In June of this year it' will be ten years since Apple made the announcement to move toward Intel on their computer side. Before this there were Motorola and IBM RISC processors. I clearly remember an Apple fan explaining to me that the RISC processors were twice as good as Intel's CISC processors which is why they were equivalent at half the speed. The reality was that twice as fast was an exaggeration, perhaps 15 percent was more like it.
In the past ten years look at how many times Intel has improved the line from the initial Core-2-Duo processors. It's kept Apple interested in Intel. Apple has touted the Haswell processors in their MacBook Pro. Apple has developed technologies like Thunderbolt with Intel. Apple and Intel have been partners.
For years now we've seen speculation that Apple was going with Intel. Some have suggested it would happen by the iPhone 6. The question now is looking less about "if" and more about "when." The problem is that once Apple goes with Intel on the iPhone line Intel will most likely keep the business locked up very long term .
As far as a Pentium processor is concerned, I don't know You would think that Apple would insist on a home brewed RISC processor from Intel. But we may be getting into a world where smartphone momentum forward is based more on the GPU than the CPU. We're used to thinking that CPU is all that matters when in the entire SIP or SoC is what is taken into account.
I'm still waiting for whacky-Ed's prediction of the fortune 500 switching from PC's to ARM based Apple PCs to come true. It's a good thing I don't trust them for any relevant information.
The problem is that unless you're talking about an iPhone or perhaps a high end Samsung device like a Galaxy or Note consumers aren't going to pay premium prices for marginal improvements in phones. The Zenfone 2 is showing that consumers are willing to embrace non-players if they can get an outstanding price/performance phone.
You get a 4K TV set and what will you do with it? How much 4K content is out there? How much will there be in 3 years? It's an interesting technology but it's probably 5 years before it hits critical mass.
I think that consumers are still in the mode that TV is something you go to in your house. The AT&T UVerse commercials were pointing out that TV can be enjoyed in places that are nowhere near an cable box. When I plugged in a current smart tv recently in my home I was surprised how much good content I could get off of wifi. But then digging deeper it was clear that it wasn't enough.
The advantage that this device has is that it's self contained and can be hidden on the back of a TV. You can pick up a 32" or 40" TV and move it around the house. Take it the backyard and put a wireless keyboard dongle on it and you've got a supersized mobile computer.
I agree the pricing is high. I would think they would sell better at $99 and $49.
If I could pick up a 32" or 40" TV with one of these units for $250 I would start placing them around the house over time. I would like to have TV sets without wires and cords in a few spots around the house. But I am still getting used to watching TV on my computer. I am still in the mindset that when I really want to watch TV I need to go somewhere comfortable.
28nm is a fabulous node. I am more skeptical about 20nm and below for ARM They should be doubling down on 28nm. 28nm offers a fabulous price/performance ratio.
I read another article that explained the Galaxy S6’s Exynos 7420. After a lot of praise the benchmarks showed marginal performance increases over the 5433 even though it was a so called 14nm true 64bit cpu versus the 5433's 8 32 bit cores. The real performance improvement came from the GPU. ARM is touting the Cortex A72, but has been very vague about real world benchmarks. It does state in it's graphics that A57 is 20nm and the A72 is 16nm finfet which should be out in 2016. And yet Samsung is claiming 14nm on their A57 (Exynos 7420). A document Nvidia provided a couple of years ago showed that 16nm will come at a price twice as great as 28nm. So, once again, how much more will it cost and what will you get for the money?
"The Cortex-A57 simply won’t end up in a lot of devices, as it only makes sense on 20nm and 14/16nm FinFET nodes, so chipmakers will have only one choice – churn out more Cortex-A53 parts at higher clocks, with faster GPUs and better LTE support. Unlike last year, they don’t have the option of using four cores (A15, A17, A9 and A7), as they can only use A57 and A53 cores, but the A57 simply doesn’t work for most market segments. The Cortex-A17 looks like a very tempting alternative and MediaTek already tapped it for some parts, but this is a 32-bit core, positioned below the Cortex-A15 and Cortex-A57. While the A17 is a good performer with a good price/performance ratio, consumers demand 64-bit chips, plain and simple."
Do you know anybody who has an Intel based smartphone? How about an Asus smartphone? Chances are you will by the end of the year. The Zenphone two comes with an Intel Atom and at least 2 GB of ram. One of the models has 4GB of RAM. Quadcore, 5.5" inch HD display with 30 megapixel camera. It starts off around $200. Currently there are million people who have pre-ordered this phone in China. Asus is expecting to sell 30 million units this year. Not bad for a company that currently isn't in the top 5 of smartphone makers.
This is just one phone model. Intel can very easily pick up 25% marketshare with no name manufacturers within a year or two. The other part of the story is that this low end phone will put pressure on the high end phones to justify their pricing. Like I said all along, when the white box manufactures in China start private labeling phones for the carriers using Intel processors and Intel marketing dollars it's going to change the game very quickly.
I believe where they are going with this is to go after the Smart TV market. I am finding that Smart TVS have around a $50 price premium. The problem with a SmartTV is that the processor lasts as long as the TV. That means that 20 years from now your TV set will have a very outdated, very weak computer.
