The rise and fall of AMD

[Phynaz]

Some detailed data about the continuous fall of Intel: 17-20% per quarter.

Who is dumb enough to make QoQ comparisons in a seasonal business?

Oh!...nevermind.

 

[Phynaz]

So you are comparing AMD 30% drop in revenues against Intel 20% drop in profits? Do you really have a clue of what you are saying here?

Yeah, he does. I'm convinced he is just trolling and laughing at us feeding him. In another thread he says AMD cpu's are better perf/watt than Intel.

Eventually the mods will ban him as they do with the others of his ilk.

 

[Idontcare]

Look up the thread regarding embedded processors. IDC holds the opinion that TI had very high margins on what they were selling to Ford. I stated a different opinion based upon when I used to work in the auto supplier industry. Neither of us posted again. We didn't going scouring the Internet to spam links to prove our point. Do you know why? Because adults don't have to "win" an argument when there is a difference of opinion.

FWIW I may have failed to post again in that thread but you did alter my opinion on the margin situation.

After thinking about it for some time it dawned on me that we (TI) sold off the entire Attleboro auto division (technically sensors and controls) and I remembered during the internal satellite broadcast the reasoning was the gross margins mix and forecast was well below where we thought (at the time) our digital CMOS and high-performance analog portfolios were headed.

I don't recall the thread galego is referring to, but I suspect I bailed on it at some point because I realized feeding trolls is a waste of my time. But my ability to stick with threads pretty much aligns with my interests in the threads, not in my interests to win an argument or to have the last word on a given topic.

I wouldn't be surprised if I just never went back because I really didn't care about the subject matter that was being discussed. It happens, not really going to apologize for that :|

But in the case of the auto stuff you and I were discussing, I think I simply forgot about it after thinking on it for a few days. I'm that way with pm's, hope no one gets offended if/when I fail to reply to a pm, it usually means I meant to think it over but then forgot to get back to it. It happens, and for that I really will apologize because it is rude of me and I wish I could be more courteous in that regard.

 

[galego]

Going only by the snippet you delivered one would probably not realize that this only applies to market share in India. Were you deliberately obscuring this fact, or do you really think India is representative of global market share?

Seeking alpha global market up to first quarter (Q1-2012)

It seems that 20% market Q4-2012 is representative. I believe that Q3-2013 is about 25-30%.

 

[ShintaiDK]

Seeking alpha global market up to first quarter (Q1-2012)

It seems that 20% market Q4-2012 is representative. I believe that Q3-2013 is about 25-30%.

You forgot to post where you got it from. Its actually part of this:
http://seekingalpha.com/article/415801-buy-amd-it-can-thrive-in-the-face-of-intel-s-dominance

Buy AMD, stocks at 7.5$ back then. Great buy! Today its only worth 2.70$.

You also forgot to tell what the chart actually covers. Because its not the marketshare you wish to portrait. Its the marketshare number from passmark users.

Yet again you shoot yourself in the foot with a unseen high caliber. But trolls regenerate, dont they?

 

[Exophase]

Here is the original source for that graph:

http://www.cpubenchmark.net/market_share.html

It's from passmark benchmark results which is probably not the best statistical sampling..

This is what Mercury Research says about AMD's market share for 2012:

http://investorvillage.com/smbd.asp?mb=476&mn=257050&pt=msg&mid=12488526

 

[Sleepingforest]

The graph itself is from PassMark CPU Benchmarks. Here's a more updated version. It shows AMD as down slightly, which makes sense since Intel gained unit share in a shrinking market.

On the other hand, it's tough to trust PassMark 100% because it's not necessarily representative.

 

[galego]

Yes Intel's revenue is down 7% and this is apparently a grave problem, yet AMD's 31% revenue fall is a sign they are on their way back. LOL

The data was merely showing that Intel is falling. Count the number of "downs" in the figure.

Some more data on Intel falling:

Meanwhile, the company has responded to lower demand for PC chips by taking some manufacturing equipment off line and readying a new production process slated for introduction late this year. As a result, Intel said Tuesday that it will reduce its large capital spending bill for 2013—which has alarmed some analysts—by about $1 billion to $12 billion.

http://online.wsj.com/article/SB10001424127887324030704578427062777641762.html

 

[galego]

In another thread he says AMD cpu's are better perf/watt than Intel.

Untrue. Another example you cannot read correctly.

 

[Sleepingforest]

Untrue. Another example you cannot read correctly.

Actually, in response to

Are you telling me AMD beats Intel in Perf or Perf/W ?

You replied:

Yes.

So it seems that you just don't remember the claims you make.

 

[Exophase]

I thought that too but I'm guessing he's saying that he just meant that one of those were true.. meaning he believes AMD beats Intel in perf. Which we all know is usually not the case but galego is probably sticking to these mysterious cases where FX-8350 allegedly beats i7-3770K by 70%...

 

[Charles Kozierok]

This entire thread is an embarrassment.

