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.
I'm saying that Intel will probably give in and fight to win the Apple foundry contract
for ipad pro cpu? i think that's extremely unlikely. if not ipad pro, when, which process? 14 nm ? it seems even altera can't start getting chips in that process till late next year and they have been at it for quite a while. 10nm?
This is the "build Apple chips" playbook
are you saying ipad pro will have an x86 in it?
half the storage
how much of that storage is really available with windows installed?
can someone quote available storage on an out of the box device for both?
i was hoping to replace my sandy bridge 17" notebook with a lenovo w540 or dell m3800 (like) with broadwell but that's not going to happen anytime soon now.
i wish someone made high(er) resolution 17" devices.
a quality ultra-book is the winning ...
interestingly i read that the prices of ultrabooks were going up because of a lack of demand and higher-end features being added to them. there is certainly a market for those but i think it's going to get incrementally smaller over time.
All very nice improvements, in my mind.
i can see that.
The PC's are vastly more compelling than a year ago.
that's a huge exaggeration, imo.
I think the average end-user would call that pretty dramatic "innovation".
alas that supposed innovation is not being reflected in numbers they are buying them. there is consolidation in larger phones/smaller tablets but overall the numbers are still increasing there and decreasing in notebooks/desktops.
i think we are starting to see stagnation of innovation in the pc market because of lack of strong competition. with increasing ultrabook prices i don't think this is good for the industry overall and intel specifically but they can bleed the golden egg laying goose while it survives.
If Microsoft thinks Google is thinking about it, you can bet that Microsoft is thinking about it.
actually microsft mnt view is already designing asics & socs.
I think Intel will do semi-custom products for these customers
intel already does semi-custom chips for at least one of them. they even had a semi-custom bug in one of those chips
i think this is interesting:
Each major system company will make their own application-specific SoC.
http://deepchip.com/items/0535-11.html
nexus 7 is on sale for $149 on groupon. nexus 8 must be coming. any credible news that it will have an x86 in it?
I was wrong....
about the weather?
Dell offers 64-bit ARM microserver proof-of-concept for hyperscale on the heels of Open Compute Summit momentum
http://en.community.dell.com/dell-blogs/dell4enterprise/b/dell4enterprise/archive/2014/02/04/dell-offers-64-bit-arm-microserver-proof-of-concept-for-hyperscale-on-the-heels-of-open-compute-summit-momentum.aspx
Just because the base layers are 20nm, doesn't mean building 16nm FinFETs on top is easy
transistors are built on base layers, not on top. What's on top (starting above contact or m1) are the metal layers.
Please don't post Nenni's garbage here.
I have read the stuff you post ...
I hope this means Krzanich finally means to take the reigns on the $7.7B acquisition, and do something useful with it.
like give it out for free on mobile devices ...
http://www.digitimes.com/news/a20131231PD205.html
Shares of MediaTek, MStar trending upward prior to official merger
Share prices of MediaTek and MStar Semiconductor have been trending upward on the Taiwan Stock Exchange (TSE) recently as the two IC design houses are to officially merge on February 1, 2014, according to market sources.
More than 1,000 or one-third of MStar's R&D staff are to be incorporated into MediaTek's handset solution business unit after the merger, which will give a boost to the competitiveness of MediaTek's smartphone solutions, said the sources.
In addition to smartphone solutions, the handset solution unit has also been engaging in developments of feature phone solutions, as well as touchscreen controller ICs, power management (PWM) ICs and other handset related IC parts.
However, MediaTek may face strong competition from China-based Tsinghua Unigroup in the China market in 2014 and beyond as Unigroup has announced the acquisition of China-based IC design house RDA Microelectronics, the sources noted.
Since Unigroup also acquired China-based handset solution vendor Spreadtrum Communications in July 2013, an integration of RDA and Spreadtrum could make Unigroup a strong contender in the smartphone solution segment, commented the sources.
MediaTek, Qualcomm and Unigroup are to compete fiercely in China's smartphone solution market later on, stated the sources.
i wonder if an opensource core can take off in any meaningful form:
Startup Taps OpenRisc for Booster
By Loring Wirbel
Cellular-repeater startup Nextivity used an Altera/Synopsys Nios CPU core, synthesized in an ASIC, for two generations of what it calls the Intelliboost Baseband Processor. This product serves in Cel-Fi, the company's indoor-coverage extender for heterogeneous networks (hetnets). But for its third-generation Cel-Fi system, Nextivity designed its own proprietary processor using six CPU cores based on the OpenRisc OR1200.
Cel-Fi boosters are cellular-repeater systems for homes and small businesses. Users place a window unit housing the network CPU indoors where the cellular signal is strongest. This unit filters, boosts, and retransmits the signal using the unlicensed 5GHz band to a companion distribution unit positioned where the signal is weakest. The retransmitted signal supports up to four 20MHz cellular RF bands carrying any combination of WCDMA, HSPA+, and LTE.
