Does AMD have any hope against Intel?

[witeken]

You should check the numbers, it s 1bn $/quarter, there s no doubt that most of this money is given to OEMs.

It's really quite insane if you think even a low-end chip like Bay Trail costs $100.

40M*$15 = 600M USD. If you do the math for the gross margin impact (Intel predicted 1%), the number is about the same. Thirdly, if you look at Intel's losses in 2013 and compare it, it has risen by about 250M per quarter.

 

[Omar F1]

I hope AMD do come back to compete with Intel core vs core, clock vs clock, but they aren't going to do it by piling on more cores, then sitting there 'demanding' games devs write consistently theoretically perfectly threaded code as a magic alternative to AMD not improving efficiency (IPC) of the CPU itself anywhere near the level of competitors. Some code (video encoding) is inherently more naturally threadable and easier to "parallel" across cores than other code (gaming). And even if gaming code were perfectly threadable, you only have to look at the half-broken state many modern games get released in to see optimization is way down on their list of priorities (if they even budgeted it in at all that is...)

Some people just can't seem to handle "MOAR CORES" is not some magic performance enhancing panacea even for supposed 2014 "next-gen" multi-threaded games:-
http://gamegpu.ru/images/remote/htt...Action-Assassins_Creed_Unity-test-ac_proz.jpg
http://gamegpu.ru/images/remote/htt...lefield_4_Dragons_Teeth-test-bf4_proz_amd.jpg
http://gamegpu.ru/images/remote/htt...ion-The_Evil_Within_-test-evilwithin_proz.jpg
http://gamegpu.ru/images/remote/htt...hadow_of_Mordor-test-ShadowOfMordor_proze.jpg
http://gamegpu.ru/images/remote/htt...f_Duty_Advanced_Warfare-test-cod_proz_amd.jpg

AAA games devs not spending an extra $2-3m delaying their game another 3 months so they can hand tweak every single line of code for the 2.0% of gamers with 6-8 CPU cores might not be how some want reality to be - but it simply is how it is. 3.19% of Steam users have 1 core, 48.3% have 2 cores, 2.7% have 3 cores and 43.6% have 4 cores. That's a combined 97.8% of the market with 1-4 cores. And being a Steam survey, this is if anything biased towards gamers with better hardware and excludes the plethora of cheap sub-2.5GHz dual-core non-gaming office boxes, netboxes, laptops, netbooks, etc, with no Steam client or games beyond Freecell, Solitare & Minesweeper installed at all.

Well, we'll are waiting a good return for AMD, I hope they're already pushing a good plan to do so.

 

[Abwx]

It's really quite insane if you think even a low-end chip like Bay Trail costs $100.



40M*$15 = 600M USD. If you do the math for the gross margin impact (Intel predicted 1%), the number is about the same. Thirdly, if you look at Intel's losses in 2013 and compare it, it has risen by about 250M per quarter.

Numbers were published by Intel, do your homework, that s 1bn/quarter losses since 2013, this year it will be 4bn for 40m chips shipped, do the maths, 100$ is what it cost to Intel to give one chip, so there s much much more than chip manufacturing costs and RD, actualy they are throwing at the OEMs the equivalent of the whole AMD s CPUS/APUs revenues.

The Mobile and Communications Group continued its slide for Q3...

Revenues were down around 100% year-over-year, at a meager $1 million. This division is also responsible for the majority of Intel’s loses, with a Q3 operating loss of $1.043 billion.

For the nine months ended September 27th, this unit has lost $3.096 billion.

http://www.anandtech.com/show/8615/intel-q3-fy-2014-quarterly-earnings-analysis

 

[Enigmoid]

Numbers were published by Intel, do your homework, that s 1bn/quarter losses since 2013, this year it will be 4bn for 40m chips shipped, do the maths, 100$ is what it cost to Intel to give one chip, so there s much much more than chip manufacturing costs and RD, actualy they are throwing at the OEMs the equivalent of the whole AMD s CPUS/APUs revenues.
http://www.anandtech.com/show/8615/intel-q3-fy-2014-quarterly-earnings-analysis

I already posted about this and I'm not posting again. Mobile and communications has always lost money. Compare before contra revenue and after to get proper numbers.

 

[Abwx]

I already posted about this and I'm not posting again. Mobile and communications has always lost money. Compare before contra revenue and after to get proper numbers.

But then provide us the numbers, or would you be left facing what you re asking me, that is proving a negative, since they didnt publish numbers for their mobile division before their started their contra revenues...

Anyway the numbers above cant be negated and only spiners will pretend that it cost 3.5bn/year to develop a chip like BT and a few generic mobile devices, the RD argument is just plain fairy tale unless, as already said, sweepers are paid 20k$/month at Intel mobile dpt..

 

[witeken]

Numbers were published by Intel, do your homework, that s 1bn/quarter losses since 2013, this year it will be 4bn for 40m chips shipped, do the maths, 100$ is what it cost to Intel to give one chip, so there s much much more than chip manufacturing costs and RD, actualy they are throwing at the OEMs the equivalent of the whole AMD s CPUS/APUs revenues.
http://www.anandtech.com/show/8615/intel-q3-fy-2014-quarterly-earnings-analysis

http://seekingalpha.com/article/2651585-intel-when-losses-go-away-they-leave-profits-behind

Scroll a bit down and you'll see that Intel has been losing money far longer than their contra-revenue in 2013.

