...s-to-return-to-profitability-by-year-end.html

http://www.reuters.com/article/us-amd-outlook-idUSKCN0YN36Y

http://finance.yahoo.com/news/u-chipmaker-amd-expects-return-093044241.html

How many sites reported this when Lisa stated it in Q2 of last year, they expected to return to profitability in 2nd half of 2016! Then it shifted to Q1 of 2017, now its Q2 of 2017. Just keeps getting pushed out.

Why, something going on with their product line up or something else happening in the market they aren't talking about? Like what Intel and nV has been doing?
Click to expand...
What are you ranting about now? I never mentioned anything about what AMD or Lisa Su "predicted" would happen. And I don't really care about their predictions.

All I said was that I think AMD is doing better than they have in the past and that they don't need a high end video card right now.

Change the subject much?
 
Creig said:
What are you ranting about now? I never mentioned anything about what AMD or Lisa Su "predicted" would happen. And I don't really care about their predictions.

All I said was that I think AMD is doing better than they have in the past and that they don't need a high end video card right now.

Change the subject much?
Click to expand...


Where the hell do you think that chart came from? you know the one that you posted?

Its from AMD!

Who is saying it?

AMD!

Am I changing the subject?

No I'm expounding upon the way AMD has been dragging their feet, and pretty much making investors fools.

Short term stock fluctuations has NOTHING to do with the health of AMD, it has everything to do with future potential based on speculation on what is going on inside of AMD which at this point has nothing do with reality or based on what AMD has stated in financial calls for the past few years, I can go back to 2015, and 14, and 13 when they have stated they will come back to profitability within the fiscal year of making such comments, and yet, nothing has materialized.

The only thing that looks good from AMD is Zen, and without it, AMD is dead in the water. So speculation of this is giving AMD's stock price a boost, but that too has been dragging out, from Q4 of 2016, to Q1 of 2017, and now Q2 OF 2017, the same dragging on AMD has been doing for years.

Investors don't know what Zen is about yet, they only see potential, as do we. But Potential has too many variables that are not in AMD's control.

What you are basically doing, is saying what ever AMD has stated so far is going to happen.

What I'm saying is that is fool hardy because we have seen AMD make these statements in the past which have NOT COME TRUE. And they won't come true unless Intel and nV play dead.
 
Last edited:
razor1 said:
Where the hell do you think that chart came from? you know the one that you posted?

Its from AMD!

Who is saying it?

AMD!

Am I changing the subject?

No I'm expounding upon the way AMD has been dragging their feet, and pretty much making investors fools.
Click to expand...
WTF? The graph I linked was put together by Market Realist, an investment and analytic website. Not AMD. And it was based on AMD's SEC filings. Not anything said by Lisa Su.

http://marketrealist.com/2016/12/drove-amds-stock-cy-2016/

Christ...

Take a chill pill.
 
Creig said:
WTF? The graph I linked was put together by Market Realist, an investment and analytic website. Not AMD. And it was based on AMD's SEC filings. Not anything said by Lisa Su.

http://marketrealist.com/2016/12/drove-amds-stock-cy-2016/

Christ...

Take a chill pill.
Click to expand...


I'm not talking about revenue, I'm talking about net profits which is a direct results of margins. Revenue is only one metric that goes into company health, that is not a good way to see things.

now your chart that you linked, did you even read the report? why did AMD's revenue go up? Because of Semi custom designs, which only happen once ever 3 years? The is probably the influx of money from Sony and MS to design their next gen chip, then after that, it all goes away back to where they were before. Hence why guidance for Q1 for 2017 is still negative.

You can sit here and think everyone is going to sit around and believe you when the article itself states the same thing.


As the seasonal effect fades in fiscal 4Q16, the company expects revenue to fall sequentially. However, it expects profit margins to improve in fiscal 4Q16 as it realizes licensing gains and improves operational efficiency.
Click to expand...


It wasn't seasonal effects as Q4 should be strong in electronic devices. Its the money Sony and MS gave to AMD for development purposes of next gen products.

And we know AMD's margins didn't improve Q4 2016, They went down!

31% to 30%!

So in Q1 of 2017, if their margins stay the same or return to 31%, and what this report is saying margins will go up how do you expect them to they will be in the red in Q1?

AMD only needs to get to 35% margins to get into the black.

I say only because their margins are piss poor much lower then any other profitable company in any industry, even companies without huge R&D needs still need margins @ or above 40% to stay afloat.

this is what you are not understanding and many investors don't either.

Operating margins that are lower than 40% when a company needs to grow (which is what state AMD is in right now with less than 40% margins), they can't grow, the cost of growth comes at the cost of not having enough money to sustain that growth.

This is why when nV was at 40% margins they weren't able to grow and they changed their view on how to shift the market to higher profitable segments vs the mainstream segment. This pushed their margins up, and look where they are now.
 