The advantage to the Windows version is that it can run Windows Media Center. This means that you can use an Silicone Dust HD Homerun with a cable card and do without your cable box while getting all the channels you pay for. An HD cable box is around $10 a month so, this pays for itself in about a year. It will also work with Homegroup so that you can easily see pictures from other computers on the network. The downside is that Windows has a ton of DRM on it so a lot of the torrents are off limits. The 32GB is enough room to blast an evening of HD TV and movies without streaming.
The Linux version is cheaper and there are a ton of torrents you can use. Aside from the pirated movies there is a universe of interesting programing that can be found on the torrents. Special interest programing and things like video podcasts. The Linux version has no DRM slowing you down. Its also less susceptible to getting hacked. The USB out means that you can hook up keyboards, a mouse, and external hard drives. The quadcore Atom is a very capable little processor. Either version provides a lot of value for not a lot of money.
I've picked up and started playing with the Raspberry Pi newest version, the B+. It's a Broadcom ARM processor (Cortex A7 running at 900MhZ). For what it is, it's fabulous. Yet, if it was a cellphone nobody would buy it. You can think of it as a clunky smartphone board with PC IO ports such as HDMI and USB 2.0. One thing I like about it was what I thought was the IDE pins. There are 40 pins, but not IDE. They can hook up to a breadboard or control relays. It can hook up to an Arduino board to control motors and thousands of very cool projects. As a module that is designed for kids to get into programing and engineering it's incredible. You can also set it up to run an XMBC media center. It runs 1080P video and does a nice job. You can control it from your smartphone. It does have it's glitches however. You can get stuck at places and five minutes of video can play in a few seconds.
Now, as far as making it a PC, it leaves a lot to be desired. It has a native operating system for it called Raspbian. It's heavily watered down Linux in GUI. The main draw is it's ability to run scripting languages such as Python. For a little unit that boots into a command line driven environment that can run fairly elaborate scripts its ideal. But booting up into a GUI and surfing the web it chokes badly. The old Atom netbooks are light years ahead which leads me to believe that Windows 10 on this will be disappointing.
But, Intel is blowing it as well. The Galileo 2 runs at 4MhZ and is nearly twice as expensive. It lacks USB and audio and video outputs. The problem is that the Raspberry Pi was designed for kids but it seems that its the favorite of the tinkering engineering crowd. The IoT is very much alive and is in the phase of garage based start ups that that in the past have given us Microsoft and Apple Computers. We're on a renaissance of a new era. I just am hoping a bit more from Intel.
Microsoft's relationship with ARM goes back decades, til around the time of ARM's humble beginnings. Microsoft correctly predicted the importance of mobile computing. They just never figured out how to make software or products people wanted to use.
Very similar to Windows 8 (except RT is gone). The third item is slightly different. I'm sorry, I am not seeing a radical departure from the old Microsoft.
The San Diego biotech scene is interesting and confusing. You'll see in the annual reports of larger drug makers that they pick up one of these small upstarts. If a company has a decent shot at FDA approval or has received approval its often cheaper for a J&J to buy the company than to try to develop inhouse. From the perspective of a stranger looking in Conatus is doing a lot of things right. Let's hope this all translates to FDA approval and all of us who hold this stock be handsomely rewarded.
I won't name names but one of my friends went to work for Conatus on Monday as a VP (they have several VPs). Her speciality is getting drugs ready to meet certain FDA deadlines. There is no promise that things will work out favorably, but at least she can help keep things on track. She does her due diligence in evaluating employers to work for and thus I am fairly optimistic about the future of this company. But being realistic 90% of companies like this one never get FDA approval. So I am cautious-optimistic.
My view of where the smartphone industry is going. It will be Apple and Samsung for the premium market. There will be a middle tier market with names like Sony, HTC and Motorola trying to stay relevant and slowly losing market share. At the carrier level each of the stores will be pushing their own Chinese/Taiwan private label brand of smartphones. This is what will be pushed on the customer who doesn't care enough to spend premium dollars on an Apple or Samsung product. Intel will be able to co-label at this level and help pay for advertising. "AT&T 4G Smartphone Powered by Intel." Peter is right, there is a large market of people who either can't afford premium or just don't care.
They are talking about 16nm and 14nm processors from GloFo as next generation A processors. How long is the current generation going to last, until 2020?
Within the next couple of weeks I will be able to break some good news about this stock. If I am not the first to break it I can at least give more insight into the rest of the story that may come out on a brief press release. This is a good time to buy this stock.
Unless the new fab is a true sub-20nm fab on 450mm wafers, it will be far too little too late.
This simply doesn't sound right on the face of the article. Sale have been good for TSMC so now they ave decided to add 16 billion in upgrades assuming that it will be ready fast enough to extend the ride they've caught recently?
I don't know, this sounds more PR in nature than anything.
Interesting that you brought up Windows, CE. As a consumer product it's dead, replaced by Windows Phone. But as an embedded platform it's still around. If you think about the code base they have for ARM with it then it stretches back about twenty years. It would be foolish to think that Intel will make every IoT device processor in the future. This code base makes a lot of sense for IoT. Perhaps there is a Windows 10 CE? I just don't see Microsoft going for another round of pain when it comes to make two versions of full blown PC operating systems for x86 and ARM. Then again, they have been known to do stupid things over and over again.