Archive

Posts Tagged ‘intel’

Intel Curie Module, Arduino 101 Board Are Being Discontinued (Too)

July 26th, 2017 No comments

Intel seems to have completely given up on its efforts to bring products specific to the Internet of Things. After discontinuing Intel Edison, Galileo and Joule boards & modules last month, forcing companies to look for alternatives, the company has now issued product discontinuance / end-of-life notices for Intel Curie Module and Arduino 101 board, itself based on the Curie module.

The two product change notification notice can be found below for:

  • Select Intel Curie Module Products – PDF
  • Select Intel Arduino 101 Products – PDF

The use of the word “Select” would normally mean some versions of the module and board won’t be affected, but I fail to see which ones here, as AFAIK there’s only one Arduino 101 board, and two variants of Curie modules, all three to be discontinued.

Arduino 101 will be phased out faster with the following milestones:

  • July 17, 2017 – Product Discontinuance Program Support Begins
  • September 17, 2017 – Last Product Discontinuance Order Date
  • September 17, 2017 – Orders are Non-Cancelable and Non-Returnable
  • December 17, 2017 – Last Product Discontinuance Shipment Date

However, you’ll still have until July 17, 2018 to get Intel Curie modules:

  • July 17, 2017 – Product Discontinuance Program Support Begins
  • September 15, 2017 – Product Discontinuance Demand To Local Intel Representative
  • January 17, 2018 – Last Product Discontinuance Order Date
  • January 17, 2018 – Orders are Non-Cancelable and Non-Returnable
  • July 17, 2018 – Last Product Discontinuance Shipment Date

Intel also posted about the news in a forum post, where they mentioned searching for alternative manufacturers for Arduino 101 board, that is until Curie is completely gone, and documentation and tools will be available until June 15, 2020, and open source code will still be found in Github.

Intel’s IoT firing squad will obviously cause collateral damages, as any company who worked on products with Curie module may have to change plans. Affected products include UDOO X86 board, Gumstix Radium 96BIE board, and Xiaomi smart running shoes. At least Intel allows companies around one year to receive enough modules to possibly continue selling their products until 2019 or 2020. It will all depend on how much stock they are ready to order before January 17, 2018.

Via Hackaday and Golem

Intel’s Movidius Neural Compute Stick Brings Low Power Deep Learning & Artificial Intelligence Offline

July 21st, 2017 6 comments

Intel has released several Compute Stick over the years which can be used as tiny Windows or Linux computer connected to the HDMI port of your TV or monitor, but Movidius Neural Computer Stick is a complete different beast, as it’s a deep learning inference kit and self-contained artificial intelligence (A.I.) accelerator that connects to the USB port of computers or laptops.

Intel did not provide the full hardware specifications for the kit, but we do know the following specifications:

  • Vision Processing Unit – Intel Movidius Myriad 2 VPU with 12 VLIW 128-bit vector SHAVE processors @ 600 MHz optimized for machine vision, Configurable hardware accelerators for image and vision processing; 28nm HPC process node; up to 100 gigaflops
  • USB 3.0 type A port
  • Power Consumption – Low power, the SoC has a 1W power profile
  • Dimensions – 72.5mm x 27mm x 14mm

You can enter a trained Caffe, feed-forward Convolutional Neural Network (CNN) into the toolkit, profile it, then compile a tuned version ready for embedded deployment using Intel/Movidius Neural Compute Platform API. Inference occurs in real-time in the stick itself, and no cloud connection is needed. You can even connect multiple Movidius Compute Sticks to the same computer to scale performance.

It can help bring artificial intelligence to drones, robots, security camera, smart speakers, and anything that can leverage deep learning. The video below also shows the USB Compute Stick connected to what looks like a development board, so the target platform does not need to be powerful with most of the hard processing going inside in the stick. It currently does need to be an x86-64 computer running Ubuntu 16.04, so no ARM support.

