PCI Express Puts Fabrics Front and (Data) Center

by Larry Chisvin, PLX VP of strategic initiatives 

PCI Express, or PCIe, has been around for longer than anybody can remember, but until recently hasn't been regarded as a viable general-purpose fabric.  That’s now changing, as we're seeing designers opting for PCIe as the main interconnect inside data center or cloud-based racks, with either Ethernet or InfiniBand connecting those racks together.  That fact that these technologies coexist and complement one another gives rise to a brand-new application for PCIe -- fabrics – which is bringing the technology to the forefront of data center architectures, while refining the role of traditional interconnect solutions.

To give this some context, Ethernet has historically been the go-to scheme for most networking and connectivity needs, since it has momentum on its side, and is well understood.  Yet, Ethernet’s performance lags in many applications, primarily due to its high latency.  InfiniBand, on the other hand, provides a low-latency, high-performance alternative to Ethernet, and its software base enables powerful systems to be deployed quickly and easily.  However, InifiniBand is expensive and, therefore, tends to be used in a limited number of environments – those in which the better performance justifies the cost premium.
Another way of looking at this is, if the application’s main objective is simple connectivity, 40G or 100G Ethernet is a reasonable choice.  If the highest performance is desired and cost isn’t an issue, then InfiniBand makes more sense, especially at the enhanced data rates.

But while Ethernet and InfiniBand each has some desired attributes and an established track record in connecting racks together, neither is well-suited for interconnect within the rack.  That distinction belongs to PCIe.  The ideal data center has PCIe inside the rack, and either Ethernet or InfiniBand – or something else, such as the less commonly used Fibre Channel -- connecting the racks together.

While PCIe may be less established as a rack-level fabric, it’s been gaining interest as a powerful interconnect solution for data center for cloud equipment.  It’s emergence is fueled by a number of factors: 1) PCIe combines both high performance and low latency; 2) almost every processor, storage device and I/O device has a native PCIe connection, so component requirements – and their associated costs and power needs -- are kept to a minimum, and 3) PCIe-based fabrics, such as PLX’s award-winning ExpressFabric technology.

ExpressFabric, which presently is being evaluated by top system makers around the world, extends the reach of PCIe by creating a powerful fabric for data center and cloud computing.   It leverages existing hardware and software to create converged, multi-host, shared-I/O systems.  With PCIe already native and dominant inside systems, ExpressFabric enables a PCIe-based box-to-box fabric within a rack, eliminating the need for expensive, power-hungry bridging devices (e.g., adapter cards) to translate PCIe to InfiniBand or Ethernet.  ExpressFabric technology is presently targeted at small- to medium-sized cloud clusters, which is where the majority of the today’s high-volume innovations are taking place.  SSD-based tiered storage, micro-servers and high-end general purpose GPU computing, for example, are ideal applications for ExpressFabric.

Looking forward, we envision PCIe-based fabrics in general and ExpressFabric in particular becoming more integral – front and center, you could say -- to data center architectures, while co-existing rather nicely with Ethernet and InfiniBand.

Proper Punctuation Marks the PCI Express Market

by David Raun, PLX president and CEO

As we hit cruising altitude in the 2014 in the tech industry, the year’s looking to be one of punctuated equilibrium that often characterizes evolution.  By that, I mean things tend to remain steady for a period, but then there’s a burst of innovation and design activity.

A prominent example that’s on everybody’s radar -- thanks to the cloud, data centers and the massiveness of content they contain – is storage.  An early adopter and now significant user of PCIe, enterprise storage has driven demand for architectures that enable highly scalable capacity yet must be efficient in terms of size, power and cost.  Solid-state storage in particular poses a challenge for storage-system designers, who’ve been handcuffed by a lack of standards to enable quick, efficient designs using off-the-shelf hardware and software.

Fortunately, though, standards such as NVMe, SATA Express, and SCSI Express are emerging to fill the void and make SSD-storage systems easier to implement and more cost-effective.  NVMe-based storage, in fact, is starting to show its colors; we expect that in 2014 this market will begin to ramp as we see systems based on this standard.

