Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Microsoft is doing the same what they did with the first Netbooks. Those were also using Linux for cost reasons. Microsoft fought back with cheap Windows XP licenses and they were successful.
Things are a bit different this time, though. There is already a cheap successful Linux based OS available this time - Android. In addition, with Google, Intel isn't fighting against me too Linux distributions but with a company that has a lot of support by hardware producers, which itself helps driver and software integration wise. In addition, Google is about to integrate Android Apps into Chrome OS, which is a very nice add-on for a PC. Once they start to support fully locally installed applications, I think Chrome OS will be much more successful than what Linux was on Netbooks years ago. Especially since Windows 8 is a weak competition compared to Windows XP.
Nevertheless, it all depends on how Google executes. They are doing nice stuff in software, still they are way behind what Apple has achieved. For Microsoft it should be sufficient though ...
Intel fabs for Panasonic.
TSMC better watch out! Hard times coming.
Is Intel finally starting to execute in mobile?
Cherry Trail atom 14nm coming late 2014
Since it is popular on this board to post ones personal PC experience:
I own a Samsung All in One PC. It runs on a Sandy Bridge 2.8 GHz Core I3 with Nvidia graphics and got Windows 8 preinstalled. I hated Windows 8 from day one since things just got worse from my point of view. Installed Win 8.1 and found it even worse. I kind of lived with it but never liked to use that machine, though hardware-wise, it's perfect. Decent performance, compact and silent, nice screen of right size (21"). It is placed in the living room, next to my TV and I also use it for Multimedia stuff via HDMI out. Therefore, a desktop is a nogo (ugly) and I need an extry screen in addition to the TV for working while others watch TV, so none of these Mini-PCs for me. The price of the AIO was just right - cost me 700$ last year, almost half of that of an iMac.
Anyway, after suffering a long time from Windows 8, I finally dumped it and replaced it with a new Windows 7 install. Took the chance and bought a Crucial/Micron MX100 SSD with 256 GB. It uses 16nm (!) MLC NAND and can be had for a little more than 100$ - I call that a nice price. In addition, I bought a Blu-Ray writer so I can watch (3D) Blu-Rays. That's just 60 bucks more.
Anyway, here are some reasons why I hated Windows 8 so much:
- DVD playback costs 10 bucks extra (is M$ nuts?)
- Upgrade to home premium with Media Center (liked that on Windows 7) costs almost 200$ (hello, still nuts?)
- They dropped the nice Windows 7 backup tool and the Windows 8 one can't backup on DVDs and when using an external HDD, it wants to format it instead of just adding the backup files to the other data (hello, nuts?)
- A fresh installation of Windows 8 is impossible since they don't provide a DVD - you can't even switch an HDD and reading out the license key is illegal (money makes the world go round - and lose customers)
- Aero UI is not available for Windows 8 (WTF - back to middle ages?)
- Installing Ubuntu next to Win 8 makes Win 8 not boot anymore due to EFI BIOS protection (what a great evolution ...)
- When starting a metro App by accident, Alt-Tab brings you to Metro menu instead of desktop (arrrggghhhh)
- In case you want to reboot your PC so you can enter the BIOS, you have to press Ctrl while left clicking the shutdown button in the Charms menu (yeah, keep it simple ...)
- Worst of it all - it refused to enter standby automatically after some minutes of idle, but powers on itself randomly for updates or whatever (had that issue also on Windows 7 though)
Ok, as you might guess, I could go on for ages like that. I really hate Microsoft and the stuff they do. The biggest problem Intel has these days is this so-called software company that makes big money with very bad products.
To install Windows 7, on the other hand, I really had to go through quite a hassle:
- Disable the EFI mode in BIOS and deactivate the fast boot option
- Windows 7 had no driver for my Wifi and ethernet and Samsung doesn't provide the drivers as downloads, just an update tool which needs the old drivers preinstalled and a network connection ...