Last edited:
Creig said:
WTF? The graph I linked was put together by Market Realist, an investment and analytic website. Not AMD. And it was based on AMD's SEC filings. Not anything said by Lisa Su.

http://marketrealist.com/2016/12/drove-amds-stock-cy-2016/
Click to expand...

Look at AMD and NVidia stock for 1 year. They go hand in hand and they are both very bubbled due to AI. And even worse, its done via QE money. That's how AMD gained ~450-500% and NVidia ~400% in a year.

You may also notice how VPs, including Lisa Su have used the time to dump stock.
 
Last edited:
Someone mentioned Polaris being better than Pascal in performance per watt per unit die area.

RX480 reference is 165w, let's call it 130w GPU.

GTX 1080 FE is 180w let's call it 150w GPUto


Polaris 10 is 230mm², GP104 is 314mm².

GTX 1080 is 60% faster (my estimate )when both reference versions are compared.

That's 100/135/230 vs 160/150/314 so 0.0032 performance per watt per mm² on Polaris vs 0.0034 performance per watt per mm² on Pascal.


Even if you look at pure compute throughput you have 6tflops vs 9 tflops so that's 100:150 and you end up with 0.0032 for both, and this is ignoring the fact that Pascal clocks well above reference {9tflops) within the 180w envelope.

Not that this matters to the consumer, we don't care about die area because that's factored into the price.

You'll notice I gave Polaris a 5W advantage in perf/w/mm², just to be nice, I don't think the difference between gddr5 and gddr5x is even 5w

It's an interesting metric from an engineering perspective though, I would have linked TPU's performance charts but I have no doubt somebody will complain about them using old drivers and biased games.

As for the ongoing debate about the value of stock price of an indication of the health of the company, I have nothing but disdain for the stock market, I struggle to make sense of it. It's a construct built on essentially public perception, which is by nature easy to skew. Seems like a recipe for disaster

Zen is poised to be a success if the performance numbers given by AMD are accurate and not merely best case scenarios.

Then again it feels like I first heard about it a lifetime ago, it's been a long time coming, and it comes at a time that Intel is moving on to skylake-e, don't get me wrong I'm quite excited for Zen, I just recognize soon after it's release, at best, I will be fantasizing about a skylake-e setup.

Vega seems like it's too late and the best it can do is lower prices. It's also massive, but hey, maybe it turns out to clock well. Something tells me amd won't exactly be making phat margins on top of Vega GPUs.

Even if matches Pascal Titan in every single benchmark, it will be doing that a year late, with a bigger die and HBM2. Hello red

Maybe this is what they mean by #BetterRed? :p couldn't resist, sorry.
 
Last edited:
razor1 said:
What numbers, you took an excerpt of their report, which has nothing to do with reality, it was AMD was hoping for, which didn't HAPPEN.

AIB-graph2.PNG


Anyone with any common sense could see AMD was talking out of their ass.

Performance is bigger than mainstream and value COMBINED! (JPR mainstream includes both mainstream and value segments)
Click to expand...

True but rx 480 I really don't consider it mainstream. It tends to give you playable settings at 1440p. To me it is more of a performance card with mainstream price, may be this only may apply to me since I got one for 170 8gb! lol
 
Last edited:
Ugh, by that time I'll have saved enough to get a 1080, I know, I know, first world problems.
 
NKD said:
True but rx 480 I really don't consider it mainstream. It tends to give you playable settings at 1440p. To me it is more of a performance card with mainstream price, this only may apply to me since I got mine for 170! lol
Click to expand...

By the way JPR does their numbers, rx480 or a gtx 1060 are mainstream cards.
 
I don't know what's going on, but now that the Vega and 1080 Ti seem a bit late, wish I had just got a 1080 when they came out.
 
Well now seems like a bad time to buy, but I really want to run a single card as I am playing more competitive multiplayer now (Overwatch). I have two 980 Ti.
 
frunction said:
Well now seems like a bad time to buy, but I really want to run a single card as I am playing more competitive multiplayer now (Overwatch). I have two 980 Ti.
Click to expand...

Then remove one :p I guess it depends on the resolution and settings you run at. If you're running 1440p one should be fine but I doubt it will hold solid 144hz or something .

If you're playing at 1440p with 4x SSAA and want high framerates it will be harder to please you with one card
 
Ieldra said:
Then remove one :p I guess it depends on the resolution and settings you run at. If you're running 1440p one should be fine but I doubt it will hold solid 144hz or something .

If you're playing at 1440p with 4x SSAA and want high framerates it will be harder to please you with one card
Click to expand...

I wouldn't think he would be using any kind of AA in a competitive multiplayer environment. Frames per second is king, and a lot of players turn settings down to get more. Single player is different though.
 