Movidius Neural Compute Stick is sold for $79 via RS components and Mouser. You’ll find the purchase links, getting started guide and support forums on Movidius Developer site.

Supermicro A2SDi-2C-HLN4F Server Mini-ITX Motherboard is Based on Intel Atom C3338 “Denverton” Processor

July 5th, 2017 15 comments

Almost exactly one year ago, I wrote about Intel Atom C3000 Series “Denverton”  processor based on the same Goldmont architecture used in Apollo Lake processor, and used in an early GIGABYTE MA10-ST0 server motherboard prototype. The latter is not out yet, but according to a YouTube video, the 16-core motherboard should be launched in H2 2017. In the meantime, Supermicro also made their own Denverton motherboard based on a lower-end Atom C3338 dual core processor.

Supermicro A2SDi-2C-HLN4F board specifications:

  • Processor – Intel Atom C3338 dual core Denverton processor @ 1.50 / 2.20 GHz with 4 MB  cache; 9W TDP
  • System Memory
    • Up to 64GB Register DIMM RDIMM and DDR4-1866MHz
    • Up to 32GB Unbuffered ECC/non-ECC UDIMM, DDR4-1866MHz, in 2 DIMM slots
  • Storage – 4x SATA 3 ports, but up to 8x SATA 3 (6 Gbps) ports are configurable; N.B.: total combined PCI-E lanes and SATA ports is up to 8
  • Connectivity – 4x GbE Ethernet via Intel C3000 SoC
  • USB – 4x USB 2.0 ports including 2x via headers, 1x USB 3.0 port
  • Expansion – 1x PCI-E 3.0 up to x4 (in x4 slot); number of PCI-E lane is configurable via BIOS setup: 0, 2, or 4.
  • Video Output – VGA
  • Serial – 1x COM port via header
  • IPMI & 2D Graphics – ASPEED AST2400 BMC
  • Misc – TPM header; AMI UEFI 2.4/SMBIOS 2.7.1 “BIOS”;  HW monitoring: 4-fan status and tachometer, CPU voltage monitoring, system temperature, VBAT, etc…
  • Power Supply – ATX power connector; 12V DC power input; CPU thermal trip support for processor protection; power-on mode for AC power recovery
  • Dimensions –  17.02cm x 17.02cm (Mini-ITX)
  • Temperature Range – Operating: 0°C ~ 60°C; storage: -40°C – 85°C

The board supports Windows 64-bit operating systems like Windows 10, Windows Server 2016, Hyper-V Server 2016, etc.. as well as various 64-bit Linux distributions such as RedHat Linux EL 7.3, Fedora 25, SuSE SLES 11 SP4, Ubuntu 16.04 LTS and others, as well as FreeBSD. You may find the full list here.

If you’d like a complete system, the company also offer SuperServer 5029A-2TN4 with the board, and the company’s SC721TQ-250B mini tower with four hot-swap 3.5″ SATA drive bays, two internal 2.5″ drive bays, and 1 slim DVD-ROM drive bay shared with one internal 2.5″ bay.

Supermicro A2SDi-2C-HLN4F motherboard is sold for just above $200 on websites such as Atacom and WiredZone. I found the barebone “SuperServer” sold for $405 on Newegg. The latter has been reviewed on ServeTheHome in April with the conclusion rather positive:

If you are looking for a 1GbE NAS unit, this is a great option. The performance is well above the previous generation dual core and sometimes quad core models. The ability to maintain a low power profile will help ROI.

Over time, we expect to see better support for the X553 NIC. The major competition for this unit will be when the higher core count variants come out and iterations that support 10GbE.

In many environments, this is going to be a perfect NAS/ network services platform. If you wanted a simple 1GbE Linux NAS or caching appliance, the Supermicro SYS-5029A-2TN4 should be high on your list.

They also ran several benchmarks to put the CPU performance into perspective.

