The AMD 790FX Chipset Explored

As you may know, AMD has just launched their brand new PC platform solution, codenamed 'Spider'. Spider is really an ecosystem consisting of the AMD Phenom processor, AMD 790FX chipset and the ATI Radeon HD 3800 series graphics. AMD is banking on these new components to revitalize its position in the consumer PC market. However, it is quite unfortunate that the one component with the biggest hype and the most promise - the AMD Phenom processor - is the only product we're missing to complete AMD's trifecta. While we aren't really one to speculate, AMD's inability to provide seeding units is troubling indeed. So while we have graphics cards and motherboards stacking up, we're left twiddling our thumbs without so much as a whiff of a CPU to test on.




Still, moving on, let's just have a look at what the new AMD 790FX chipset has to offer. The main features of the 790FX chipset is really to upgrade the backbone support of AMD's platform in order to fully take advantage of the AMD Phenom processor and ATI Radeon HD 3800 series GPUs. These features come in the form of HyperTransport 3.0, split power planes and PCI Express 2.0. The AMD 790FX also officially supports DDR2-1066, but since the memory controller is embedded in the processor, this is more of a CPU upgrade than chipset.



With HyperTransport 3.0, the 790FX will effectively double the bandwidth available between the Northbridge and CPU from 1GHz of the previous generation chipsets running with HyperTranport 2.0 to 2GHz today. Split power planes is also a new feature supported by AMD Phenom (and above) processors where the there will be two power planes feeding the CPU, one for the CPU cores and another for the memory controller. This is a highly anticipated feature since AMD features a built-in memory controller. With separate power planes, users can better tweak their systems. This should theoretically help power consumption and efficiency as well. Because of the upgrade to HyperTransport 3.0 and split power planes, the 790FX chipset will support what AMD calls Socket AM2+.


Basically, Socket AM2+ is fully pin-compatible with Socket AM2 and vice versa, so you can technically run a Phenom on any current Socket AM2 motherboard or run a current Athlon 64 X2 on the 790FX. However, running a Phenom on an AM2 will mean running in compatibility mode, without the benefits of HyperTransport 3.0 or split power planes (which consequently affects the memory controller's performance potential). Similarly, an older Athlon 64 X2 processor will not benefit from Socket AM2+ as it does not support the new features.





As expected, AMD has also moved towards PCI Express 2.0 and the 790FX chipset features a total of 42 PCIe 2.0 lanes. 32 lanes are dedicated for graphics with two full speed PCIe x16, each of which can split again evenly into four PCIe x8. This will facilitate the new CrossFireX design, which can support 2, 3 or 4-way CrossFire setups. For the remaining ten lanes, six are dedicated PCIe x1 lanes for peripheral connection while the last four seem to be reserved to form the interconnect between the North and Southbridge. In this light, AMD's solution trumps Intel's 3-series chipset because AMD is boasting PCIe 2.0 for almost the entire system, top to bottom, whereas Intel only offers PCIe 2.0 on their high-end X38 chipset, limited again to supplying the GPU lanes and nothing else.

However, there is a weak link in the AMD 790FX chipset, and that happens to be the Southbridge. It would seem that AMD has not been able to roll out their next generation SB700 Southbridge yet, which means that the current batches of 790FX motherboards will still be paired with the older SB600 Southbridge, which doesn't support PCIe 2.0. Compared to chipsets today, the SB600 is also slightly lacking in features with only four SATA 3.0Gbps ports and no built-in Ethernet MAC.

Lastly, the AMD 790FX is supposed to feature great power efficiency. Built on a new 65nm process technology, AMD claims that the 790FX runs with a nominal 8W and a maximum power draw of 10-12W at full load. Because of this, the chipset can even be passively cooled with a simple heatsink. Though from what we've seen, motherboard manufacturers are still sticking to their heat-pipe designs for enthusiast boards.