Intel should have been quiet about their contra-revenue program, because the amount of FUD and lies that have been spread under that moniker is enormous.

 

[mrmt]

I already posted about this and I'm not posting again. Mobile and communications has always lost money. Compare before contra revenue and after to get proper numbers.

Contra-revenue is 1 billion per year, anyone advocating something far above this number doesn't have a clue on how accounting works.

 

[Arachnotronic]

Anyway the numbers above cant be negated and only spiners will pretend that it cost 3.5bn/year to develop a chip like BT and a few generic mobile devices, the RD argument is just plain fairy tale unless, as already said, sweepers are paid 20k$/month at Intel mobile dpt..

Intel's MCG develops the following:

- cellular baseband + RF transceiver solutions
- carrier certification across many carriers for the above
- low power connectivity (Wi-Fi, Bluetooth, GPS, etc.)
- a pipeline of low power SoCs
- all of the IP that is not shared with PC/server chips (image signal processor, low power variants of Gen X GPU, video processors, sensor hub, and SoC fabric)
- drivers/software
- hardware reference designs

All of this combined is expensive . Broadcom alone spends $1 billion on R&D for just its connectivity business. Qualcomm spends nearly $6 billion per year on R&D, again most of that is likely for mobile chips.

Don't underestimate how much it costs to develop these chips and the attendant platforms.

 

[AtenRa]

I already posted about this and I'm not posting again. Mobile and communications has always lost money. Compare before contra revenue and after to get proper numbers.

The big difference is that before contra revenue they had 300-400M of Revenue in Mobile and Communication sector per quarter, in Q3 2014 Revenue was 1M. :whiste:

Also, the above picture of BOM reduction cost is about the Board and peripherals, not the APU.

Edit: And yes Contra-Revenue is 1 Billion, it is why revenue went to almost ZERO in Q3 2014 in the M&C Group. The Operating income loss of the MCG is another thing and it should be added to the overall losses meaning adding the Contra Revenue 1B.

 

[MiddleOfTheRoad]

Astute observation. A lot of people simply write off others' opinions as "trolling," when in reality, people very rarely post things just to get a rise out of others. Unfortunately, the moderation team here doesn't seem to understand that at times.

Yeah, because using Nazi overtones / racist remarks is always used for fact-finding on hardware comparisons..... Obviously, not the work of a troll. This guy simply joined the forums to bait and troll. Enough said.

 

[Arachnotronic]

But then provide us the numbers, or would you be left facing what you re asking me, that is proving a negative, since they didnt publish numbers for their mobile division before their started their contra revenues...

Anyway the numbers above cant be negated and only spiners will pretend that it cost 3.5bn/year to develop a chip like BT and a few generic mobile devices, the RD argument is just plain fairy tale unless, as already said, sweepers are paid 20k$/month at Intel mobile dpt..

The big difference is that before contra revenue they had 300-400M of Revenue in Mobile and Communication sector per quarter, in Q3 2014 Revenue was 1M. :whiste:

Also, the above picture of BOM reduction is about the Board and peripherals, not the APU.

They also had a very sharp fall-off of their 3G modem business that made up the majority of the revenue for that division.

 

[AtenRa]

They also had a very sharp fall-off of their 3G modem business that made up the majority of the revenue for that division.

You mean they were selling more modems than APUs and at a higher ASP ??? dont thing so.

 

[Arachnotronic]

You mean they were selling more modems than APUs and at a higher ASP ??? dont thing so.

Did I ever make such a claim?

I merely stated that part of the y/y decline in Intel's MCG revenue is due to, in addition to the contra-revenue, a substantial decline in the company's 3G business.

 

[AtenRa]

Did I ever make such a claim?

I merely stated that part of the y/y decline in Intel's MCG revenue is due to, in addition to the contra-revenue, a substantial decline in the company's 3G business.

You said the majority of the MCG were due to 3G Modems. In order for the 3G modems to have the majority of the revenue is only if you sell more modems than APUs and at higher ASP.

 

[MiddleOfTheRoad]

I doubt it. He's a new member trying to learn more about hardware. Just look at his posts, it's clear he didn't do his due diligence before but now he is learning to and wants to learn what the current state of the market is.

Someone who joins a forum to immediately start an inflammatory thread (we all know the AMD vs Intel debates always go that way) is the definition of a troll. To follow up with "Hail Hitler" and "Die in Hellfire" remarks. Seriously, the mods should boot the guy. His only purpose is rabble rousing.

 

[Arachnotronic]

You said the majority of the MCG were due to 3G Modems. In order for the 3G modems to have the majority of the revenue is only if you sell more modems than APUs and at higher ASP.

I said that the majority of the revenue in that business comes from the sale of those 3G modems. I did not say that the decline of the 3G business is responsible for the majority of the loss.