Also punctuating the market are enhancements to the PCIe specification, notably the new downstream port containment extension.  This key enhancement tackles the pervasive problem of data errors that often plague availability and reliability.  DPC has been in the works for a while, and we expect it to bear fruit this year, in the form of actual systems using the enhancement.

Not to be overlooked among the PCIe-related advances to evolve this year are the M.2 specification (once known as the Next Generation Form Factor) and M-PCIe (the M being for mobile applications).  M2 is poised to drive adoption of highly dense SSD-based storage on a variety of platforms.  This form factor, while not specific to PCIe, is looking like a catalyst for taking PCIe deeper into storage than it’s ever been.  For its part, M-PCIe was developed to allow PCIe to work efficiently in power-critical mobile devices – significant because heretofore, the alternative methods to achieve power efficiency involved moving between different protocols.

At the architecture level, 2014 will see an expansion of PCIe as a fabric, notably those based on PLX’s emerging ExpressFabric technology.  With PCIe already penetrated in applications inside the box, ExpressFabric should help accelerate the technology’s growth externally by connecting all boxes within racks – and in the process eliminate costly, power-hungry protocol translation hardware.  The beauty of this approach is that with some straightforward extensions of the existing standard, PCIe can be deployed as a low-power, cost-effective, high-performance fabric.

Yes, 2014 is shaping up to be a great year for PCIe in innovation, standards development and new applications.  And to that, let’s add an exclamation point – arguably the most appropriate punctuation of all!

PLX Defines Scalable PCI Express Within the Rack as the Next Disruptive Data Center Fabric

Free Access to Presentation, Demo at Linley Tech Data Center Conference

SUNNYVALE, Calif., Jan. 30, 2012 -- PLX Technology, Inc. (NASDAQ: PLXT), a leading global supplier of software-enriched, high-speed connectivity silicon for the data center, today announced it has been selected to present at next week’s Linley Tech Data Center Conference, held February 7-8 in San Jose, Calif., and hosted by respected industry analyst specialists The Linley Group.  Registration is free if completed by February 1, 2012.

The PLX® presentation, February 7 at 1:00 p.m., highlights how PCI Express (PCIe) is evolving by extending its dominance within the box, to be an external connectivity of choice within the rack to create clusters of server, switch and storage appliances.  An ExpressFabric® based on PCIe Gen3 and Gen4 inside the rack is complementary to InfiniBand and Ethernet in next-generation cloud-driven data centers.  PCIe does not replace the existing network itself, but instead extends the benefits of PCIe outside the box by moving network interface cards (NICs) to the top of the rack -- or edge of the cluster -- thereby reducing cost and power while maintaining features offered by other network fabrics like InfiniBand and Ethernet.

PCIe is the lowest power, lowest cost solution, and it negates the cumbersome need to translate multiple interconnects, thus resulting in lower latency and higher performance.
PLX will showcase its industry-leading PCIe Gen3 switch and 10GBase-T PHY solutions at the conference with demonstrations highlighting as-yet-unannounced cutting-edge hardware and customer examples on Tuesday evening from 5 p.m. to 7 p.m.

The free two-day, single-track event is the only one of its kind focused on processors, switches, backplanes, and interconnects for data center networking and servers.  The conference is intended for network-equipment vendors, server OEMs, system designers, network service providers, enterprise-network managers, and software developers.

The Linley Data Center Conference will be held from 9 a.m. to 5 p.m. daily, February 7-8, 2012, at the DoubleTree Hotel in San Jose, Calif.  For free registration, visit http://www.linleygroup.com/events/register.php?num=17.

PLX today offers 11 ExpressLane™ PCIe Gen3 devices and is the world’s number-one supplier of PCIe switching solutions, with more than 68 percent market share due to its superior technology and first-to-market leadership.  With an available portfolio ranging from PCIe Gen1, Gen2 and industry-first, production-ready Gen3 devices, PLX today offers more than 40 highly configurable, low-power, low-latency PCIe switches backed by elite productivity- and performance-enhancing free development software.