So in a nutshell, the AMD 790FX and Socket AM2+ isn't all that different from previous generation AM2 motherboards. It serves as an interim before AMD can roll out AM3. Current AM2 users have no real reason to upgrade to the new chipset unless you are investing in the complete Spider platform - motherboard, CPU and GPU. It is only then, will the Spider platform perform at its peak. However, having seen the workstation platform equivalent from AMD with their Barcelona processors, so far there doesn't seem to be much in the way of expectations in either performance or power savings with all the new features on the processor end. So as we've pointed out, this is just a platform renewal for AMD, but don't expect anything exceptional. With that said, we share with you a couple of new motherboards bracing the new AMD 790FX platform on the following pages.


Specifications

The AMD 790FX Chipset is designed for PC enthusiasts,
and performance seeking overclockers*

HyperTransport™ 3.0 technology More than doubles your CPU communications bandwidth to graphics as compared to HT1.
Auto Xpress* The technology for qualified performance boosts that automatically activates with AMD/ATI components to deliver better CPU, GPU, and system performance
* Boosts graphics bandwidth when using AMD processors with HT1
* Reads special Enthusiast DIMM settings for higher memory performance
* Enhanced DDR2 tuning to deliver better system performance. Maximizes performance from new AMD Phenom CPUs

PCI Express® Generation 2.0 Double your graphics bandwidth over earlier PCI Express for improved performance.
GPU-Plex Technology Scalable twin-engine technology, all on a single chip, enables flexible configurations for multiple graphics cards on a single x16 link.
Quad PCIE Blocks* Native PCI-E cores on one chip means fast GPU to GPU communications by eliminating the latency and possible bandwidth issues of multiple PCIE chip designs. Double pumped with Xpress Route for fast Core to Core transfers.
ATI CrossFireX™ Technology The ultimate in scalable graphics performance with a widely certified infrastructure and dedicated memories with improved memory performance.
Backwards compatibility Ensures flexibility to build platforms with AMD Athlon™ today and AMD Phenom™ tomorrow.
AMD OverDrive™ Shift your system performance into next gear. Enables control of the 7-Series Chipsets to allow configuration of system settings in Microsoft Windows.
AMD RAIDXpert Easily configure your RAID setup from remote locations to personalize your media for extra performance or enhanced reliability.
Low Power Design Ensures you have more power available for other components when you need it.

Open Patent Alliance Formed to Advance WiMAX 4G Technology

Alcatel-Lucent, Cisco, Clearwire, Intel, Samsung, Sprint Ally to Form WiMAX Industry Patent Pool

SANTA CLARA, Calif., June 9, 2008 – To accelerate the widespread adoption and deployment of WiMAX technology and products, Alcatel-Lucent, Cisco, Clearwire, Intel Corporation, Samsung Electronics and Sprint today announced the formation of the Open Patent Alliance (OPA). The OPA will advance a competitive and open intellectual property rights model, thus stimulating a larger WiMAX industry that supports innovation through broader choice and lower equipment and service costs for WiMAX technology, devices and applications globally.

WiMAX is a 4G, IP-based broadband wireless technology that provides low-cost, multi-megabit speed and greater throughput for the mobile Internet era of video-rich content and bandwidth-intensive applications. It is based on the IEEE 802.16e standard.

To accomplish its goals, the OPA will form a WiMAX patent pool to help participating companies obtain access to patent licenses from patent owners at a predictable cost.

The patent pool will aggregate essential patent rights needed to implement the WiMAX standard as defined by the WiMAX Forum and the IEEE 802.16e standard. To help ensure product differentiation and interoperability at a more predictable cost, this approach will focus on providing a more competitive royalty structure by charging only for the features required to develop WiMAX products. The patent pool will incorporate a variety of royalty licensing solutions, including accounting for cross-licensing among individual members within the pool.

In addition to the formation of a WiMAX patent pool, the OPA will educate the WiMAX ecosystem about, and serve as a central resource for WiMAX intellectual property rights topics.