 

[AtenRa]

I said that the majority of the revenue in that business comes from the sale of those 3G modems. I did not say that the decline of the 3G business is responsible for the majority of the loss.

I really dont have the data so i will stop here

 

[Abwx]

Intel's MCG develops the following:

- cellular baseband + RF transceiver solutions
- carrier certification across many carriers for the above
- low power connectivity (Wi-Fi, Bluetooth, GPS, etc.)
- a pipeline of low power SoCs
- all of the IP that is not shared with PC/server chips (image signal processor, low power variants of Gen X GPU, video processors, sensor hub, and SoC fabric)
- drivers/software
- hardware reference designs

All of this combined is expensive . Broadcom alone spends $1 billion on R&D for just its connectivity business. Qualcomm spends nearly $6 billion per year on R&D, again most of that is likely for mobile chips.

Don't underestimate how much it costs to develop these chips and the attendant platforms.

Broadcom spend 1bn out of 8bn revenues, Qualcomm, a quasi patents troll, is spending 6bn out of 24bn revenues, and for only 26 000 employees, that s absolutely exceptional, they must count the bought societies as RD...

Actualy most of their cash must come from the royalties they get from enrolling the gullible com authorities in using their patents for the UMTS system, other than that you did name a relevant exmple since they are also quite good at practicing anti competitive methods.

Anyway this doesnt change the fact that Intel has a cost of 100$/chip, it used to be gloom and doom in their mobile dpt but so far there was never talks of such amazing losses.

Edit : Their whole RD is 10-11bn , no way the mobile dpt use 3bn in RD.

 

[Arachnotronic]

Anyway this doesnt change the fact that Intel has a cost of 100$/chip, it used to be gloom and doom in their mobile dpt but so far there was never talks of such amazing losses.

The business is not at scale right now, and the only way to bring that "effective" cost down is to get to scale.

It's going to be a very long, hard road for Intel's mobile group. The incumbents are powerful and focused, and Intel is still in the process of fundamentally reinventing itself in order to be able to quickly respond to market demands and requirements.

 

[Abwx]

The business is not at scale right now, and the only way to bring that "effective" cost down is to get to scale.

It's going to be a very long, hard road for Intel's mobile group. The incumbents are powerful and focused, and Intel is still in the process of fundamentally reinventing itself in order to be able to quickly respond to market demands and requirements.

That has nothing to do with business not being at scale, you re using hollow formulaes to not name things by their actual name, said simply they are producing a chip that even if sold at 0% gross margin, that is at a 10-15$ price, wouldnt be accepted by the market it target, early infos was that OEMs were not willing to pay as much as 25-30$ as projected by Intel.

On the W8.1 front the chip had the advantage of a better perf/watt CPU, but not GPU, against AMD s Temash but on the Android side there was chips that were good enough CPU wise and that have not worse GPUs, the outcome was clear, neither the AMD nor the Intel offering could provide W8.1 with a competitive offering when price, perfs and perfs/watt were all three taken into account, the ARM thing still managed to get the optimum solution, it s only when AMD announced that they ll get the good perf/watt, perfs ratios that Intel reacted as they were to be outmatched in both Android and W8.1 fronts, if they could stand being competed by ARMs using Android they saw the irruption of AMD as a much bigger threat in the long term, the cause of thoses contra revenues is nowhere else to be find, getting in the financial numbers will only reinforce this logic.

 

[ultimatebob]

Intel can squash AMD anytime they want. Unfortunately, for all consumers Intel don't even consider them a big enough competition to bother taking them down.

I really don't like AMD to be in such a big mess on the processor front. The only thing that seems to work for them is the APU. But, there too they are too conservative. They haven't have really shown the potential that they have.

AMD needs to really shift to a smaller node process to be relevant in the game. Problem is, the more time they take to actually reach there. The more hard it would become for them to come out of the quick sand.

I don't think that they could press AMD any harder than they are now without worrying about Antitrust legislation being forced on them. They need AMD to prove that they don't have a virtual monopoly in Desktop, Laptop, and Server processors.

Saying that Intel has competition in the Mobile space that may encroach into their other product lines isn't enough for the lawyers looking for a big pay day or Attorney Generals looking to make a name for themselves.

 

[Phynaz]

You should check the numbers, it s 1bn $/quarter, there s no doubt that most of this money is given to OEMs.

There's plenty of doubt, do you have any evidence besides your made up numbers to backup your claim?

 

[Lepton87]

Does anyone really think that comparing an old crippled Intel Product on a bigger node comapared with a newer fully-enabled AMD CPU that was designed from scratch puts AMD in a good light? When it even can't beat it when both are overclocked.
i750 is clocked very conseratively and it has about two times as much OC headroom as BD, not to mention it is feature disabled because it lack HT which would help it a lot in MT benchmarks, exactly where it loses. Compare fully enabled Lynnfield with similar OC headroom, otherwise it's not fair.

http://anandtech.com/bench/product/192?vs=434

Lynnfield wins more benchmarks and when it wins it usually does so with a higher margin. If someone wants to bother, a weighted arithmetic average would be nice. And this CPU wasn't even a competitor, its sucessor was.