Separate names with a comma.
Discussion in 'CPUs and Overclocking' started by Roland00Address, Sep 26, 2012.
Must have been using old Cadence tools. Maybe they're still on Silicon Ensemble and PKS?
LOL, I think management forgot to open the second box that came with the installation disks, if they had then they would have realized they were supposed to pass these out to their engineers:
How does this compare to the ARM/low power cpu scene?
Good stuff going on in here. I'm sure it more than likely has been beaten into the ground in old threads, but does anyone have any solid info on why Intel chose not to grant NV a license? As IDC pointed out, they really don't have the funds to take on Intel at the high end and at best might be able to outpace Atom (which Bobcat did resoundingly with nary a dent on Intel's fortunes). That and the fact that they have quite limited CPU experience (tacking a small GPU on a few pre-designed ARM cores provides little if any useful experience for designing an X86 CPU from scratch). I don't know if I'd consider getting into X86 throwing money into a black hole though. Wouldn't be easy, but if someone took a Transmeta type approach they might be able to capture the smartphone/tablet/entry level notebook market.
Too bad really, because that will mean the death of things like this
AFAIK, other manufacturers don't produce embedded dev boards like that, only ti really.
Have faith in the natural cycle of things. Something will rise up in the face of this opportunity. Something always does. From death comes new life, oftentimes it is a new life that could not have otherwise existed had death not preceded it.
We all love hearing about the dinosaurs but the age of mammals (Cenozoic age) could not have existed were it not for the ending of the age of dinosaurs.
TI was a living dinosaur, a relic of the mid-20th century. In the late 90's they began to metamorphose - they divested nearly all of their business units (defense, solar, memory, etc) and bought Burr-Brown (analog). They've spent the last 10yrs taking steps to eliminate the last vestiges of the TI that once was - CMOS.
10yrs from now the TI that will exist will be 100% analog, metamorphosis complete. Same name but having nothing else in common with its heritage. Exactly the same as AT&T.
Freescale. Pretty sure the i.MX 6 is reference.
Also all those China chip companies are producing rather capable chips (when everyone thought their chips would fall flat).
Example is Rockchip's RK3066. This dual core 1.6GHz beat the Tegra 3 in quadrant, and not by a hair, I think it was something like 4288/4085? But yea, you get the point.
Hey, I'll bet they're still selling the Ti-83 Plus until the earth crashes into the sun
Looks like the fallout is a layoff of 1700 people.
That sucks. There's too many players left in the mobile chip design business and so I wonder who's next. On the other hand most of those players have a large cash cow driving them.
I wonder what NVIDIA is thinking.. they don't have integrated wireless right?
nVidia bought Icera last year and is taping out 28nm discrete radio modems and SoCs with an integrated modem right now.
Freescale may be changing things up, we will see how i.MX fares.
"Freescale, which makes and designs microprocessors, will fire 150 of its 450 workers at its Herzliya headquarters. The company fired 120 workers in 2009, when it still had 600 workers. The firings were somewhat surprising, said analysts, because the company just last month advertised that it was hiring 30 engineers."
Isn't that a "soft"-modem? How will that hold up power consumption wise? And didn't Intel buy out a wireless company 2-3 years ago (Infineon Technologies)? Where's the product from them?
Sometimes you got to fire the Wally's and hire a few Dilbert's and Tina's. HR doesn't like it when you do it, but companies go bankrupt if they don't clear out the deadwood that accumulates over time.
Icera is "programmable" radio modem, so no software thing.
Power consumption? No clue but it will be integrated and on 28nm.
Letting go that high of a percentage probably means dropping some products.
True, but probably for good reason. If your products aren't generating enough revenue to cover the payroll then it is time to cut the bottom (both in terms of products as well as employees).
The issue with TI cutting 1700 design employees is emblematic of the problem with TI in the first place (i.e. the reason why they find themselves in the position of needing to layoff 1700 employees)...it is the same problem TI had when they let go of the majority of their CMOS R&D process development engineers: management doesn't have a clue what to do with them.
The problem with TI is that TI's executive team has so little vision for the future of TI that they look at 1700 experienced seasoned design engineers, then they look at the semiconductor market, and then they look at each other and go "eh!? wtf do we do with all these design engineers?".
And because there isn't a sound strategy, no vision beyond M&A within the existing analog sector, management has no idea what to do with those engineers. So they just let them go.
I feel for the 1700 TI engineers and their families because of the insecurity and unknown future that comes with the layoff, but honestly they will be far better off for it in 2-3 yrs time when they have their careers firmly growing under the helm of actual business leadership as employees at other companies.
Well, makes sense why we have so many TIers banging down the doors at work.
Samsung must be delighted.
A question for the wise IDC mind
How does samsung fit into this picture in the future?
They are diversified as hell - is that a stategy to keep afloat if say RAM fails or some other fabbed design circuitry does?
Is it their relatively large fab capacity\ability that also keeps them going?
When the apple v samsung thing finally settles - would they be interested in AMD's IP?
Not sure if they have the culture\moxy - but they do have the financial resources to gain something from a firesale no?
Samsung is a behemoth corporation with revenue that absolutely dwarfs Intel.
I don't think many people truly understand the size and scale of Samsung, they know what they see of it as electronics consumers (LCD's, phones, ram) and that is probably about it.
Samsung's 2011 revenue was $248B. Intel's was a relatively paltry $54B in 2011.
Samsung employs 369,000 employees. Intel just 100,000.
Samsung is responsible for generating one fifth (20%) of S.Korea's GDP.
Think about that for a moment, imagine what it would mean to everyone's livelihood and social structure here in the US if Apple or Intel was responsible for generating 20% of the USA's GDP.
Samsung is not just "yet another business" that the competition must contend with. Samsung is essentially a highly capitalistic national movement, an anti-communistic planned economy of sorts.
They are diversified because they are literally larger than any single business segment within which they might attempt to participate.
Samsung is to S.Korea what the Mubadala Development Company (parent company of GloFo) is to the United Arab Emirates.
Samsung would be interested in AMD's IP only if Samsung was able to envision a way to compete with Intel using that IP.
Personally I don't think they are. I think Samsung sees the future of high volume sales and ubiquitous computing as being that of ARM and smartphones.
But make no bones about it, if Samsung decided to put their resources into it then they could outspend Intel by nearly the same margin that Intel outspends AMD. For now though, they seem satisfied to spend their time and money growing in a different direction. The impending collision between Intel and Samsung thanks to Intel's atom will be interesting.
Ehhh, you have to look at Net Income. Samsung's revenues are high because it also ships a lot of high revenue/low margin products. Look at Net Income of INTC versus Samsung, and the playing field levels out immensely. Then take a look at Samsung's electronics division and see how it does compared to Intel's - not even close.
WTH IDC? Samsung's heft is due to multiple industries at generally lower margins than what Intel has. What are you saying, that Samsung can just spend nothing on all of their other divisions and pour everything into one division to compete with Intel? Please think about that for a moment.
It is a more fair comparison to compare Samsung's semiconductor divisions to Intel, if that is even broken down.
However you are right about there being the potential for long-term cross-subsidization. But it's not like the enormous advantage you are making it out to be. And pouring $ into something doesn't necessarily work--see Samsung's camera division for instance.
Samsung Electronics 2011 Net Income: $12.06 billion
Intel 2011 Net Income: $12.6 billion
Seems pretty close to me.....