The OPA will issue a call for WiMAX essential patents for inclusion in its patent pool. An independent third-party reviewer will serve as the "patent referee" and will evaluate submitted patents to determine how essential they are to the WiMAX standard and WiMAX Forum profiles. While the OPA initially will focus its efforts on the WiMAX standard, it may work with other industry groups in the future.

Made up of six founding companies at its inception, the OPA expects to secure participation of an additional six to nine investor companies to encompass the WiMAX value-chain and broad geographic representation. Each investor company will have a seat on the OPA board of directors. The OPA will have dedicated staff that is directed by and accountable to the OPA board, and the alliance will serve as the licensing agent for the WiMAX patent pool, representing the licensors who agree to participate in the pool.

OPA Founded With Broad, Cross-Industry Support
"As an early and consistent supporter of the WiMAX 802.16e ecosystem, Alcatel-Lucent believes that end-user experiences will drive the broadband wireless revolution," said Philippe Keryer, president of Alcatel-Lucent's mobile access business. "We are confident that the formation of the Open Patent Alliance, which is fully in line with Alcatel-Lucent's Open CPE Program, will ensure that a broad and diverse set of devices will be available to the consumer, enabling service providers to deliver an unparalleled selection of applications and optimized device form-factors that can unleash the huge potential of broadband wireless anywhere, anytime, that WiMAX enables."

"Cisco supports the Open Patent Alliance and its charter to stimulate further adoption of WiMAX broadband wireless technologies and enhance the mobile experience around the world," said Brett Galloway, senior vice president of Cisco wireless and security. "By driving the adoption of WiMAX, innovative technologies will help enable people to move freely throughout the world while maintaining continuous connectivity to cool new applications and services."

"Clearwire fully supports the OPA as a vehicle that will ensure the rapid development of an open, worldwide WiMAX ecosystem," said Scott Richardson, chief strategy officer of Clearwire. "We are committed to innovation in all aspects of WiMAX technology, from the network infrastructure to end-user devices. Just as mobile phones now connect to people rather than places, Clearwire is giving our customers the ability to access the Internet, at true broadband speeds, when they are on the go."

"Intel's vision in driving the formation of the Open Patent Alliance is to bring the openness of the Internet and computing industries to the wireless industry and WiMAX" said Sriram Viswanathan, vice president, Intel Capital and general manager, Intel WiMAX program office. "By facilitating an open intellectual property rights model, we expect the alliance to in turn create a more competitive, innovative and broader WiMAX industry that ultimately benefits the consumer with more choice for WiMAX technology and services."

"Samsung Electronics is very proud to join the OPA with industry-leaders in WiMAX technology," said Hwan Chung, senior vice president of Samsung Telecommunications America. "We believe that this initiation can be a significant momentum to expand mobile WiMAX services. Thanks to a flexible and easily adoptable scheme for participants, we can accelerate the mobile WiMAX deployment and marketing so that people can enjoy the benefits of fast, revolutionarily new, and more cost-effective data communications."

"We expect to unleash wireless innovation with this patent alliance," said Barry West, president of Sprint's XOHM business unit. "The open licensing program will help prompt pervasive WiMAX technology. This speeds the day when consumers can enjoy the convenience of WiMAX mobile broadband services on new devices in new places, and experience all the open Internet has to offer."

For information regarding the submission of patents for testing, visit the OPA Web site at www.openpatentalliance.com. Information about and a replay of the OPA web cast held earlier today can be found at http://newsroom.cisco.com/dlls/2008/prod_060908.html.

About Alcatel-Lucent
Alcatel-Lucent (Euronext Paris and NYSE: ALU) provides solutions that enable service providers, enterprise and governments worldwide, to deliver voice, data and video communication services to end-users. As a leader in fixed, mobile and converged broadband networking, IP technologies, applications and services, Alcatel-Lucent offers the end-to-end solutions that enable compelling communications services for people at home, at work and on the move. With operations in more than 130 countries, Alcatel-Lucent is a local partner with global reach. The company has the most experienced global services team in the industry, and one of the largest research, technology and innovation organizations in the telecommunications industry. Alcatel-Lucent achieved revenues of Euro 17.8 billion in 2007 and is incorporated in France, with executive offices located in Paris. For more information, visit Alcatel-Lucent on the Internet: www.alcatel-lucent.com.

