Semiwiki on LinkedIn Semiwiki on Facebook Semiwiki on Twitter Register / Log in 
Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/where-do-we-see-ti.5387/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Where do we see TI?

P

Pawan Fangaria

New member
TI's strategy of getting out of wireless and expanding its Analog and Embedded processing business seems to be paying well for now.
On 26 Jan, TI would be reporting Q4 2014 along with the whole year 2014 earnings.

It's expected to close on a strong note - seekingalpha report here

In last 3 quarters, its revenue increased by 6.5% and net income by 20.9% compared to last year.
With its increased focus on micro-controllers for industrial, automotive and consumer electronics segments, including IoT, can it maintain or rise in top 10 rank?

The flip side is that micro-controller prices will keep on declining with rise of IoT and several players in that space.
 
The 4th quarter financial report is just out. TI Investor Relations - TI reports 4Q14 and 2014 financial results and shareholder returns
That seekingalpha report was probably written by two people. It gave positive comments and very low share price suggestion. :rolleyes: Errrrrrrr
TI is quite robust now. After shifting from digital base band battle field to automobile arena, they gain the full momentum in growth in my opinion.
Can Finfet make analog circuit? Perhaps with higher budgets and more efforts, we can get somewhere. The point is that analog share little scaling risk. And their debt profile is quite healthy. Accordingly I expect TI to rise.
 
Yes, results are good with 8% increase in revenue, 85% of revenue coming from embedded processing.
However, its automobile focus may not prove to be a long running business, considering low base in that segment.
 
Pawan Fangaria said:
The flip side is that micro-controller prices will keep on declining with rise of IoT and several players in that space.
Click to expand...

I don't agree with that, at least to the degree some are projecting. 8-bit and 16-bit MCU prices have been at commodity levels (34 cents is the lower end for TI MSP430) for years and have not dropped all that much as volumes expanded. Low-end MCU volume pricing is really a function of package and features like flash and analog integration. In other words, that isn't going to go from 34 cents to 3.4 cents if volume suddenly expands 10x - which it won't on a mature part.

32-bit ARM MCUs are closing in, the cheapest Freescale Cortex-M0 part I can find readily is 58 cents. Higher end parts with more features climb into the $1-$5 range.

TI is also doing a good job of adding value to MCUs. Last week they announced a member of the Piccolo F2806x family, a 32-bit DSP core, for motion control at around $10.

Is there pricing pressure and competition? No doubt. I don't believe the IoT changes the economics radically, however, at least while the technology is what it is: silicon, flash, plastic packaging. I think MCU volumes go up at a reasonable rate - 12% to 15% a year - and prices on those $1-$5 parts work their way into sub-dollar as they reach volume.

The chances of "trillions" of IoT devices driving the price of everything - MCUs and MEMS sensors in particular - to something like 2 cents ... is zero. Most of the industry goes out of business at those kinds of price points. We may have a few break through technologies, like the ThinFilm/Xerox printed electronics deal, that can get close in RFID-like applications.
 
I agree that the MCUs for IOT will not go down like 2 cents, but they will be on declining path until being stagnant somewhere on the lower side. Having said that, TI can have good fortune even there, provided it caters to larger segments of IoT. Automotive segment has a smaller base, okay it can show significant growth in next couple of years, but can stagnate after that.
 
Pawan Fangaria said:
I agree that the MCUs for IOT will not go down like 2 cents, but they will be on declining path until being stagnant somewhere on the lower side. Having said that, TI can have good fortune even there, provided it caters to larger segments of IoT. Automotive segment has a smaller base, okay it can show significant growth in next couple of years, but can stagnate after that.
Click to expand...

Dear Pawan, I have to disagree with your forecast of the Auto segment. Automotive is almost the perfect opposite of wireless: very long design (and qualification) phase (in the 5 y range), it takes also quite a long time for an IC to penetrate all the car segments, starting from the high-end. But, once installed, the volume quantities are decent (several M units) and stay for long time (10 years?). Thus, if you have been patient, the ROI will come, but you have to calculate it over 10 years - not 6 months like with Mobile Apps Proc!!
 