OpenSSL Sign Benchmark – Higher is Better

You’ll find more details about the motherboard on the product page.

Thanks to Paul for the tip.

 

 

Intel Quark S1000 “Sue Creek” Processor to Support On-Chip Speech Recognition

June 19th, 2017 3 comments

Intel may have announced plans to discontinue several of their IoT boards, but based on some documents I received, the company has not given up on the Quark family, although they may have given up on the Intel architecture for low power microprocessor, as Intel Quark S1000 – codenamed “Sue Creek” – will feature two Tensilica LX6 cores (yes, just like ESP32), and is designed to handle speech recognition at the edge (e.g. locally), so some of your voice commands should still work when Internet is down.

Intel Quark S1000 key features and specifications:

  • Digital Signal Processors
    • Dual Tensilica LX6 cores @ 400 MHz with HiFi3 DSP
    • Single precision scalar floating-point instructions
    • 16KB 4-way I$; 48KB 4-way D$
    • Up to 2400 DMIPS, 3.2 GMACS (16×16), 800 MFLOPS of Compute
  • Speech Accelerators
    • A GMM (Gaussian Mixture Model) and neural network accelerator
    • Low power keyboard and limited vocabulary recognition
    • Up to 9.6 GMACS (16×16) of compute
  • Internal Memory
    • 4MB shared embedded SRAM
    • 64KB embedded SRAM for streaming samples in low power mode
  • External Memory Interfaces
    • Up to 8MB external 16-bit PSRAM
    • Up to 128MB external SPI flash
  •  I/O Interfaces
    • Host I/O – SPI for command and control, I2S for streaming audio, IRQ, reset, wake, optional USB 2.0 HS device
    • Microphone – I2S/TDM 9.6 MHz max. bit clock
    • Digital Microphone – 4 PDM ports 4.8 MHz max. bit clock
    • Speaker – I2S/TDM 9.6 MHz max. bit clock
    • Instrumentation – I2C master @ 100/400 MHz
    • Debug – UART Tx/Rx/RTS?CTS up to 2.4 Mbaud/s
    • GPIO – 10 mA sink/source, 8x PWM outputs
  • Power Management / Consumption
    • Low power idle (memory retention); voice activity detection; play through; full active
    • Clock and power gating support
    • < 20 mW voice activity detection
    • < 250 mW full active
  • Package (preliminary) – FCCSP132 7.45 x 8.3mm 0.6/0.7mm pitch staggered/orthogonal
  • Temperature Range – Commercial: 0 to 70 °C; industrial: -40 to +85 °C

The diagram above shows Quark S100 is supposed to be connected to a host processor providing network connectivity, getting commands over SPI, audio over I2S, and the Intel processor can handle some speech recognition likely for a limited subset of words, and use cloud based recognition for more complex requests. The solution could be used in product like Google Home or Amazon Echo look-a-likes, or other voice-controlled appliances.

I don’t know when the processor will be available, and I could not find any information online yet.

Intel Issues End-of-Life Notices for Galileo / Galileo 2, Edison and Joule Boards & Modules

June 19th, 2017 12 comments

While I’m not sure many of my readers are using them, Intel introduced several IoT development kits and modules over the years, with products like Intel Galileo, followed by Galileo 2, Edison module development board all based on Quark processors, and more recently Intel Joule modules powered by Intel Atom T550x / T570x processors.

The three boards / modules and corresponding modules will soon be no more, as Intel issues three end-of-life (EOL) notices for:

  • Intel Galileo Board, and Intel Galileo Gen2 Board Products – PDF
  • Select Intel Edison Compute Module, Intel Edison Breakout Board, Intel Edison Kit for Arduino, and Intel Edison Breakout Board Kit Products – PDF
  • Intel Joule 570x Compute Module, Intel Joule 550x Compute Module, Intel Joule 570x Developer Kit and, Intel Joule 550x Developer Kit Products – PDF

All three follow the same “forecasted key milestones”:

  • June 16, 2017 – Product Discontinuance Program Support Begins
  • July 16, 2017 – Product Discontinuance Demand To Local Intel Representative
  • September 16, 2017 – Last Product Discontinuance Order Date
  • September 16, 2017 – Orders are Non-Cancelable and Non-Returnable After
  • December 16, 2017 – Last Product Discontinuance Shipment Date

EOL notices are normal, and potentially understandable for Galileo and Edison products announced in 2013 and 2014, but Intel Joule modules were just announced last August, so they may not have been used any products considering development time, and I can’t even find Atom T5500/550X or T5700/570x on Intel Ark website. So it does not look good for Intel’s IoT initiatives. Note that some Edison modules will be sold in 2018, so not all SKUs are discontinued.

Via Hardware.info. Thanks to Sander for the tip.

Intel DLIA is a PCIe Card Powered by Aria 10 FPGA for Deep Learning Applications

May 29th, 2017 No comments

Intel has just launched their DLIA (Deep Learning Inference Accelerator) PCIe card powered by Intel Aria 10 FPGA, aiming at accelerating CNN (convolutional neural network) workloads such as image recognition and more, and lowering power consumption.

Some of Intel DLIA hardware specifications:

  • FPGA – Intel (previously Altera) Aria 10 FPGA @ 275 MHz delivering up to 1.5 TFLOPS
  • System Memory – 2 banks 4G 64-bit DDR4
  • PCIe – Gen3 x16 host interface; x8 electrical; x16 power & mechanical
  • Form Factor – Full-length, full-height, single wide PCIe card
  • Operating Temperature – 0 to 85 °C
  • TDP – 50-75Watts hence the two cooling fans

The card is supported in CentOS 7.2, and relies on Intel Caffe framework, Math Kernel library for Deep Neural Networks (MKL-DNN), and works with various network topologies (AlexNet, GoogleNet, CaffeNet, LeNet, VGG-16, SqueezeNet…). The FPGA is pre-programmed with Intel Deep Learning Accelerator IP (DLA IP).

Intel DLIA can be used by cloud services providers to filter content, track product photos, for surveillance and security applications for example for face recognition and license plate detection, in the factory to detect defects automatically, and in retail stores to track foot traffic, and monitor inventory.

You’ll find more details including links to get started and the SDK in the product page.

Intel Compute Card Apollo Lake and Kaby Lake SKUs, Block Diagrams, and Specifications

May 29th, 2017 2 comments

Intel Compute cards are the latest ultra-thin CPU cards introduced by Intel at the beginning of the year, with a concept similar to EOMA68 CPU cards, that it to allow  CPU card upgrades or replacements, and interoperability across compatible devices such as smart kiosks, IoT gateways, and so on. But at the time, Intel did not reveal that many details about the different cards, although we know NexDock is working on laptop dock compatible with Intel Compute Cards. But I’ve recently received some block diagrams for Apollo Lake Pentium/Celeron, and Kaby Lake Core M/Core i5 compute cards.There will two SKUs for Apollo Lake compute cards sharing the same specifications, except for the processor:

  • SoC
    • CD1C64GK SKU – Intel Celeron N3450 quad core processor @ 1.1 / 2.2 GHz (base/turbo) with 12EU Intel HD Graphics Gen9; 7.5W TDP
    • CD1P64GK SKU – Intel Pentium N4200 quad core processor @ 1.1 / 2.5 GHz with 18 EU Intel Gen9 HD graphics; 6W TDP
  • System Memory – 4GB dual channel LPDDR3-1866
  • Storage – 64 eMMC flash, SPI flash for BIOS
  • Connectivity – Intel Wireless-AC 7265 (2×2 802.11ac & Bluetooth 4.2)
  • Compute Card connector:
    • USB type C part with  USB 3.x, USB CC (Configuration Channel), DisplayPort 1.2 and USB 2.0 signals
    • Extended part with USB 2.0, DisplayPort 1.2, USB 3.x, 2x multiplexed SATA & PCIe x1 interfaces
  • Others – PMIC, Embedded Controller, and Crypto Element Device