About Cisco
Cisco, (NASDAQ: CSCO), is the worldwide leader in networking that transforms how people connect, communicate and collaborate. Information about Cisco can be found at www.cisco.com. For ongoing news, please go to newsroom.cisco.com.

About Clearwire
Clearwire, founded in October 2003 by wireless pioneer Craig O. McCaw, is a provider of simple, fast, portable and reliable wireless high-speed Internet service. Clearwire customers connect to the Internet using licensed spectrum, thus eliminating the confines of traditional cable or phone lines. Headquartered in Kirkland, Wash., the company launched its first market in August 2004 and now offers service in 50 markets across the U.S. as well as in Europe. For more information, visit www.clearwire.com.

About Intel
Intel (NASDAQ: INTC), the world leader in silicon innovation, develops technologies, products and initiatives to continually advance how people work and live. Additional information about Intel is available at www.intel.com/pressroom and blogs.intel.com.

About Samsung Electronics
Samsung Electronics Co., Ltd. is a global leader in semiconductor, telecommunication, digital media and digital convergence technologies with 2007 consolidated sales of US$103.4 billion. Employing approximately 150,000 people in 134 offices in 62 countries, the company consists of four main business units: Digital Media Business, LCD Business, Semiconductor Business, and Telecommunication Business. Recognized as one of the fastest growing global brands, Samsung Electronics is a leading producer of digital TVs, memory chips, mobile phones and TFT-LCDs. For more information, please visit www.samsung.com.

About Sprint Nextel
Sprint Nextel offers a comprehensive range of wireless and wireline communications services bringing the freedom of mobility to consumers, businesses and government users. Sprint Nextel is widely recognized for developing, engineering and deploying innovative technologies, including two wireless networks serving nearly 53 million customers at the end of the first quarter 2008; industry-leading mobile data services; instant national and international push-to-talk capabilities; and a global Tier 1 Internet backbone. For more information, visit www.sprint.com.

MRAM(Magnetoresistive Random Access Memory)

Unlike conventional RAM chip technologies, in MRAM data is not stored as electric charge or current flows, but by magnetic storage elements.

The elements are formed from two ferromagnetic plates, each of which can hold a magnetic field, separated by a thin insulating layer. One of the two plates is a permanent magnet set to a particular polarity, the other's field will change to match that of an external field. A memory device is built from a grid of such "cells".

Reading is accomplished by measuring the electrical resistance of the cell.A particular cell is (typically) selected by powering an associated transistor which switches current from a supply line through the cell to ground. Due to the magnetic tunnel effect, the electrical resistance of the cell changes due to the orientation of the fields in the two plates. By measuring the resulting current, the resistance inside any particular cell can be determined, and from this the polarity of the writable plate. Typically if the two plates have the same polarity this is considered to mean "0", while if the two plates are of opposite polarity the resistance will be higher and this means "1".

Data is written to the cells using a variety of means. In the simplest, each cell lies between a pair of write lines arranged at right angles to each other, above and below the cell. When current is passed through them, an induced magnetic field is created at the junction, which the writable plate picks up. This pattern of operation is similar to core memory, a system commonly used in the 1960s. This approach requires a fairly substantial current to generate the field, however, which makes it less interesting for low-power uses, one of MRAM's primary disadvantages. Additionally, as the device is scaled down in size, there comes a time when the induced field overlaps adjacent cells over a small area, leading to potential false writes. This problem, the half-select (or write disturb) problem, appears to set a fairly large size for this type of cell. One experimental solution to this problem was to use circular domains written and read using the giant magnetoresistive effect, but it appears this line of research is no longer active.