Most of what's coming with Zen and Vega architectures is predictable at this point. Should have a rather significant performance and efficiency boost. A boost that may very well not have shown up in benchmarks without an updated compiler. One of Raja's tweets would seem to indicate people figured it out. What's not known are the specifics: core counts, cache sizes, etc. That said, Vega is far larger than it should be for what's been demoed so far. I wouldn't be surprised if it went well beyond Titan, especially with more futuristic loads using the new shaders and SM6.
 
Anarchist4000 said:
Most of what's coming with Zen and Vega architectures is predictable at this point. Should have a rather significant performance and efficiency boost. A boost that may very well not have shown up in benchmarks without an updated compiler. One of Raja's tweets would seem to indicate people figured it out. What's not known are the specifics: core counts, cache sizes, etc. That said, Vega is far larger than it should be for what's been demoed so far. I wouldn't be surprised if it went well beyond Titan, especially with more futuristic loads using the new shaders and SM6.
Click to expand...

The Titan X (TXP) is almost, and sometimes is, twice as fast as AMD's best GPU (Fury X). That would be a hell of a jump to match the TXP never mind beat it.

Overexpectations only hurt AMD... kinda like I had my expectations too high for the TXP. Even though it's a great card I had something different in my mind and felt let down.
 
Last edited:
Anarchist4000 said:
Most of what's coming with Zen and Vega architectures is predictable at this point. Should have a rather significant performance and efficiency boost. A boost that may very well not have shown up in benchmarks without an updated compiler. One of Raja's tweets would seem to indicate people figured it out. What's not known are the specifics: core counts, cache sizes, etc. That said, Vega is far larger than it should be for what's been demoed so far. I wouldn't be surprised if it went well beyond Titan, especially with more futuristic loads using the new shaders and SM6.
Click to expand...


Core counts are pretty much know now. Also going by Polaris's die size, Vega's die size adds up ;), we even discussed this prior to seeing Vega's die, it was going to be bigger than GP102 is exactly what I stated with 4096 cores..... the GP102 has a die size of 471 mm^2, so Vega coming just north of that, was to be expected with a core count of 4096, which is the die that Raja showed Linus and others.
 
I really don't think amd is demoing a full vega chip. I think they are being sneaky here, everything is too tight lipped.
 
razor1 said:
Core counts are pretty much know now. Also going by Polaris's die size, Vega's die size adds up ;), we even discussed this prior to seeing Vega's die, it was going to be bigger than GP102 is exactly what I stated with 4096 cores..... the GP102 has a die size of 471 mm^2, so Vega coming just north of that, was to be expected with a core count of 4096, which is the die that Raja showed Linus and others.
Click to expand...
True, but we truly dont know what they are demoing is the full chip or not. The one we have seen. Since Vega is suppose to make a new series of cards.

It looks like AMD tricked every one with clock speeds on ryzen. going from 3ghz to 3.2 to 3.4 and now confirmations that they had on chip running at 3.6ghz stock. So it seems like they kept alot of people in dark about what the stock speeds would be. They might be doing the same with Vega, may be not demoing a full chip.
 
Last edited:
NKD said:
I really don't think amd is demoing a full vega chip. I think they are being sneaky here, everything is too tight lipped.
Click to expand...


They are, I have no doubt they are, cause Vega doesn't have a full set of DP units either, so there might be a Vega with all of its DP units, which will account for a chip to go against the GP100, but that is only if it can muster up performance to it.

At this point, AMD not showing its best is just kinda worthless endeavor to fool nV the timing just doesn't make sense. For Vega to come out mid this year and Volta coming out 6 months after, why even bother with trying to fool nV, cause nV's schedule is set, there is nothing nV can do at this point, to speed up their timelines. All AMD can do if they had a bigger chip (more performance) would be showing it and slow down nV's current sales, which would be good for AMD. The blindsiding thing just doesn't work with the window of time available to either of the companies.
 
NKD said:
True, but we truly dont know what they are demoing is the full chip or not. The one we have shown. Since Vega is suppose to make a new series of cards.
Click to expand...

Hmm they stated its their top tier chip (now don't know where its going to fit into that tier or if their top tier is only performance tier, but going by the Doom and Battle Front numbers performance looks likely)
 
razor1 said:
Hmm they stated its their top tier chip (now don't know where its going to fit into that tier or if their top tier is only performance tier, but going by the Doom and Battle Front numbers performance looks likely)
Click to expand...

True, but I think they have always referred to it as vega being high end chip. Now it could be any of the chips based on vega. I usually see them referring to the card as vega in general.
 
Have we ever seen AMD or any company for that matter show something less then what they are capable of? ;)
 
NKD said:
LOL they did kinda with ryzen!
Click to expand...

Not really, they are showing their best they can do in particular tests, clocks could have been lower, but they also down clocked Intel's chip to compensate. When AMD was able to get clocks up, they didn't down clock Intel's chip anymore.

And we still don't know the end results cause those tests could be very specific to the best Ryzen can do in particular applications.
 
You must log in or register to reply here.
Back
Top