Watch a related video tutorial hosted by a PLX engineering executive explaining PCIe as a data center clustering interconnect.

• http://plxtech.com/files/products/expresslane/pcie_clustering.html

PLX Enables High-Performance PCI Express on ASRock’s Flagship X79 Graphics Platform

Optimized Package Ideal for Cost-Conscious Consumer Market

SUNNYVALE, Calif., USA. and TAIPEI, Taiwan – December 21, 2011 – PLX Technology, Inc. (NASDAQ: PLXT), the leader in high-speed connectivity solutions for the enterprise and the home, and leading motherboard maker ASRock Inc. (3515:TT) today announced PLX’s ExpressLane™ PEX8605 switch is providing the high-performance PCI Express (PCIe) interconnect on ASRock’s new flagship X79 Extreme9 motherboard.  The integrated PLX® PCIe switch connects the Intel-based x79 processor and other endpoints to five PCIe Gen3-ready expansion slots for unparalleled performance and flexibility, satisfying bandwidth-hungry graphics requirements such as extreme gaming and Adobe Photoshop.
“Our X79 Extreme9 motherboards, supported PLX PCI Express switches, were designed to deliver a premium experience for gamers and high-end graphics,” said James Lee, ASRock vice president of sales and marketing.  “PLX has again provided a perfect match for our cutting-edge Extreme9 design with unmatched performance and scalability, while housed in an ideal package.”
“PLX is once again delivering superior high-performance PCI Express fan-out capabilities to ASRock motherboards,” said Gene Schaeffer, vice president of global sales at PLX.  “Working closely with the ASRock development team has been instrumental in our understanding of the needs of this exciting market and we look forward to opportunities in its next generation of products.”
The PLX ExpressLane PEX8605 (four lanes, four ports) Gen2 switch offers very low power requirements (down to 0.6W typical) due to its high-efficiency power management feature designed to meet the difficult power constraints of battery- and bus-powered applications.  In addition, PLX has integrated reference clocks into the device, which in turn sheds the requirement for multiple external clock and timing ICs, thereby reducing the number of vendors and overall bill-of-material (BOM) costs.
PLX is the world’s number-one supplier of PCIe switching solutions, with a commanding 68 percent market share due to its superior technology and first-to-market leadership.  With an available portfolio ranging from PCIe Gen1, Gen2 and industry-first, production-ready Gen3 devices, PLX today offers more than 40 highly configurable, low-power, low-latency PCIe switches backed by elite productivity- and performance-enhancing free development software.
Product Information

• For more information on the PLX PEX8605 go to www.plxtech.com/8605.
• For information on the entire PLX PCIe portfolio go to www.plxtech.com/pcie.
• For more information on ASRock’s X79 Extreme9 Gen3 motherboard, go to http://www.asrock.com/mb/overview.asp?Model=X79%20Extreme9

