Meet the little startup that Intel just bought so they can say they understand AI.              

Well, it can transfer or "carry over" some machine learned "AI Vision" executions, so that is an improvement of sorts, but not enough improvement to make Intel x86 chips stay relevant for very long.

https://www.youtube.com/watch?v=VioTPaYcF98

And here is a more complete picture of Movidius main use as a "AI vision program transfer stick" or a jump drive for AI vision programs, carrying the new learned item between laptop and your various toys, little hover drones and other gimmicks.

https://www.youtube.com/watch?v=4xud1T9DaFY

Actually, x86 main PC functions has nothing much to do with what Movidius is doing right now in AI vision with it running through a USB cable like that.   Nor vice versa.   It is a "try me" gimmick, to let programmers try it out some on some AI vision programming toys before buying big new equipment to do it right.  

Like the Optane memory, the reality does not live up to the huge early on hype Intel puts out on it.

It would take the old Wintel working together diligently to make that Movidius vision related stuff work on PC .......   it would have to move to the silicon heart of the CPU/GPU set up and be integral to the OS and have all the apps rewritten to utilize it properly.    And this has yet to start in Wintel world.

..... and currently Tel is busy pointing racked and ready legal shotguns at Win's face, Win who is busy making plans to run off with Qualcomm and the ARM boys in the very near future.

                                                       

      ...... BTW,  do you think those little blue heat sheild things with the huge air vent slots in them might mean Movidius gets a wee bit HOT when it runs by any chance?



========================================



Intel discontinues WiGig hardware





Remember this ????   Intel swore up and down it was going to replace all cables and actually peddled it to some machine builders who paid big bucks to include it in their machines, but it never got any presence in the "other equipment" side of the equation, so it never got used anywhere except in this particular picture.

Intel has been pushing the idea of wireless homes and offices for years. Bring your laptop into your workspace and it’ll automatically connect to your display, printer, mouse, and other hardware without the need to plug anything in. And wireless charging means you don’t even need to plug in a power cable.

But now it looks like Intel is giving up at least part of that vision. The company is discontinuing its WiGig (Wireless Gigabit) products for laptops and other mobile devices.

As spotted by AnandTech, Intel is phasing out its existing WiGig products and telling hardware partners to place any final orders by September 29th. The last shipments will go out by December 29th, 2017.



What this really means is that the current BlueTooth 5.1 tech has this tech beat all hollow as BT 5.0 and up all has 100+ foot range and IT GOES THROUGH WALLS, it has real devices all over the place using BlueTooth tech and it costs next to nothing to use the proven BlueTooth 5.1 tech compared to this overly expensive Intel WiGig which cannot go through a single layer of drywall.

Once again, Intel tech loses out to a nobody group pushing a simpler, always improving FOSS based technology.   This ever improving cheap to free competition is why ALL of Intel's much hooted "Technical Innovations" die out within 3-5 years time and the high cost of using Intel tech explains why it never goes anywhere during the short time while it is current.

   Intel has a history of big big claims and no real performance, a history which is repeating itself all the time lately.

http://www.zdnet.com/article/crossbar-aims-to-bury-intels-3d-xpoint-with-supe...

Crossbar aims to bury Intel's 3D XPoint with superior technology



INTEL GROWS A MORE CAPABLE COMPETITION

Enter Crossbar. Founded in 2010, Crossbar has designed a filimentary ReRAM with excellent performance, scalability, and manufacturability.

Crossbar says writes are 1,000x faster than NAND flash, at 1/20th the power consumption, and with over 1,000x the endurance of flash. Not quite as fast or durable as DRAM, but surprisingly close - and much more power effficient than flash or DRAM - perfect for mobile devices and IoT. And it performs well over a wide temperature range, from -40C to 125C.

Manufacturability is critical to achieving volume, and here Crossbar has an excellent story as well. They can use a standard CMOS fab, adding a couple of steps at the tail end of the process, using existing tools and processes, and they have an easy 3D stacking design to achieve high density.

But 3D isn't their only path to density. They've also demonstrated that their design can shrink to feature sizes of less than 8nm, while also increasing the ratio between ON states and OFF states. They could produce 1TB chips in a few years, given smaller feature sizes and enough layers.

NAND flash, on the other hand, had to go the 3D route because as features sizes shrink, endurance and stability decrease. Stacking flash cells is the only way that vendors could increase density.

3D XPoint's clumsy start has been a blessing to NVRAM companies like Crossbar and Nantero. The slippage in schedule and specs has given them breathing room, while the software folks - Microsoft in particular - responded with an urgency that no startup could command.

