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.
Monster,
I am starting to more noise about quantum computing- which might be a game changer???
Andy,
and now the watch back plate burns the "Intel Inside" logo on to ones wrist.
FPG, on the flip side, from the same article:
There are reports that Apple A10 series processor will be integrated into MacBook series instead of Intel, although we can’t confirm the development.
Borusa,
That is one of the dumbest things you have said, so far.
Flumoxed,
Nice going, unless Intel gets to $70 within a year or so. Fact is, you know nothing.
Click-bait. EOM
IdealInv,
iPad sales have been falling & I am pretty certain they will continue their slide.
Morrowinder,
Android Tablets Are Dying
LOL, with a yearly salary of $950,000, she might afford a toolshed next to Mark Zuckerberg in Palo Alto.
Fortunately her stock benefits are much more lucrative. And if AMD stock heads further down into the toilet, the board can always "reprice" them ...
Tenchu
Chipguy, sure took a while for this case to work its way through the judicial system.
Is anyone working on Itanium these days? Seems like Oracle made the right decision in dropping Itanium support, though the manner by which they did was legally questionable.
Tenchu
Morrowinder, I too didn't have much problems with McAfee AV software, but that wasn't the point.
The point was that there was no purpose in acquiring McAfee. Very little value-add to Intel's core businesses, very little headway into new markets for digital security (despite there being a VERY high demand for it), and kind of a WTF move for a hardware company.
Personally I think Intel should just acquire the Golden State Warriors and rename their home court the Intel Arena. At least that would have advertising value ...
Tenchu
TCaE,
No, that'll be all. I'll be waiting for the proof.
Tom,
ALL antivirus-software-companies are currently in a bad shape.
Ibc,
Competitors have developed a "Cache Coherent Interconnect for Accelerators (CCIX)"
Flumoxed,
The trillions spent on defense budgets over the past 10 decades have led to lots of commercial products...in fact to all of civil aviation. What's your point?
WT,
Intel need only worry about stuff that people can (and will) buy of the shelf. Because given enough budget and a target application that's well enough defined, some kind of freak custom chip will always be able to beat a general purpose Xeon.
Thanks for the article, Dmcq.
This is indeed a major milestone for China, something that should have implications both in semiconductors and in global politics.
Tenchu
Hey Chippie, have you seen the news?
China creates world’s fastest computer without relying on US hardware for first time
https://www.rt.com/news/347620-china-supercomputer-sunway-taihulight/
The key news item here isn't the fact that China now has the world's fastest supercomputer, as it seems that record is made to be broken every year.
No, the most interesting part is that this supercomputer is based on a "completely homegrown processor" instead of Intel Xeons.
Got any insight into this? How valid are the claims? Seems like quite an accomplishment, if true.
Tenchu
FPG,
Depends on the price, doesn't it?
Morrowinder,
By using Intel, their phones would not support CDMA so it makes it harder to switch to say Verizon.
Chipguy,
My point is there is a lot of chip business out there that is marginal at best that good companies with alternatives should avoid. I hope multi-sourced smart phone modem isn't another console socket win for Intel.
Chipguy,
The question is how badly Intel wanted to get a foot in the mobile door at Apple. It makes no sense to lose money with a customer that will drop you in an instant to save a penny.
Flumoxed,
I don't believe these "models" any more than you do.
Flumoxed, having lived in the Pacific Northwest for six years, I am very familiar with the nine months of constant rain each year.
I have also spent my entire career writing computer simulations. I know all about the limits of simulation and what we can and cannot do in trying to predict whether a digital design will work coming out of the fab.
Hence when you say that "computer models show the Pacific Northwest will become a coastal desert," I will naturally be skeptical of said computer models.
Because your claim is just as extreme as the claim that Arizona will soon become beachfront property, courtesy of rising sea levels due to "climate change."
Tenchu
Elmer, I forgot to mention the role emulation plays.
Somewhat of a cross between simulation and real silicon, emulation is where we program the design onto FPGAs and run it like an actual chip.