 

[Sleepingforest]

I thought that too but I'm guessing he's saying that he just meant that one of those were true.. meaning he believes AMD beats Intel in perf. Which we all know is usually not the case but galego is probably sticking to these mysterious cases where FX-8350 allegedly beats i7-3770K by 70%...

The 8350 actually does win in Linux by a bit. But that's a use-case for about 1% of the world at best.

 

[galego]

You forgot to post where you got it from. Its actually part of this:
http://seekingalpha.com/article/415801-buy-amd-it-can-thrive-in-the-face-of-intel-s-dominance

Buy AMD, stocks at 7.5$ back then. Great buy! Today its only worth 2.70$.

You also forgot to tell what the chart actually covers. Because its not the marketshare you wish to portrait. Its the marketshare number from passmark users.

Yet again you shoot yourself in the foot with a unseen high caliber. But trolls regenerate, dont they?

If you have access to the second part of the article, why do not you read what says about the chart and post here. Just the part where uses the word "reliable".

 

[Centauri]

Why do the moderators allow every AMD thread to turn into a slugfest? This doesn't happen on other boards of the same caliber.

 

[galego]

This is what Mercury Research says about AMD's market share for 2012:

http://investorvillage.com/smbd.asp?mb=476&mn=257050&pt=msg&mid=12488526

It looks more like a poster in a forum saying it.

 

[galego]

I thought that too but I'm guessing he's saying that he just meant that one of those were true.. meaning he believes AMD beats Intel in perf.

Right! Dasrtral wrote "OR" And my "yes" was to the first part of his question.

 

[galego]

The 8350 actually does win in Linux by a bit. But that's a use-case for about 1% of the world at best.

Not only linux. AMD also win some benchmarks on windows. E.g. FX-8350 wining an i7-3770 on crysis 3.

Recall that dastral tried to convince me that Intel always wins, which is not true.

 

[ShintaiDK]

If you have access to the second part of the article, why do not you read what says about the chart and post here. Just the part where uses the word "reliable".

You have no intention in it anyway. The chart is based on 1000s of passmark users. Its about as useless as it gets. But its the only chart that somewhat remotely can be used iny your trolling to avoid the painful fact for you, that Intel now sits on 85% and AMD is another VIA and irrelevant.

 

[Ajay]

Why do the moderators allow every AMD thread to turn into a slugfest? This doesn't happen on other boards of the same caliber.

What do the mod do on other boards to prevent this? Which boards? I'm just curious.

 

[Exophase]

The 8350 actually does win in Linux by a bit. But that's a use-case for about 1% of the world at best.

You're referring to Phoronix right? http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=1

This is a summary of every result:

On average, FX-8350 is over 10% behind. There are no outliers where FX-8350 wins by 70% - the highest is C-Ray at 41.6% and this is way off from the next highest win of 18.9%. Its biggest loss is by 58.6%.

Overall it loses 15 benchmarks and wins 7. And of course you can't claim that there's an artificial disadvantage due to compiler games.

It looks more like a poster in a forum saying it.

You think a post on an investment forum is lying about getting it from Mercury Research why?

Right! Dasrtral wrote "OR" And my "yes" was to the first part of his question.

But you really should have been clearer. I think you're probably doing this on purpose.

 

[jhu]

You're referring to Phoronix right? http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=1

This is a summary of every result:

On average, FX-8350 is over 10% behind. There are no outliers where FX-8350 wins by 70% - the highest is C-Ray at 41.6% and this is way off from the next highest win of 18.9%. Its biggest loss is by 58.6%.

Overall it loses 15 benchmarks and wins 7. And of course you can't claim that there's an artificial disadvantage due to compiler games.

One thing I wish Phoronix would do with their tests is to actually test which compiler switches produce better faster code. In that link they should have used "-march=bdver1" for the FX8150 and "-march=core-avx-i" for the Intel processors. As is both processors are underperforming from their potential probably.

 

[Phynaz]

Why do the moderators allow every AMD thread to turn into a slugfest? This doesn't happen on other boards of the same caliber.

It happened due to one poster. The rest of us were actually have a decent conversation about the article and the insights it provided.

We can try to get back on topic, what did you think of the two articles?

 

[Exophase]

One thing I wish Phoronix would do with their tests is to actually test which compiler switches produce better faster code. In that link they should have used "-march=bdver1" for the FX8150 and "-march=core-avx-i" for the Intel processors. As is both processors are underperforming from their potential probably.

They do some of this in other articles:

http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=3

Not really a lot to gain from bdver1 or bdver2, I don't think.. at least not from those benches.

Here's core-avx-i:

http://www.phoronix.com/scan.php?page=news_item&px=MTA3NjE

Some arguably more appreciable benefits.

 

[Sleepingforest]

But you really should have been clearer. I think you're probably doing this on purpose.

Eh, maybe he's a programmer and he thought he was just supposed to return a Boolean. :awe: