Intel’s Haswell processors drive into embedded

Paul Boughton

Intel’s fourth generation Core processors, codenamed ‘Haswell’ have been launched in formats for the embedded market.

congatec in Germany has launched a Type 6 COM Express compact module supporting the low-power single-chip solution, codenamed Haswell-ULT, with integrated chipset (PCH) and graphics. Despite increased performance, the maximum thermal design power (TDP) is just 15 watts.

These processors are an optimisation of the existing microarchitecture with new microcode and extended registers, an expansion of the vector processing unit, larger and much more powerful graphics units, as well as standard hardware support for AES encryption in all models.

The Core i7-4650U and i5-4300U have improved power management as well as expanded, individually configurable turbo-boost modes and extensive TDP configuration management for adapting to the chosen cooling solution. Models of the Haswell U-Series integrate the processor controller hub (PCH) as multi-die in the same package.

Compared to previous Core i processor models, such as the i7-3517UE, it is possible to save not only 2 watts of TDP, but also approximately 4 watts previously required for the separate PCH solution. Further savings come from the new voltage regulator units which are also integrated in the housing. In total, power consumption is about a third lower compared to the previous generation.

The conga-TC87 can currently be equipped with the embedded dual-core single-chip processors Intel Core i7-4650U, i3-i5-4300U or i3-4010U. For each of these processors, the appropriate chipset is integrated in the housing. The module supports fast and energy-efficient dual-channel DDR3L memory up to 16GB.

The integrated graphics is considerably more powerful than preceding versions and supports Intel flexible display interface (FDI), DirectX 11.1, OpenGL 4, OpenCL 1.2 and high-performance, flexible hardware decoding to decode multiple high-resolution full HD videos in parallel. 4K resolution with up to 3840 x 2160 pixels for DisplayPort is natively supported. It is also possible to connect up to three independent display interfaces via HDMI, LVDS and embedded DisplayPort (eDP). When using DisplayPort, the individual displays can be daisy chained to take advantage of simple wiring.

Native USB 3.0 support provides fast data transmission with low power consumption. A total of eight USB ports are provided, two of them support USB 3.0 Superspeed. Four PCI Express 2.0 lanes, four SATA ports with up to 6 Gb/s, RAID support and a Gigabit Ethernet interface enable fast and flexible system extensions. Active fan control, LPC bus for easy integration of legacy I/O interfaces, I²C bus and Intel High Definition audio complete the feature set.

The strength of this new COM Express compact module lies in the flexibility and scalability of the graphics and processing power; and superior performance within a maximum TDP power envelope of 15 watts. All of this makes the conga-TC87 an ideal solution for high-performance mobile systems.
MSC Vertriebs in Germany also offers its high-end MSC C6B-8S COM Express Type 6 module family with the Haswell processors for higher performance applications.

Besides the first module with Intel Core i7-4700EQ (2.4GHz, 3.4GHz in turbo boost mode) quad-core processor, four more economical Intel Core i3 and i5 variants with two processor cores are now available. The COM Express products are offered with the Core i5-4400E (2.7/3.3GHz), i5-4402E (1.6/2.7GHz), i3-4100E (2.4GHz) or i3-4102E (1.6GHz) processor. The thermal design power (TDP) is 37W or 25W. Depending on the type, the processors support the Intel AMT 9.0 Technology, Intel 64, the Intel Virtualization Technology, VT-d Virtualized I/O, Intel‘s Trusted Execution Technology, the Intel Advanced Encryption Standard and the Intel Turbo Boost Technology. The Intel Advanced Vector Extensions (AVX) 2.0 Technology allows for high-end imaging applications.

The 4600 (GT2) Intel HD graphics integrated into the processor die offers a significantly improved video and graphics performance over the third generation. Accelerated coding and decoding functions for high resolution videos, DirectX 11.1 and OpenGL 3.2 are supported. OpenCL 1.2 allows for the additional use of the graphics engines for applications with extensive use of floating-point computations.

