[WSJ] Intel To Combine Mobile, PC Chip Groups

[NTMBK]

This comment from AnandTech shows that it isn't just to blur the division's losses:

We'll see when they actually release the next quarter's results. "Haven't decided yet" can easily mean "we've decided, but don't want to tell you yet".

 

[Abwx]

Intel 2014 tablet goal: 40M

Q1: 5M
Q2: 10M
Q3: 15M
Q4: 20M (projected)
-----------------------
'14: 50M

That is a 5X tablet unit growth in 1 year and isn't even counting the non-tablet Bay Trail market.

I bet they will double again in 2015.

To double again in 2015 they ll have to cram 3 other bn in subsides to get their losses up to 7bn, once the subsides end the amounts shipped will collapse in a matter of weeks...

 

[witeken]

To double again in 2015 they ll have to cram 3 other bn in subsides to get their losses up to 7bn, once the subsides end the amounts shipped will collapse in a matter of weeks...

If I can show you the insignificance of Intel's contra-revenue in yet another way (the 4th), consider this:

In Q3, Intel's tablet volume increased by 50%, yet their mobile division's loss decreased. In fact, the Atom volume tripled compared to Q1, yet they only lost an additional $0.1B.

The contra-revenue will cease in 2015 anyway.

 

[Nothingness]

If I can show you the insignificance of Intel's contra-revenue in yet another way (the 4th), consider this:

In Q3, Intel's tablet volume increased by 50%, yet their mobile division's loss decreased. In fact, the Atom volume tripled compared to Q1, yet they only lost an additional $0.1B.

You can't make any such inference since the MCG is delivering many more things that are not subject to contra-revenue.

Mobile and Communications Group: Delivering platforms designed for the tablet and smartphone market segments; and mobile communications components such as baseband processors, radio frequency transceivers, Wi-Fi, Bluetooth*, global navigation satellite systems, and power management chips.

 

[tential]

If I can show you the insignificance of Intel's contra-revenue in yet another way (the 4th), consider this:

In Q3, Intel's tablet volume increased by 50%, yet their mobile division's loss decreased. In fact, the Atom volume tripled compared to Q1, yet they only lost an additional $0.1B.

The contra-revenue will cease in 2015 anyway.

This forum had people crying about Contra-Revenue for god knows how long now. "One Contra-Revenue stops Intel will fail!!!" or "Intel's Contra-Revneue scheme will be the hdeath of them", or any number of ridiculous comments.

You can argue til your blue in the face it won't change their opinion. The only thing that will stop them from whining about Contra-Revenue will be the numbers once intel actually stops their Contra-Revenue scheme.

 

[Abwx]

You can't make any such inference since the MCG is delivering many more things that are not subject to contra-revenue.

Seems that these are obviously not accounted in this dpt, didnt you notice this ?..:

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.

1 million revenues....

You mean that all thoses other chips are also given for free.??.

 

[mrmt]

You can argue til your blue in the face it won't change their opinion. The only thing that will stop them from whining about Contra-Revenue will be the numbers once intel actually stops their Contra-Revenue scheme.

I doubt it. The contra-revenue numbers are clear for anyone that read accounting 101, the relevant market agents aren't complainign about the accounting on the program and the audit company doesn't seem to be saying anything about it. But yet we hear fairy tales about a stealth contra-revenue program spilling to the PC Client Group, and how Intel is spending 1 billion per quarter in contra-revenue, all that because people don't want to believe that AMD chips are bad products and refuse to believe that OEMs don't want to buy them.

 

[Abwx]

all that because people don't want to believe that AMD chips are bad products and refuse to believe that OEMs don't want to buy them.

That s just FUD at its worst, Anandtech did a review wich contradicts completely your sayings.

http://www.anandtech.com/show/7974/...tecture-a10-micro-6700t-performance-preview/3

http://techreport.com/review/26377/a-first-look-at-amd-mullins-mobile-apu

You keep ignoring the facts.

So wich is the chip that needed contra revenues when looking at thoses reviews.??.

 

[Sweepr]

That s just FUD at its worst, Anandtech did a review wich contradicts completely your sayings.

http://www.anandtech.com/show/7974/...tecture-a10-micro-6700t-performance-preview/3

http://techreport.com/review/26377/a-first-look-at-amd-mullins-mobile-apu

You keep ignoring the facts.