Altera Interoperates First 28-nm FPGA with PLX Technology's PCIe Gen3 Switch

SAN JOSE, Calif., Dec. 12, 2011 -- Altera Corporation (Nasdaq: ALTR) today announced it has successfully interoperated its 28-nm Stratix® V GX FPGA with PLX® Technology's (Nasdaq: PLXT) ExpressLane™ PCI Express® (PCIe®) Gen3 switch. Stratix V GX FPGAs feature hard PCIe Gen3 IP blocks and are the only FPGA shipping today that interoperate with a PCIe Gen3 switch.
"Interoperating the industry's highest-performance 28-nm FPGA with the industry's first PCIe Gen3 switch saves significant development time by enabling customers to focus on design issues rather than verifying electrical compliance between devices," said Patrick Dorsey, senior director of component marketing at Altera. "The hard PCIe Gen 3 IP blocks in Stratix V conserve logic resources, while increasing throughput and reducing power consumption for high-performance, leading-edge applications."
Stratix V FPGAs feature up to four hard PCIe Gen3 x8 IP blocks. The PCIe Gen3 IP blocks support x1, x2, x4 and x8 lane configurations and provide transfer rates up to 8-Gbps per lane, sustaining 2X higher throughput using Gen3 x8 lanes compared to the previous Gen2 x8 version. Hardening the PCIe IP blocks in Stratix V FPGAs delivers a savings of over 100,000 logic elements when compared to alternative soft implementations. The hard PCIe Gen3 IP blocks embed the PCIe protocol stack into the FPGA and include the transceiver modules, physical layer, data link layer and transaction layer. Stratix V FPGA's PCIe Gen3 IP targets PCIe Base Specification Rev. 3.0, 2.x, and 1.x.
"Interoperation with Stratix V GX FPGAs and PLX PCIe Gen3 switches is critical in the development of the ecosystem and we are pleased to see leaders like Altera developing Gen3 products," said David Raun, vice president of marketing and business development at PLX.  "Having launched the industry's first PCIe Gen3 silicon in 2010, and remaining the only company shipping production-ready Gen3 switches today, PLX is well-prepared to service this rapidly expanding market."
PCIe Gen3 is the latest version of the industry's most popular high-speed interconnect technology, and PLX switches blend valuable innovation and high port counts to enable new, more powerful designs in servers, storage and communications platforms. The PLX Gen3 portfolio includes 11 devices ranging from 12 to 48 lanes and three to 18 ports with more configurations in development. All devices are available today from PLX.

PLX Technology Expands 10GBase-T PHY Family; Lowers Power, Cost; Adds Powerful Features

SUNNYVALE, Calif., December 5, 2011 – PLX Technology, Inc. (NASDAQ: PLXT), the leader in high-speed connectivity solutions for the enterprise data center, today announced the expansion of its TeraPHY® 10GBase-T PHY family with three new single-, dual- and quad-port integrated transceivers.  Designed on a 40-nanometer lithography node, these devices represent the fourth generation of PLX® 10GBase-T PHYs and its second generation in 40nm.  This latest PLX TN8000 family of devices bring to market several unique features and attributes requested by multiple Tier-1 customers, and are already being designed into several networking platforms.  The demand for 10GBase-T has escalated due to the phenomenal growth in cloud-based data distribution, combined with the availability of powerful, yet affordable and easy to implement, 10 Gigabit PHY devices, which will now replace the widespread Gigabit infrastructure.
Low power dissipation is crucial in today’s data centers and PLX’s upgraded TN8000 PHY devices require only two watts for 10-meter distances over standard copper cabling, which is a popular networking length.  PLX TeraPHY products automatically determine the length of a cable and only use the necessary power to reach that distance.  If long reach is needed, the TN8000 is capable of 10Gbps operation at distances in excess of 120 meters on standard Cat6A cabling.
Another key feature includes Enhanced Energy Efficient Ethernet (eEEE) for both 10 Gigabit and Gigabit speeds -- this PLX customization of the IEEE 802.3az standard implements special low-power idle modes, resulting in lower costs and further contributes to saving the environment.  Appealing to system designers requiring high-security networks, PLX enhanced its Media Access Control Security (MACSec) circuits with second-generation MACSec features that include options for either variable or fixed latency for network protection.  MACSec allows the setup of very secure private networks by identifying and preventing threats within and beyond a local area network (LAN).  Additional value-added features include RXAUI and enhanced KR interfaces and Wake on LAN support.  The new TN8000 devices will also be able to use a 39.0625 MHz crystal, in addition to the 156.25 MHz oscillator, resulting in reduced bill-of-materials (BOM) cost.
Important for high-density switch applications, the new TN8045 quad-port member of the family is housed in a smaller 25mmX25mm flip-chip BGA package and therefore requires less overall board space than previous solutions.
“Clearly, 10GBase-T is poised for wide-scale deployment as data centers convert from the decade-old 1G to 10G,” said Brad Smith, senior vice president of market researcher LightCounting, which tracks high-speed networking.  “Data centers are filled with Cat6 and Cat6A copper cables, and the cost of re-cabling the infrastructure can dwarf capital expenses on hardware.  Therefore, 10GBase-T PHYs such as those PLX is bringing to market can enable cost-efficient transition to 10G in data centers.”
“We reached out to real-world suppliers of IT solutions that service data center managers and performed a comprehensive cost-comparison study between a Direct Attached plus Fiber solution versus a Cat6a cable implementation using 10GBase-T,” said Ron Cates, PLX vice president of marketing, networking products.  “The analysis revealed that with the new 10GBase-T transceiver prices, a Cat6a-based solution is significantly more cost-effective.  When these lower costs are added to the other inherent advantages of 10GBase-T technology, such as interoperation with legacy 1000Base-T links and the ability to reach 100-meter distances, they open the door wide for IT managers looking to reap the advantages of a 10 Gigabit performance upgrade.”
With the market’s broadest 10GBase-T PHY portfolio, PLX has enabled the industry’s first 10GBase-T equipment, and the company today owns more than 60 percent market share for 10GBase-T PHY devices.
Articles, Videos:

• 10GBase-T Technology Revisited (EE Times)
• Electromagnetic Interference and 10GBase-T in the Data Center (Network World)
• The Benefits of 10GBase-T Connectivity in Data Centers (Data Center Journal)
• Achieve 300 Meter Reach with 10GBase-T Transceivers (YouTube)

Press Graphic, Tools

• A downloadable high-resolution press graphic can be found on this page
• Product information can be found on www.plxtech.com/teraphy

EDN Magazine Names PLX TN8022 Dual-Port 10GBase-T PHY to Prestigious Hot 100 Products of 2011 List

SUNNYVALE, Calif., November 30, 2011 – PLX Technology, Inc. (NASDAQ: PLXT), the leader in high-speed connectivity solutions for the enterprise data center, today announced the influential EDN magazine has awarded the PLX® TeraPHY® TN8022 dual-port 10GBase-T physical layer (PHY) device a “Hot 100” product of 2011 from among thousands of candidates.  EDN editors noted the TN8022’s breakthrough performance and power achievements based on unique advancements, such as built-in noise and interference mitigation, robust DSP algorithm, and ultra-low power capabilities that feature a 50 percent reduction in power dissipation for typical long-term traffic patterns.
EDN’s assessment of the TN8022 was elevated when PLX announced it had successfully demonstrated the PHY in 10Gbps Ethernet data and electrical transfer, in full compliance with the IEEE 802.3an standard, over a distance of 300 meters using typical, widely installed Cat6A unshielded twisted pair copper cable.  The unique networking demonstration, enabled by PLX-designed Power-over-Ethernet (PoE) repeaters, utilizes a TN8022 dual-port 10GBase-T PHY transceiver.  This breakthrough distance achievement with low-cost Ethernet cabling brings new capabilities to IT managers seeking high-performance networking between multiple server rooms, which, for example, may be located on different floors of a high-rise building or in a campus environment.
“We are pleased that EDN has honored PLX by recognizing the TeraPHY 10GBase-T transceiver,” said Ron Cates, vice president of networking products, PLX.  “With its lower power and cost, excellent EMI immunity and advanced security features such as MACSec, the TN8022 is well-positioned to stimulate widespread adoption of 10GBase-T technology in the data center.”
The TeraPHY 10GBase-T PHY family includes single-, dual- and quad-port integrated transceivers designed on a 40-nanometer lithography node, with products in full production and more to be announced.
Articles, Videos:

• 10GBase-T Technology Revisited (EE Times)
• Electromagnetic Interference and 10GBase-T in the Data Center (Network World)
• The Benefits of 10GBase-T Connectivity in Data Centers (Data Center Journal)
• Achieve 300 Meter Reach with 10GBase-T Transceivers (YouTube)

PLX TeraPHY product information can be found on www.plxtech.com/teraphy
EDN’s Hot 100 List can be found on its Website.