- So, I had to go to the device manager, look for the PID and Vendor of the hardware, do a search on my netbook, transfer the file via SD card and install the network driver together with all the other stuff
- Then I could use the update tool and install the latest drivers from Samsung
- Still some issues with the graphics, so installed the latest Intel GMA drivers and the Nvidia drivers from their pages
- To playback (3D) Blu-Rays in Media Center, one needs PowerDVD Ultra - costs only 90 bucks, 40 more than the Blu-Ray writer and about twice as much as a standalone player (I am not going to tell you the russian site where I took that from ...)
Windows 7 is much better than Windows 8, which is actually not difficult, but I am still not loving that either, because:
- Usability of StartMenu and Explorer got worse (more clicks, missing menus etc.)
- It loads tons of updates, that's really a pain and takes days to complete
- When shutting down and applications refuse to close you have to interrupt the intermediate shutdown, close the app manually, and click on shutdown again - WTF?
- Media library can't exclude folders (don't want to listen to my son's tales)
These are minor issue compared to Windows 8, but it is far from great, especially when compared to my wife's MacBook - surfing is a charm on that thing and it just works.
So, why am I telling all this? Because I really understand people not wanting to spend much money for a new PC anymore. It just sucks, it is no fun, it's work and anger. It's all about the software, the hardware available today is great - ultra fast and silent. But Windows, also Windows 7, is no fun to use. That's the reason why I'm counting on Chrome OS - since it eases the usage of the PC, makes it less to maintain and more like "just works". Microsoft will never get it I suspect.
By the way, tried Ubuntu on that machine - it is slow and lags while typing - so, looks like I have to stay with Windows 7 - this time ...
Poor little Bernstein. The market didn't care about his thesis
Looks like his only follower has been Fastpathguru
Especially worth to note the catalysts he mentioned, stuff like the Windows XP replacement, 14nm transition/Moore's Law and mobile success (plan for 40 million tablets announced by Intel) have all been known for a long time before. How can he, as an analyst, justify being wrong with things he must have taken into account quite a while ago? Who is listening to those people anyway (except FPG that is)?
Having said that, and I like the current run up, I guess you all know that this is just as unjustified as the beaten down price a year ago. There is still no real success in mobile visible until now. Moorefield is good but let's hope the BOM issue and power drain is not as bad as it seems (7$ a piece is basically a giveaway, everyone can win market share like that). Intel's future roadmap still seems to slow in execution and for Broxton there really is only hope that Intel gets it right this time (3rd or 4th trial that is).
The Nexus 8 (or 9 or whatever) is pretty sure based on Nvidia's Denver (they demoed a game with great visuals on that platform - what hint do we need more?). So, again, no great design win in mobile in sight.
The recovery at the PC side is not sustainable so far. Windows 8 still sucks and that drags down PCs. Chromebooks won't replace that for a while and initially not at the high end. Emerging markets will continue to pull PC demand in the (long) future. I am very confident about that, since those people also need a workhorse at home (no matter what FPG claims). This alone will lead to only slow growth for PCs in the years coming. Not enough to offset for the high costs of 10nm fabs and beyond.
The server market will remain strong for Intel, very strong I think, so that's definitely a plus. Can't see any whimpy ARM making inroads here. Intel's execution is just great and I hope they keep going on like that.
What is really missing is progress in Intel's foundry business by winning key customers as well as filling those fabs with new business, e.g. in-package high-speed memory. There's still no visible plan as to how to fill that excess capacity in the future.
If Intel fixes all that, a much higher share price than the current one is justified. As of now, I don't think Intel is priced right, with the information we have. There's much hope priced in (right after depression was priced in a year ago).
I'd expect Intel to never produce with bad yields since that doesn't make economical sense for them. It certainly also doesn't for foundries, but they have contracts to fulfill. I'd say they'd serve Apple with their chips even in case yields are bad since otherwise, Apple would need to shift their iPhone launch. I doubt Apple has contracts which even remotely would allow for such a delay and shifting an existing design back to 28 nm quickly is impossible.
For Intel, it is different. They deliver their own parts once they are ready and ready there means, once it makes economically sense for them. This may also be the reason that Apple had to do these lukewarm upgrades of the Mac line, since Broadwell just isn't there yet. Those contracts seem to look a bit different, which is good for Intel.