Nextivity engineers considered a single-chip processor incorporating six ARM9 cores based on the ARMv5TE ISA, but they instead turned to OpenRisc for greater flexibility and lower licensing costs. OpenRisc is the flagship design of the OpenCores community, a Linux coalition formed in 1999. OpenRisc 1000 is an architectural description of a family of 32- and 64-bit processors, and OpenRisc 1200 (OR1200) is its first instantiation, written in Verilog. So far, Beyond Semiconductor, Orsoc, and Samsung have developed commercial OR1200 implementations, although Cadence also uses the synthesizable core as a reference design for EDA tools.
Microprocessor Report subscribers can access the full article:
http://www.linleygroup.com/mpr/article.php?id=11134
Cloverfield+ phone is unlikely to provide a compelling user experience
if that's true, what if it teaches reviewers/users that intel phones aren't good enough?
Why would they? Snapdragon 800 is currently better than Z3770 for all intents and purposes for this space.
how do you think intel 14nm chips will compare against 20nm chips from qcom/samsung/apple?
the speculation right now is that the new "large" tablet from apple will have an a8/9 in it (my speculation . it seems samsung is also not going to use an x86 in their large tablet: http://www.phonearena.com/news/Samsung-could-name-its-12-tablet-Galaxy-Note-Pro-more-specs-leak-out_id50558
custom-built network routers, and custom-built Ethernet controllers, and custom-built hard drives for their storage networks
they already design their own switches and fpga based flash controllers ...
http://www.theregister.co.uk/2013/11/18/cisco_cloud_problem/
even though Apple is using TSMC to make the 20 nm A8 for 2014, word is Apple will be using Samsung for the 14 nm A9 chip sets in 2015!
interesting.
http://www.deepchip.com/items/0528-10.html
TSMC moves 16nm FinFET to risk production
http://www.digitimes.com/news/a20131212PD211.html
there will not be a single customer for these chips outside of Google research.
the question is whether google platforms will use those chips if they are made. if they do, it would be for a good reason and if google platforms group have that reason, facebook platforms (or what ever that department is called over there) will see it too which will force fb take similar route too.
I think this is important to note:
Google ... has made no decision and plans could change, said the person.
so there is no plan, and that non-existent plan could change. one of the known unknowns, eh?
google chip jobs:
I think it's misleading to check if google is advertising for new asic/soc design jobs for the arm processors. I think a better metric would be to go to linkedin and do an "advanced search" with asic keyword and current company at google. This returns 111 names, with a significant chunk of people who have a lot of experience designing asics and socs.
This is 100% pure bullshit
i would not be that hasty dismissing this. google designs its own network switches, nand controllers with fpgas, x86 motherboards. they have a decent number of people who know how to make socs/asics. it is not beyond the realm of possibility for them to design or have designed a fully custom (for them) arm soc suitable for their server farms.
as to arm for macbooks, i think you should consider whether the rumored 12 inch ipad will have an a8 or x86 in it. what is exactly the difference between a 12 inch ipad with a bundled keyboard and an macbook air ? bloomberg's timing might be early but that is the deciding factor as far as I am concerned. what cpu will the large ipad have?
Where do these analysts get their stuff from?
Ashraf ?
http://news.investors.com/121213-682683-intel-might-acquire-broadcom-to-expand-in-wireless.htm
He upgraded Intel to neutral.
IMO ARM based servers will be a novelty that will be lucky to grab even 1% market share by revenue.
at what time frame, 2015, 2020? ever ?
In addition, contract orders for Apple's 20nm A8 processor chip are expected to account for about 4% of TSMC's sales starting the second quarter of 2014
http://www.digitimes.com/news/a20131205PB200.html
here is what they said "It is anticipated that Qualcomm Gobi 9x35 and WTR3925 will begin sampling to customers early next year. "
not in 2014 but early next year. xilinx 20nm chips are sampling now with wider 20nm in q1. it makes sense that they will sample in q1 too.
The schoolyard children are now singing creepy jumprope rhymes about them while at recess.
that's not new; it's been going on since netburst days.
I still stick to the idea that Bay Trail is "Pentium M". We're still few years away from "Merom/Conroe". Now I'm not completely optimistic about "Merom/Conroe" against ARM incumbents either.
for that to work, arm incumbents have to be as incompetent as amd. do you see that with apple/samsung/qcom etc?
I have been waiting for the lenovo w540 for more than 3 months now. sometimes announcing a product and making it available are different things ...
Intel's tablet processor shipments for 2014 are expected to reach only 30 million units, accounting for about 10% of total tablet processor shipments and lower than the company's forecasted 40 million units
so they can't sell them in the numbers they want even with a $25 wrapped around the chip? not promising.