The thing about emulation is that no single FPGA can support a complex design like a CPU or even a functional cluster of said CPU. Instead, there is a network of machines that house these FPGAs. The design block that is to be emulated will somehow be programmed onto these FPGA host machines. Then tests will be run.
The nice thing about emulation is that we can get many more cycles through than RTL simulation. The problem, of course, is that you lose the checking functionality of RTL test benches. Simulations are nice in that if something goes wrong, you can rerun the test and create waveforms of any signals in the design. You can't do that with emulation, just like you can't do that with actual silicon.
It just reinforces the absurdity of putting any faith in climate models from a science stand point and it sure explains them from a political standpoint.
Flumoxed,
But, I have negative faith in a country run by oligarchs who rape and pillage the environment (including the atmosphere) for their own financial gain, and who have bought the government to permit them to do it.
Elmer, there are many different levels of simulation. When I was working on server CPU design verification, I would run uncore models where we either run the entire uncore or just various portions of it.
We also run tests where a single CPU core is modeled along with the uncore. The other CPUs are replaced by behavioral (i.e. dummy) models because the simulation performance is already quite slow. I don't know if we ever modeled the entire chip with all of the CPU cores instantiated with the uncore.
Of course all of these are RTL simulations. I don't know the extent of the models used to run fault grading, gate-level simulations, or whatever. I do know that certain blocks are also run on analog mixed-signal models, which may have not been around back during your time. These models are very slow but are crucial in bridging the gap between the analog and digital world.
As for the machines we run these simulations on, they're standard dual-CPU rack-mounted servers. At Intel, there are of course tens of thousands of these machines allocated across different projects based on schedule and need. I don't think RTL simulations as of yet are able to take advantage of more than one or two CPU cores at a time, since the multithreading in these simulations is still not where you would expect them to be. Hence we just make up for it by running at least several jobs per machine at a time, limited more by local memory than CPU cores.
Now let me answer your questions directly:
Is that for a single CPU or a multicore device?
Is that 15 minutes of RTL simulation or gatelevel simulation with parametrics?
How long do such simulations take of wall clock time and how many System CPUs are involved in running those simulations?
(Climate models) need to not be accurate. Otherwise who would get a grant to make a better one?
Elmer,
We can't simulate more than a few seconds of CPU operation and you put faith in climate models that project out decades/centuries?
Flumoxed, maybe I chose the wrong word. I should have said "Disheartening."
Tenchu
Flumoxed,
He isn't focused on the main problem...the survival of Intel.
Koog,
Stated simply, Tench, Krzanich seems to have absolutely no strategic vision. He seems enamored with buzzwords (IoT) and toys (drones) without the vision of how any of these things can be made into major businesses. He seems intent on purchasing or investing in small companies without a thought on how they can be integrated into a strategic thrust.
Koog,
Krzanich is not CEO material.
Muzosi,
so that's the delusion of a conservative these days, eh?
Ibc,
Major WTF here...
Chipguy,
Charlie is NEVER wrong (just ask him).
What's the good part, Borusa? The part where Jerry Sanders rides in on his white horse and saves humanity from the evil Intel monopoly?
The part where Athlon and Opteron finally reach the 30% critical market share that they were denied before of Intel's "targeted rebates"?
The part where the market for x86 processors is finally leveled and consumers are free to choose between Intel and AMD?
You're stuck in the past, waiting for something that no longer applies in the ever-changing world of high-tech.
Tenchu
Borusa, the real question is why you are stuck focusing on "targeted subsidies" when that has nothing to do with why Intel failed to make inroads into handheld devices.
It's 2016, not 2006. Time to reboot your system and install the necessary updates.
Tenchu
Andy,
Without server growth, Intel is a $24 stock......at best.
Borusa, you're stuck in a feedback loop. Someone needs to hit your reset button. EOM
Ibc,
Apple has a CPU team that is seasoned - they have engineers from DEC, IBM and Intel among others.
I guarantee you that those people don't migrate between teams