@Tenchu
No, I meant fastpathguru. Sorry for the confusion. The second quote was from his post.
@fpg
I understand it just sucks to be wrong. Intel at 29,99 just now - nice scratching at 10 year highs.
Anyway, we'll see in the next ten years about the future of the PC. As of now, it is the tablets that start to stagnate, not the PC. Seems like hopes for tablets have been too high and people still need a device to work on. But yeah, now it's the phablets that replace the tablets and soon also the PC - and then it's your watch which does all that ... (to be continued)
Great, what an analysis. We don't do our work on tablets from now on. Who did ever believe in that story? Ah, Andy Grave did, right ...
Don't let them fool you about what 14nm actually means. Since when did any of the semicos or foundries leapfrog a full node? In this case, Samsung would just skip the 22/20nm node and be going straight to 14nm Finfets. Even Finfets haven't been seen by foundries until today. Only Intel has them. That's no piece of cake and I bet Samsung can't do all of this in a single step. I suspect their so called 14 nm is just a relabled 16nm which actually is the 20nm node with Finfets. It's just names.
Well, that's simple. If TSMC would say no, Intel would ask Globalfoundries or Samsung to take over. TSMC doesn't have the standing to say no. Intel has.
Fudzilla on Braswell.
It seems that Braswell is just the Airmont successor of the Silvermont/Bay Trail based Celerons. So it seems that Cherry Trail is shifted to Q1 2015. I really hope Intel did also improve the core a little to realize higher speeds at less power consumption for tablets. An Airmont successor to Moorefield would be nice as well, until Broxton releases. Intel just needs to do some incremental steps and improvements from here to stay head to head with the competition until Broxton arrives.
I really hope they are making progress quickly with the integration of comms in 14 nm. Hope to see Broxton with integrated comms in Q1 2016. If Broxton really is the best performing and lowest power SoC in the market by that time, it will really hurt ARM based vendors. A 10nm release of Broxton right in time may erase them.
By the way: My impression is that Krzanich really changed Intel from the inside. It's good to have an old school engineer back at the steering. He seems to have the insights to make his technological staff do what is necessary to just plain succeed. I'm impressed so far.
Russ Fischer about Intel's Fab capacity.
I think he's scratching the surface of the most important question about an investment in Intel. I never see any questions about that in the Q&A by analysts. Either those guys don't care or they don't have a big picture.
Anyway, Russ Fischer's numbers are a bit off at time. He claims the mobile business is only worth 10 billion. I'd estimate it at 30-40 billion once mature. But, in principle, he's right. For Intel to grow and pay for those future fabs, it needs more than a win in mobile. The foundry business is another possibility, but Intel likely won't take the customers TSMC and Samsung take (i.e. competitors). On the other hand, they can get the well paying, high end ones like Altera. So that's a nice add up with good margins I guess.
But still, this alone won't do it for big 10nm fabs. I think Russ Fischer is pretty correct with his assumption that Intel will somehow reenter the memory business. Times have changed, only a few players are left and their margins are actually quite good. Not to forget that filling Intel's fabs increases their margins of other products as well, since fixed costs decrease for each chip. On the other hand, producing memory is different to producing logic, therefore it is not going to be easy to switch between these two. Still, Intel is already producing DRAM as some sort of extended cache with Iris Pro and also for Xeons. I think that makes sense and gives a nice boost to today's systems where the memory tends to be the bottleneck.
What is left probably is the most interesting: The joint venture with Micron in flash production. Flash is going to grow big time since it will replace almost all HDDs. It is fast, robust, low power and reliable. Prices are falling but those flash cells can't be shrunken much further, since it costs data rentention time. Therefore, I suspect flash to not shrink that much in the coming future. That's actually a good, since it keeps prices up and eats much wafer volume.
What is known is that Micron is not up to date process wise with their flash tech. To catch up, they could team up with Intel and build large volumes of flash in mothballed Intel fabs. Also 3D packaging would be nice for Intel and Micron to develop together, since it is needed for both, Flash and mobile SoCs.