Just an adder about TI: I have joined TI in 1992, at that time the mantra was "Application Specific Products" (ASIC, DSP and soon wireless) and I must say I have enjoyed a lot my job of ASIC FAE (then Marketing). I am not sure that I would be so happy to work with TI today... Analog is growing more than the rest of the ICs, see:View attachment 13117

But, as far as I am concerned, Analog sounds like commodities... no more fun on the design side!
 
Eric Esteve said:
Dear Pawan, I have to disagree with your forecast of the Auto segment. Automotive is almost the perfect opposite of wireless: very long design (and qualification) phase (in the 5 y range), it takes also quite a long time for an IC to penetrate all the car segments, starting from the high-end. But, once installed, the volume quantities are decent (several M units) and stay for long time (10 years?). Thus, if you have been patient, the ROI will come, but you have to calculate it over 10 years - not 6 months like with Mobile Apps Proc!!
Click to expand...

Hi Eric,
I must admit that Automotive has long life-cycle and provides opportunities to reap benefits for longer years on 'more or less' one time investment. In fact, if you see one of my last year's blog on TI strategy (https://www.semiwiki.com/forum/content/3395-secret-ti%92s-success-analog-embedded-space.html), I mentioned about a quote by Rich Templeton in which mentions about "~10 years life cycle of a product or design in automotive and industrial segment against a product in personal electronics (aka smartphone) having ~12 to 18 months of life cycle".

But today I see, the point is automotive has very small base. Even then it would have worked provided TI would have, say 70% market share. But that seems to be a far fetched goal. Last year, I was seeing TI's to represent about 13-15 percent of global embedded market. And microcontrollers' market share was below 10%. Well, high prices can sustain for some time, but volume is the key in long-run.
 
By the way, even after exiting mobile business, TI still has presence in WiFi chip business.
It unveiled SimpleLink CC3100 and CC 3200, internet-on-chip solution for home, industrial and consumer electronics.

Look for details - TI SimpleLink Wi-Fi Family

This provides good business for TI in the web connected embedded segment. Looking from IoT business perspective, does that provide leading edge to TI?
 
Pawan, I saw this News, and I was impressed. Did you noticed that these chips were using Network-on-Chip from Semiwiki good friend Arteris? It look like the NoC is now an active component of on-chip power management, and that's really make sense as the NoC can access all the main chip functions. NoC integration into such SoC is also a proof of pervasion for this IP...
 
Yes Eric, NoC is going to become an integral part of SoC with several components that need to optimize traffic between them. And you know the advantage it provides to an SoC in many aspects. I guess, TI is doing well here!
 
Pawan, we know pretty well how NoC can optimize traffic, avoid layout congestion, but so far NoC has been integrated into large SoC, I should say SoC being really system-on-Chip: multiple CPU and GPU, DDRn memory controller, multiple high speed I/F like USB2, USB3, HDMI, SATA and many more (MIPI DSI or CSI, or PCI Express), security, video... up to100 functions in some cases.

Not to under-valuate the CC3100, this is a low cost SoC (good idea to target IoT) which is probably 1/10 complexity of the above described SoC, if not less: it fit into a 64 QFP. See the CC3100 block diagram below.

The main two reasons why I am (very positively) surprised by NoC integration into CC3100 are:
- NoC is explicitely used to manage Power Management
- NoC pervasion into mid to low complexes SoC is strong signal of NoC penetration in various segments and many more IC than just the very large apps processor, CE (HDTV) or server and network processor

See the CC3100 moderate complexity here

View attachment 13467
 
Thanks Eric! Yes, you made it explicit :) I was meaning that NoCs will be used more frequently in SoCs. Here is the live example, and it's good to see it in low cost SoCs for IoT applications. If TI can do well here, they can grab good share of IoT market.
 
You must log in or register to reply here.
Back
Top