So wich is the chip that needed contra revenues when looking at thoses reviews.??.

Funny, the same guy that bashes Intel for demoing Core M under a good light (metal chassis) keeps posting old AMD reference platform Mullins tests ad nauseam. Quick reality check, if Mullins was really a BT/ARM perf/watt killer then AMD wouldn't have to hide power consumption numbers in their shinny preview and their reference platform would like a proper BT/ARM competitor (mainly cheap thin/light 7-9'' tablets) intead of bulky >10'' tablets - their 11.6'' Discovery Platform looks a lot more like a Core M competitor than anything you'd find BT inside.

 

[Abwx]

Funny, the same guy that bashes Intel for demoing Core M under a good light (metal chassis) keeps posting old AMD reference platform Mullins tests ad nauseam. Quick reality check, if Mullins was really a BT/ARM killer then AMD wouldn't have to hide power consumption numbers in their shinny preview and their reference platform would like a proper BT/ARM competitor (mainly cheap thin/light 7-9'' tablets) intead of bulky >10'' tablets - their 11.6'' Discovery Platform looks a lot more like a Core M competitor than anything you'd find BT inside.

Power comsumption can be estimated with the numbers available from other such chips, the Discovery plateform has been more or less used as basis for a design and its numbers cope well with what is computable, they state 6-8h autonomy with a 32Wh battery, this is adequate with the Stream 14 that do 5.28h with the same battery but a 14" screen.

It is indeed a Core M competitor but only because of the latter s unability to get a better perf/watt ratio as well as Intel insistence to use it in low power devices, otherwise they are not in the same segment at all, Mullins is essentialy a chip whose price is in the 20-30$ bracket, BT was intended to be sold about 25$ before the lack of interest from OEMs and AMD s soon to come chip did lead Intel to push the panic button and start their contra revenues, seriously, how much BTs would have been sold if there was no subsides..?.

 

[Ajay]

Intel 2014 tablet goal: 40M

Q1: 5M
Q2: 10M
Q3: 15M
Q4: 20M (projected)
-----------------------
'14: 50M

I bet they will double again in 2015.

It's a failure because Intel is using BT as a loss leader, but they don't have a more profitable product to up-sell into. This just won't work. It's looking like the high-end tablet market is going to stagnate. If that is true, Intel needs to sell to Apple. IIRC, Apple takes home the vast majority of profits from the entire tablet market - therefor, without Apple, there are only scraps left. Intel doesn't have a profitable market to sell into. That's Intel's big problem.

 

[StrangerGuy]

It's a failure because Intel is using BT as a loss leader, but they don't have a more profitable product to up-sell into. This just won't work. It's looking like the high-end tablet market is going to stagnate. If that is true, Intel needs to sell to Apple. IIRC, Apple takes home the vast majority of profits from the entire tablet market - therefor, without Apple, there are only scraps left. Intel doesn't have a profitable market to sell into. That's Intel's big problem.

They spent billions chasing a ARM speedboat that has long sailed into the good enough territory with a canoe. Already past the time to abandon it and lick the wounds.

Oh please don't tell me about that Intel performance, process advantage blah blah blah because it flies off in the face of economics and market conditions of reality.

 

[III-V]

The market conditions of reality actually put Intel in a fairly respectable place in terms of market share. The profits will come later, after Intel gets their costs down.

 

[Ajay]

The market conditions of reality actually put Intel in a fairly respectable place in terms of market share. The profits will come later, after Intel gets their costs down.

Without winning Apple, or at least Samsung, Intel will have to get it's cost very low to compete in the mass market space for tablets or phones. That why I suggested it may not happen till Intel gets EUV + 450mm wafers + SoC advances. I don't expect Intel to bow out of such a huge market, but though they will be ratcheting down their sales & marketing expenses, they may be eating their development costs for a while.

So, rolling mobile into PCCG makes sense from a technology standpoint and from a leadership/management standpoint - there should be some cost synergies as well (we'll have to wait and see). This way CCG becomes an incubator for Intel mobile until they become profitable.