Another approach, the toggle mode, uses a multi-step write with a modified multi-layer cell. The cell is modified to contain an "artificial antiferromagnet" where the magnetic orientation alternates back and forth across the surface, with both the pinned and free layers consisting of multi-layer stacks isolated by a thin "coupling layer". The resulting layers have only two stable states, which can be toggled from one to the other by timing the write current in the two lines so one is slightly delayed, thereby "rotating" the field. Any voltage less than the full write level actually increases its resistance to flipping. That means that other cells located along one of the write lines will not suffer from the half-select problem, allowing for smaller cell sizes.

A newer technique, spin-torque-transfer (STT) or Spin Transfer Switching, uses spin-aligned ("polarized") electrons to directly torque the domains. Specifically, if the electrons flowing into a layer have to change their spin, this will develop a torque that will be transferred to the nearby layer. This lowers the amount of current needed to write the cells, making it about the same as the read process ^[1]. There are concerns that the "classic" type of MRAM cell will have difficulty at high densities due to the amount of current needed during writes, a problem STT avoids. For this reason, the STT proponents expect the technique to be used for devices of 65 nm and smaller.^{[citation needed]} The downside is that, at present, STT needs to switch more current through the control transistor than conventional MRAM, requiring a larger transistor, and the need to maintain the spin coherence. Overall, however, the STT requires much less write current than conventional or toggle MRAM.

Comparison with other systems

Density

The main determinant of a memory system's cost is the density of the components used to make it up.DRAM uses small capacitors as a memory element, wires to carry current to and from it, and a transistor to control it – referred to as a "1T1C" cell. Capacitors basically consist of two small metal plates separated by a thin insulator, a single element that can be built as small as the current fabrication technology allows. This makes DRAM the highest density RAM currently available, and thus the least expensive, which is why it is used for the majority of RAM found in a computer.
MRAM is physically similar to DRAM in makeup, although often does not require a transistor for the write operation. However, as mentioned above, the most basic MRAM cell suffers from the half-select problem, which limits cell sizes to around 180 nm or more. Toggle-mode MRAM offers a much smaller size before this becomes a problem, apparently around 90 nm ^[2], the same size as most current DRAM products. To be worth putting into wide production, however, it is generally believed that MRAM will have to move to the 65 nm size of the most advanced memory devices, which will require the use of STT.

Power consumption

Since the capacitors used in DRAM lose their charge over time, memory assemblies using them must periodically refresh all the cells in their chips approximately 1000 times a second, reading each one and re-writing its contents. This demands a constant power supply, which is why DRAM loses its memory when power is turned off on the computer. As DRAM cells decrease in size, the refresh cycles become shorter, and the power draw more continuous.

In contrast, MRAM requires no refresh at any time. Not only does this mean it retains its memory with the power turned off, but also that there is no constant power draw. While the read process theoretically requires more power than the same process in a DRAM, in practice the difference appears to be very close to zero. However, the write process requires more power in order to overcome the existing field stored in the junction, varying from three to eight times the power required during reading . Although the exact amount of power savings depends on the nature of the work – more frequent writing will require more power – in general MRAM proponents expect much lower power consumption (up to 99% less) compared to DRAM. STT-based MRAMs eliminate the difference between reading and writing, further reducing power requirements

Speed

DRAM speed is limited by the speed at which the charge stored in the cells can be drained (for reading) or stored (for writing). MRAM operation is based on measuring voltages rather than charges or currents, so there is less "settling time" needed.
IBM researchers have demonstrated MRAM devices with access times on the order of 2 ns, somewhat better than even the most advanced DRAMs built on much newer processes

The only current memory technology that easily competes with MRAM in terms of speed is Static RAM, or SRAM. SRAM consists of a series of transistors arranged in a flip-flop, which will hold one of two states as long as power is applied. Since the transistors have a very low power requirement, their switching time is very low. However, since an SRAM cell consists of several transistors, typically six or four, its density is much lower than DRAM. This makes it expensive, which is why it is used only for small amounts of high-speed memory, notably the CPU cache in almost all modern CPU designs.