With Intel 3D XPoint partner Micron strangely silent on their plans, there may be more potholes on the 3D XPoint roadmap than we've heard about. But the good news for us consumers of fine storage technology is that it looks like we'll have more choices for NVRAM than expected.

OK, it's got an industry standard to back it now and a real name, NVRAM.    It can run on existing NORMAL planar style production process machines and it can support lithography at less than 10nm right now.    (bit mining card graphical SOCs and their support memory are being run off at 7nm as we speak)

Micron isn't fighting this new stuff to any large degree (and is likely to swing over to it as the costs are superior to Intel's super expensive 3-D X-Point mess).

In addition, Samsung is making their current N-RAM products better and cheaper to the point you can use it for an entire machine's OS and data drive.   This isn't the final answer by any means, but it is a good interim step to where memory needs to go inside the next 2 years.

Intel's INTENTIONALLY locking down their new main CPU processors and the motherboards that mount them in an attempt to FORCE people to use slower and more expensive Intel 3-D X-Point daughter board memory is pure suicide for Intel -- don't buy one of these machines as you are paying big big bucks for Intel's latest round of stubborn bad decisions.

--------------------------------------------------------------

Shout from Sept 17th -- Micron has joined OpenCAPI and as such has signalled that the future of fast durable memory does not lie with Intel's Optane 3-D Xpoint memory at all ......  

So you would be stupid to buy you an Intel processor that RESTRICTS you to only use Optane memory.      

As long as it was her choice all the way she will tend to stay happy with it.

https://www.economist.com/news/business/21717430-success-nvidia-and-its-new-c...




Time for Intel to be paranoid

Instead of making ASICS or FPGAs, Intel focused in recent years on making its CPU processors ever more powerful. Nobody expects conventional processors to lose their jobs anytime soon: every server needs them and countless applications have been written to run on them. Intel’s sales from the chips are still growing. Yet the quickening rise of accelerators appears to be bad news for the company, says Alan Priestley of Gartner, an IT consultancy. The more computing happens on them, the less is done on CPUs.

One answer is to catch up by making acquisitions. In 2015 Intel bought Altera, a maker of FPGAs, for a whopping $16.7bn. In August it paid more than $400m for Nervana, a three-year-old startup that is developing specialised AI systems ranging from software to chips. The firm says it sees specialised processors as an opportunity, not a threat. New computing workloads have often started out being handled on specialised processors, explains Diane Bryant, who runs Intel’s data-centre business, only to be “pulled into the CPU” later. Encryption, for instance, used to happen on separate semiconductors, but is now a simple instruction on the Intel CPUs which run almost all computers and servers globally. Keeping new types of workload, such as AI, on accelerators would mean extra cost and complexity.



As AMD has done already (and is doing more and more) Intel needs to blend AI graphics type processing into the same die as the main CPU, sharing a pool of fast memory at the die level.

Trouble is, Intel still believes in SEPARATE chips for separate jobs and lots & lots of daughter boards which is what killed Intel in mobile over the last 10 years.

Intel still cannot shrink their lithography any more and Intel still can't seem to make a complex SOC chipset (because it simply doesn't believe in them).    

Intel has bought lots of little companies over the years and has then carelessly run them into the ground by NOT incorporating the tech inside their main CPU processor designs.


=====================================================


https://www.cnbc.com/2017/07/10/jefferies-downgrades-intel-to-underperform.html

Intel has 'most to lose' from 'tectonic shift in computing,' Jefferies says in downgrading stock

Jefferies downgraded Intel from hold to underperform on Monday, saying the chipmaker has the "most to lose" in the "4th tectonic shift in computing."

Jefferies said it is downgrading Intel because "its Xeon/Xeon PHI platform is disadvantaged vs NVidia in emerging parallel workloads like deep neural networking."

Jefferies calls out several areas of concern for Intel, including Microsoft's Windows new support for ARM processors and the rapid 200 percent growth of Nvdia's data-center business year over year.

Nvidia has been one of the market's hottest stocks recently; SoftBank Group bought a $4 billion stake in Nvidia in May. Shares of Nvidia are up 28 percent this year.

In a separate note on the semiconductor sector, Jefferies says it sees a major "tectonic shift" in the industry that will favor parallel computing platforms already used by AMD, Nvidia, Cavium and Xilinx.

Earlier tectonic shifts noted by Jefferies included the mainframe era in the 1950s; the minicomputer era in the 1970s; the personal computer era in the 1980s and 1990s; the cellphone/server era in the 2000s and the parallel processing IoT era we're just now entering.