There will also be two more powerful and expensive Kabe Lake compute card with the following specifications:

  • SoC
    • CD1M3128MK SKU – Intel Core m3-7Y30 dual core / quad thread processor @ 1.0 / 2.6 GHz (base/turbo) with Intel HD Graphics 615; 4.5W TDP
    • CD1IV128MK SKU – Intel Core i5-7Y57 dual core / quad thread processor @ 1.2 / 3.3 GHz with Intel HD Graphics 615; 4.5W TDP; Support Intel vPro
  • System Memory – 4GB dual channel LPDDR3-1866
  • Storage – 128GB PCIe SSD, SPI flash for BIOS
  • Connectivity – Intel Wireless-AC 8265 (2×2 802.11ac & Bluetooth 4.2)
  • Compute Card connector:
    • USB type C part with  USB 3.x, USB CC (Configuration Channel), DisplayPort 1.2 and USB 2.0 signals
    • Extended part with USB 2.0, DisplayPort 1.2, USB 3.x,  1x multiplex Gigabit Ethernet / PCIe x1, and 1x PCIe x1
  • Others – Voltage regulators (VRs), Embedded Controller, and Crypto Element Device; CD1IV128MK only: TPM

I was not aware of any Core i5 processors with such a low TDP, which can also be tuned up to 7W, and down to just 3.5W. The processor was just launched in January, a few devices are equipped with the processor, but I still managed to find Lenovo ThinkPad X1 Tablet 20JB ( $1,450), as well as some benchmarks for reference.

Intel To Make Thunderbolt 3 Royalty-free, Release Specifications

May 25th, 2017 10 comments

Intel Thunderbolt 3 is a single port connect that supports multiple standards (Thunderbolt, USB 3.1, DisplayPort, PCIe), leverages USB-C connector, and offers up to 40 Gbps throughput. So far, it’s mostly found in higher end computers and laptops, but Intel has now announced plans to make it royalty-free, and “release the  specifications to the industry” (so maybe not completely free/public) in order to increase adoption of the standard.

Thunderbolt-3 main features:

  • Thunderbolt, USB, DisplayPort, and power on USB-C
  • USB-C connector and cables (small, reversible)
  • 40 Gbps Thunderbolt 3 – double the speed of Thunderbolt 2
    • Bi-directional, dual-protocol (PCI Express and DisplayPort)
    • 4 lanes of PCI Express Gen 3
    • 8 lanes of DisplayPort 1.2 (HBR2 and MST)
      • Supports two 4K displays (4096 x 2160 30bpp @ 60 Hz)
  • USB 3.1 (10 Gbps) – compatible with existing USB devices and cables
  • DisplayPort 1.2 – compatible with existing DisplayPort displays, devices, and cables
    • Connect DVI, HDMI, and VGA displays via adapters
  • Power (based on USB power delivery)
    • Up to 100W system charging
    • 15W to bus-powered devices
  • Thunderbolt Networking
    • 10Gb Ethernet connection between computers
    • Daisy chaining (up to six devices)
    • Lowest latency for PCI Express audio recording

That means eventually we may get devices with a single Thunderbolt/USB-C like smartphones to connect all peripherals including 4K displays,  NVMe SSDs, Ethernet, external GPU, etc.., and power/charge the device though a Thunderbolt 3 dock, which can be purchased for $200 and up. One example is StarTech.com Thunderbolt 3 Docking Station with 4K display, Ethernet, Audio and USB ports going for $199.99 on Amazon US.

Intel also announced plans to integrate Thunderbolt 3 into future Intel CPUs.

Categories: Hardware Tags: displayport, intel, thunderbolt, usb