It is hard to find in-depth information about the IM-Flash joint venture. So, here's the question: Does anyone know some of the details of this joint venture? How Intel and Micron are cooperating in this business and what announcements have been made? How production lines and fabs are shared and how this is handled financally?
Samsung smartphone design win with Moorfield.
That's, on one side, good news, on the other side there are things mentioned that don't sound that great at all:
@Andy
You can buy that soon. It is called Asus Transformer V. Have a closer look at it. Ah, wait, no, sorry, doesn't suit the way you play your options ...
Yeah, these Nexus 8 rumours are really interesting. We have three participants that are rumoured for the design win: Nvidia with Tegra K1, Intel with Moorefield and Qualcomm with some unknown SoC.
I think what we can be sure about is that Google will introduce an updated version of Android (already announced for Google I/O) which will support 64 bit CPUs. Now, the only one in the market today with capable 64 bit SoCs is Intel. Together with its decent LTE modem XMM7260 (available June they said) it makes an almost perfect match for the Nexus 8. Also, since the Nexus devices initially were meant as technical references for new hardware architectures, Intel's x86 would seem like a natural choice.
Nividia, on the other hand, would likely deliver the most capable 64 bit platform with Denver. On the other hand, this was said to be a second half 2014 story and I would be surprised if Nividia will be ready soon enough. That would be quite a risk for Google - I kind of doubt that Nvidia can execute faster than Qualcomm here but I may be wrong. I am pretty sure that they both didn't plan a 64 bit design that early, until Apple's A7 popped up. Additionally, an external LTE modem may be costly since, as far as I know, Nvidia doesn't provide one and using e.g. a Qualcomm modem would probably be connected with a quite high price in order for Qualcomm to kind of force their own design win. So Nvidia for 64 bit would probably be a high risk for Google's schedule and a not so cheap option (the latter which they may not care that much about). On the other hand, Nvidia may accept very low prices in order to push their somehow desperate mobile SoC plans. There are already rumours that Nvidia complained about Intel's aggressive pricing - makes kind of sense in that context.
And last but not least we have Qualcomm. I don't think they care much about this in terms of this design win - their designs already run on many tablets and phones so they are very well supported. On the other hand, they don't want Nvidia and especially Intel to win this, so they might be very aggressive in pricing, especially since they have a very competitive cost structure due to their high integration level. Then, again, there's no high end 64 bit design available early enough, so I'd guess this alone would kick Qualcomm out of the race. Google may be sacrificing performance though, since they rarely used the top notch SoCs in their devices - but on the other hand this might be something that cost them sales significantly in the past.
To me, the logical choice would be to use Intel's Moorefield. Google will get mature 64 bit hardware, a decent LTE modem at supposedly low price and very strong support for the software porting - something that Nvidia and most probably also Qualcomm will not provide to that extent. The only thing standing against Intel as the winner here seems to me that Qualcomm and Google just want to keep Intel out here by force. Qualcomm could offer an LTE modem for Nvidia for a cheap price and Google could select Nividia just to not let Intel grow too strong - something they are not interested in.
In this context, all the existing rumours make quite some sense and Google may have started to design multiple versions of Nexus 8 in order to finally choose in a late stage (say, Moorefield tablet from Asus, Tegra K1 from HTC). In the end, the decision may depend on how well Nvidia and Qualcomm execute on their 64 bit plans - I do have some doubts with Nvidia here and Qualcomm's original schedule doesn't allow for a decent 64 bit design win for the Nexus 8.
In any case, this design win will not make money for any of the participants except Google - which doesn't care much about it anyway.
Moorefield in the Nexus 8 would be great news for Intel and an important step towards better x86 app support. I am really looking forward to Google I/O.
Thanks for the Info! Qualcomm is without doubt the king of integration. Great company and a very tough competitor for Intel to beat.
Does anyone know whether those are integrated in common mobile SoCs?
First benchmark results of Moorefield
As expected, it seems to be the fastest on the market at the moment. Let's hope for design wins in phones.