"NVIDIA was the first to recognize and successfully invest in a HW/SW platform (GPU/CUDA) targeted specifically at parallel processing applications, and our field checks suggest it is years ahead of its competition," Jefferies said, referring to Nvidia's strategy to take advantage of computing power from graphics processing units versus standard processors.

Jefferies reset its 2018 price target for Intel to $29 from $38. Shares of Intel were at $33.30 in Monday's premarket, down 1.7 percent.


The Economist and CNBC both agree, Intel is busy losing it ......

Here is a plain English read on the 10 times faster FOSS based OpenCAPI interface wars which are at the heart of the AI roll out for everything next year ......

https://www.theregister.co.uk/2016/10/14/opencapi_declaration_of_interconect_...



Why OpenCAPI is a declaration of interconnect fabric war

Any standard but Intel in another CPU-memory interconnect consortium
By Chris Mellor 14 Oct 2016 at 22:38 11

An OpenCAPI consortium has sprung into life, promising a new, open specification that can increase data center server performance up to 10x through the use of a new CPU-memory-IO adapter interconnect scheme – and it doesn't include Intel in its membership.

The Open Coherent Accelerator Processor Interface represents, the consortium says, a data-centric approach, and "provides an open, high-speed pathway for different types of technology – advanced memory, accelerators, networking and storage – to more tightly integrate their functions within servers."

It "puts the compute power closer to the data, removes inefficiencies in traditional system architectures to help eliminate system bottlenecks and can significantly improve server performance," according to the consortium.

The design specification is "built to minimize the complexity of high-performance accelerator design. Capable of 25Gbits per second data rate, OpenCAPI outperforms the current PCIe specification, which offers a maximum data transfer rate of 16Gbits per second."

The consortium's announcement release declares: "Many technology companies have developed innovative solutions that today's data center technology cannot effectively leverage due to limited legacy interfaces. New technologies such as storage class memory and accelerators to support emerging workloads do not fit well on existing interfaces and a closed, proprietary approach does not allow for full industry participation nor innovation."

OpenCAPI was founded by AMD, Google, IBM, Mellanox Technologies, and Micron. Other members include Dell EMC, HPE, Nvidia and Xilinx.

Several members have product/tech introductions coming:

IBM plans to introduce Power9-based servers that leverage the OpenCAPI specification in the second half of 2017.
IBM will enable members of OpenPOWER Foundation to introduce OpenCAPI-enabled products in the second half 2017.
Google and Rackspace's new Zaius server under development, announced at the OpenPOWER Summit in San Jose, will leverage Power9 processor technology and plans to provide the OpenCAPI interface in its design.
Mellanox plans to enable the new specification capabilities in its future products.
Xilinx plans to support OpenCAPI-enabled FPGAs.
So it's mostly a Power9-driven effort currently.

The OpenCAPI consortium plans to make the OpenCAPI specification fully available to the public at no charge before the end of the year. Find out about this and membership details at opencapi.org.

Oh, and the supplier of the vast majority of server CPUs, Intel, is not a member. It's welcome to join though, the consortium says.

The background includes the point that Intel bought FGA supplier Altera for $16.7bn in December 2015, and has its own QPI (Quick Path Interconnect) technology, not to mention Silicon photonics.

It seems to us that OpenCAPI overlaps with the Gen-Z Consortium open memory fabric initiative. Both address the memory/storage-class memory interconnect area.

Gen-Z aims to provide a direct-attach, switched or fabric topology – a flexible, high-performance memory semantic fabric – to interconnect Compute (SoC memory), FPGA and GPU accelerators, pooled memory, and network/storage IO links.

OpenCAPI is meant to allow any microprocessor to attach to coherent user-level accelerators and I/O devices, and advanced memories accessible via read/write or user-level DMA semantics.

Then there is the CCIX (Cache Coherent Interconnect for Accelerators) group formed by AMD, ARM, Huawei, IBM, Mellanox, Qualcomm, and Xilinx (but not Nvidia or Intel) in May, to devise and make a cache coherent interconnect fabric linking multiple suppliers' CPUs, accelerators – FPGAs and GPUs – and network adapters so they can exchange data and share main memory.

Why are there three such overlapping efforts? The simple answer is that none of the members of these consortia want to let Intel own the CPU-memory-accelerator-network interface technology and so grab an even higher percentage of customer spend on servers, plus favor its own CPU and FPGA efforts.

They each and all think that the PCI interface is too slow and a higher-speed, lower-latency interconnect scheme is needed to let faster processors talk to faster accelerators, DRAM and the various storage-class memories emerging (think XPoint-type stuff), and RDMA-accessed external arrays.