 

[Enigmoid]

Even without the desired margins intel may continue to sell chips at low prices precisely (not give them away but abandon their high margins) simply for marketshare and to prevent competitors from grabbing profits and eventually scaling up to challenge high end servers/desktop/etc.

The huge number of BT android tablets intel has pumped out has surely taken profits away from mediatek and the like. I except them to continue selling at lower margins (though without counter revenue) in the future.

I don't like the sneaky underhanded way this is going but this is business.

 

[Shivansps]

The thing is Intel has no cheap SOC, its worse BT, the Z3735G, its still a LOT better of what Mediatek and AllWinner can offer with those quad A7s, no to mention duals. They need to put up a cheaper SOC to win more market share among chinese tablets.

My wild guess is that they are specting to archive that with Rockchip. The thing is, that has yielded no results so far.

 

[Nothingness]

The thing is Intel has no cheap SOC, its worse BT, the Z3735G, its still a LOT better of what Mediatek and AllWinner can offer with those quad A7s, no to mention duals. They need to put up a cheaper SOC to win more market share among chinese tablets.

MediaTek MT6595 with its 4 Cortex-A17 is certainly better than BT.

 

[StrangerGuy]

The thing is Intel has no cheap SOC, its worse BT, the Z3735G, its still a LOT better of what Mediatek and AllWinner can offer with those quad A7s, no to mention duals. They need to put up a cheaper SOC to win more market share among chinese tablets.

My wild guess is that they are specting to archive that with Rockchip. The thing is, that has yielded no results so far.

Except Mediatek is not just a mere SoC designer, it's killer advantage is to sell complete mobile solutions where phone makers can just slap in screens, batteries, casing etc to put finished products onto the market in the least amount of effort and cost possible.

Winning in performance alone is everything, riiiiiiight.

 

[III-V]

MediaTek MT6595 with its 4 Cortex-A17 is certainly better than BT.

Are you sure about that? I've not seen benchmarks for it.

Winning in performance alone is everything, riiiiiiight.

Because that's totally what he was saying.

If you're just here to rain on everyone's parade, and refuse to listen to reason, then bug off.

 

[Nothingness]

Are you sure about that? I've not seen benchmarks for it.

I'm afraid I'll have to link Geekbench which many don't like: Dell Venue 10 7040 vs Meizu MX4. Basically look for Meizu MX4 reviews on the Web as it seems to be the only widespread device using MT6595.

On the GPU front, Moorefield is ahead, and about the same as Baytrail. The GPU in MT6595 is indeed a disappointment when you see what the CPU is able to achieve...

 

[III-V]

I'm afraid I'll have to link Geekbench which many don't like: Dell Venue 10 7040 vs Meizu MX4. Basically look for Meizu MX4 reviews on the Web as it seems to be the only widespread device using MT6595.

On the GPU front, Moorefield is ahead, and about the same as Baytrail. The GPU in MT6595 is indeed a disappointment when you see what the CPU is able to achieve...

I still have doubts regarding Geekbench's ability to accurately compare x86 to ARM, despite the author's commentary here. But thanks for that.

I'm curious, though. Even Intel's Saltwell was fairly competitive with A15, and the initial claims for A12 seemed to indicate it was slightly slower than the A15. Of course, it seems A12 exceeded expectations, and is now appropriately labeled as an A17 part. I guess if anything, it shows that I was right about A15 being a piece of garbage.

Regardless of my reluctance with Geekbench, certainly Silvermont is getting a bit long in the tooth. We really could use Cherry Trail right about now.

 

[Nothingness]

I'm curious, though. Even Intel's Saltwell was fairly competitive with A15, and the initial claims for A12 seemed to indicate it was slightly slower than the A15. Of course, it seems A12 exceeded expectations, and is now appropriately labeled as an A17 part. I guess if anything, it shows that I was right about A15 being a piece of garbage.

My take is that A12/A17 is a more balanced/tuned design. Think about P4 vs P-M to some extent

Regardless of my reluctance with Geekbench, certainly Silvermont is getting a bit long in the tooth. We really could use Cherry Trail right about now.

Cherry Trail micro arch is supposed to be basically the same as Silvermont with only some increase in frequency (2.7 vs 2.4 GHz).