Overall

MRAM has similar speeds to SRAM, similar density of DRAM but much lower power consumption than DRAM, and is much faster and suffers no degradation over time in comparison to Flash memory. It is this combination of features that some suggest make it the "universal memory", able to replace SRAM, DRAM and EEPROM and Flash. This also explains the huge amount of research being carried out into developing it.

In comparison, MRAM is still largely "in development", and being produced on older non-critical fabs. The only commercial product widely available at this point is Freescale Semiconductor's 4 Mbit part, produced on a several-generations-old 180 nm process. As demand for Flash continues to outstrip supply, it appears it will be some time before a company can afford to "give up" one of their latest fabs for MRAM production. Even then, MRAM designs currently do not come close to Flash in terms of cell size, even using the same fab.

IGEL’s Latest Microsoft® Windows® XP Embedded Firmware

IGEL’s Latest Microsoft® Windows® XP Embedded Firmware Allows Customers to Access their Virtual PCs Using an Open Hypervisor Strategy

The Inclusion of the Ericom® PowerTerm® WebConnect Client Allows Customers to Choose from 14 Hypervisors When Delivering Virtual PCs. New Functionality also Gives Higher Degrees of Control that Result in More Environmentally Friendly Devices that are Remarkably Flexible

IGEL Technology today announced the availability of its latest Microsoft® Windows® XP Embedded firmware on the broadest and most powerful range of thin clients on the market. The latest firmware gives users increased choice of hypervisors for virtual PCs, improved device management and further energy saving capabilities.

IGEL Universal Desktops powered by Microsoft Windows XP Embedded now have additional Ericom features included as standard. The new firmware version includes the Ericom® WebConnect client. This provides users with secure, local and remote access to applications and desktops running on Windows Terminal Servers, Virtual Desktops (VDI), Blade PCs and legacy host systems. With support for 14 hypervisors, including VMware® ESX, Citrix® XenServer, Microsoft Hyper-V, Oracle® VM, Virtual Iron and others, Ericom® PowerTerm® WebConnect enables organisations to flexibly choose the VDI environment that best suits user requirements.

Following research from the renowned German Fraunhofer Institute last year, which showed IGEL thin clients used up to 50% less power than conventional business PCs, the new firmware also delivers even more environmentally friendly control.

Using the IGEL Remote Management Suite software bundled with every thin client, the power management configurations of each unit deployed across an organisation can be controlled centrally. Sleep times can be set and managed remotely, ensuring that thin clients are powered down across the business when not in use – saving power and reducing CO2 emissions. User experience is also improved with sessions able to remain active during sleep mode and allowing faster power return, meaning users can instantly return to work without lengthy reboot and connection times.

The latest firmware also includes File Based Write Filter (FBWF) which delivers improved security and management capabilities to the IGEL thin client by allowing tightly defined, secure access rights to the flash memory to deploy local applications that require permanent local files, such as anti-virus updates and browser cookies. All features on FBWF can be configured at the local set-up console or using IGEL’S Remote Management Suite.

Additional features in the firmware include the Internet Explorer 7 web browser with its improved tabbed browsing, instant search and improved security features. Users will also experience enhanced video performance and support for a broader range of media with the latest Media Player 11. A feature to enable and disable IP v6 has also been included to increase network performance if IP v6 is not required.

“These latest features ensure that our Windows XP Embedded based products provide more digital services, are easier to manage and more environmentally friendly than ever before,” explained Stephen Yeo, worldwide strategic marketing director for IGEL Technology. “The inclusion of Ericom’s PowerTerm® WebConnect client is particularly exciting since it gives our customers freedom of choice in hypervisor when designing their virtual PC infrastructures.”

The IGEL Windows XP Embedded firmware is immediately available at www.MyIGEL.com.