Gen-Z consortium members are AMD, ARM, Broadcom, Cavium Inc, Cray, Dell EMC, HPE, Huawei, IBM, IDT, Lenovo, Mellanox, Micron, Microsemi, Red Hat, Samsung, Seagate, SK Hynix, Western Digital Corporation and Xilinx.

Google and Nvidia are the only OpenCAPI members not in the Gen-Z consortium.

Current Support for OpenCAPI

OpenCAPI declares: "Backed by a total of more than 30 leading technology companies, the three organizations now welcome each other's announcements as part of a collaborative industry effort to create an open data center architecture for the future."

As an example of vendor support, here's Tom Eby, VP and GM of Micron's compute and networking business: "Because open standards present the best opportunity for rapid innovation, the OpenCAPI, Gen-Z, and CCIX consortiums are an important step in ensuring that developing architectures can quickly adapt to capitalize on the dramatic benefits provided by new memory technologies."

Here's Gaurav Singh, CCIX Chair: "The CCIX group of companies believe in the need to foster innovation in the industry. We welcome the efforts of and look forward to collaborating with the Gen-Z and OpenCAPI consortiums, which will further the development of key technologies that will define the data centers of tomorrow."

Completing the fulsome threesome's joint supportive comments is Kurtis Bowman, Gen-Z Consortium president: "The Gen-Z Consortium is committed to establishing an ecosystem where members, the broader industry, and customers work together to deliver robust, high-quality specifications that enable new data center architectures. The formation of these three new consortia (CCIX, OpenCAPI, and Gen-Z), backed by more than 30 industry-leading global companies, supports the premise that the data center of the future will require open standards. We look forward to collaborating with CCIX and OpenCAPI as this new ecosystem takes shape."

Key customers for any non-proprietary alternative to Intel fabric interconnect technology are:
The server and hyperscaler suppliers: Cisco, Dell, Fujitsu, HDS, HPE, Huawei, IBM, InSpur, Lenovo, Oracle, Quanta and Supermicro on the server side.  Businesses like Amazon, Apple, Facebook, Google and Microsoft on the hyperscale data center side.   End users will basically buy what the server vendors adopt.

The three consortia, in a crude view, represent overlapping anti-Intel groups, wanting to constrain and limit chipzilla's ability to grab more of the server silicon component market to itself. It would be more hopeful for them if they could combine in a single CCIX, OpenCAPI, Gen-Z organization, but the individual members' concerns and technology preferences would make such an attempt unwieldy and difficult to bring to a successful conclusion. Hopefully, and nonetheless, they will try anyway.

....... yep, they just did this "combine the standards together" thing and then said they were going to actually going to build the resulting chipset and actually prove it out in test boards and products before selling it as part of the first wave of DynamIQ SOCs ......  (ARM's and Google's relatively cautious influence is showing through)

Intel would rather face, we're sure, three partly or ill-coordinated and overlapping groups that collectively confuse interconnect-buying customers, and contrast the trios's confusing messages and timescales with its own, hopefully clearer and simpler message.

The CPU-memory-network interconnect technology wars are on.    

Wuups, only ONE message and only ONE production proven FOSS chipset for Intel to face off against now ...... and wow, it's a lot cheaper and 10 times faster now too .......  poor poor Intel.

....... bet that Beeg One hurts a whole buncha of a lot going in --- and there is only what, 30 of them OpenCAPI dudes all lined up now jest awaiting for their turn at Intel's fundaments .......

                     

THAT'S the CPU?


The days of being satisfied with a yearly 20% speed bump from Intel or from ARM are over -- so is Moore's Law.  

Could you expand on that??
And what do you see in your crystal ball re: Moore's Suggestion?

I'm gonna be needing the Readers Digest version.

http://www.androidauthority.com/what-ever-happened-to-lgs-nuclun-processors-7...

Intel fails again in attempts to make a 10nm SOC for LG.



This is baseball rules and Intel has struck out three times already on making an SOC for LG off of their 10nm process.

Furthermore, Intel is also choking on building its own larger processors on their 10nm process as well, reporting yield issues and relatively poorer performance off of the ones that do test out as "all present on all functions".

LG is now planning to use the Snapdragon 835 and deeply regrets the year plus they spent screwing around with Intel.    Intel is back to playing with themselves over there in the corner.

 


Now, let's see graphically and clearly how non-competitive Intel's best 10nm LG effort was size-wise as compared to the very successful Snapdragon 835 as tuned and produced  by Samsung.




Snapdragon 821 is a little smaller than the LG SOC                        Snapdragon 835 is in the middle