The MSC C6B-8S module family also integrates the improved QM87 Intel 8 series platform controller hub (PCH). Two fast dual-channel DDR3L (1333/1600) SDRAM modules with a maximum storage capacity of altogether 16 GB assure a great computing performance and a low power consumption. An Infineon Trusted Platform Module offers a hardware based security functionality according to the Trusted Computing Group (TCG).

The MSC C6B-8S modules have comprehensive display support offering three DisplayPort 1.2 with multi-stream-transport or HDMI digital display interfaces with resolutions of up to 3800 x 2400 pixels to connect three independent displays. Furthermore the modules provide two embedded DisplayPort interfaces, along with a LVDS and CRT interface. The processor modules feature four fast USB 3.0 and four USB 2.0 ports, seven PCI Express x1 channels, a PCI Express graphics (PEG) x 16 interface, a LPC bus, Gbit Ethernet, HD audio, and four SATA interfaces at up to 6Gb/s.

The high-performance platform runs under the Microsoft Windows 7, Windows 8 and Linux operating systems. The BIOS firmware is based on the UEFI platform Aptio from AMI and MSC also offers starter kits and carrier boards to ensure a quick evaluation of the COM Express modules and corresponding cooling solutions like a passive and an active heat sink.

The MSC C6B-8S module family is especially suitable for sophisticated applications such as image recognition and processing, for displaying multiple high-resolution (up to 4k) videos or for the control of large screen displays.

For embedded military applications such as image recognition and radar, GE Intelligent Platforms has developed a rugged dual node quad core 6U OpenVPX High Performance Embedded Computing (HPEC) deployed server platform that uses the Intel Core i7 Haswell chip sets.

Designed for a broad range of mission-critical intelligence, surveillance and reconnaissance (ISR) and image-, signal- and data processing applications, the DSP281 brings more than double the theoretical peak performance of its predecessor within the same size, weight and power (SWaP) envelope.

Deployed on manned- and un-manned airborne, ground and naval platforms, it can enable optimum benefit to be derived from the significant volume of data obtained from large numbers of high resolution sensors by providing meaningful information in an actionable time frame to decision makers and war fighters across an asymmetric, dynamic and rapidly-changing threat landscape.

The DSP281 is a software-compatible upgrade to the DSP280, which was based on a dual Intel Core-i7 processor architecture, and provides defence and aerospace prime contractors, systems integrators, OEMs and end users with a clear, cost-effective path to increased mission capability through planned technology refresh across a wide range of platforms. It also brings high levels of data security with GE’S new on-board Security Hub and full support for the latest Intel platform security features.

“Since the GE DSP280 platform was announced two years ago, the nature and complexity of the threats faced by military organizations around the world has continued to evolve at a rapid rate,” said Michael Stern, Global Product Manager for Military & Aerospace High Performance Embedded Computing at GE Intelligent Platforms. “The DSP281 responds to this in two key ways. Not only does it offer significantly enhanced processing performance, but it also takes advantage of the increasing availability of software tools, expertise and experience coming from commercial high performance computing (HPC) data centres. This makes the DSP281 not just a very powerful solution, but also one that minimises risk, cost and time-to-market.”

GE’s HPEC architecture, as implemented in the DSP281, can scale from one to many processor nodes per enclosure via RDMA-enabled InfinBand and Ethernet dual port network interface cards, delivering very high interprocessor bandwidth at extremely low memory to memory latencies. In addition, system integrators can minimize card count by mapping multiple platform functions such as control, DSP, image and video processing and graphics onto one or more DSP281s in order to reduce spares and simplify logistical support in the field.

The DSP281 is supported by GE’s AXIS Advanced Multiprocessor Integrated Software development environment. AXISPro includes a high performance IPC middleware and GUI for task level programming and fast prototyping to reduce development cost and shorten time to solution of multi-core, multi-threaded and multi-node distributed system architectures across multiple operating systems including Linux, Windows and VxWorks and includes quick start signal and image processing examples to further accelerate time-to-solution.
 
For more information, visit www.congatec.com;www.mscembedded.com;
defense.ge-ip.com/products/dsp281/p3719