Linley Wire: July 19, 2016

 weSRCH's Best of the Internet Award

Linley Wire
Volume 16, Issue 9
July 19, 2016

Please feel free to forward this to your colleagues

Independent Analysis of the Networking-Silicon Industry

Editor: Bob Wheeler
Contributor: Loring Wirbel and Tom R. Halfhill

In This Issue:

- Decawave Steers Impulse UWB to IoT
- Qualocmm’s GigaDSL Uses FDD
- Broadwell Accellerates the DPDK
- ConnectX-5 Target HPC Co-Design

To read these articles on-line, click here

Communications Semiconductor Market Forecast 2015-2020

The Linley Group is pleased to announce the upcoming release of the tenth edition of our Communications Semiconductor Market Forecast report.

"Communications Semiconductor Market Forecast 2015-2020" provides the detailed information needed to understand the complexities of this market. Chip vendors, investors, and OEMs can readily see how large are the mature product markets and how fast the emerging categories are growing. New to this edition are forecasts for 100G Ethernet PHYs and 2.5/5GBase-T PHYs.

A detailed description of the report, including the preliminary table of contents, is available on our web site:

Special offer: Get a pre-publication discount of $500 off the corporate license or $300 off a single license until August 19, 2016. For more information or to order the report, please email us at

Decawave Steers Impulse UWB to IoT
By Loring Wirbel

Impulse-based ultrawideband radio (IR-UWB), once an afterthought to the multiband version of the IEEE 802.15.4 WPAN standard, may find applications beyond real-time location services (RTLS). Decawave, an Irish company offering chips and modules for RTLS asset tracking, is trying to meet emerging demands that the wireless protocol serve multiple roles in medical and industrial IoT markets.

After interest in 802.15.4 OFDM-based ultrawideband (UWB) radio waned, developers at Decawave and the European research organization Imec pushed for an impulse version of the standard: 802.15.4a. Dublin-based Decawave heads a small roster of chip startups applying low-bit-rate UWB to in-building asset tracking and statistics gathering. But even as that company expands trial networks into automotive point-to-point radio and factory-floor IoT, some of its competitors are seeking RTLS footholds through differentiated products.

Decawave can claim a longer design and development history than IR-UWB startups such as BeSpoon and BlinkSight. Founded in 2004, it sampled its first single-chip transceiver in 2009. The company has raised more than $30 million over 12 years, including a $4.5 million angel round in 2015 that was the precursor to a planned $20 million round in 2016. This planned funding will aid new automotive- and industrial-IoT design efforts, but it will focus primarily on consumer applications. Decawave is talking with IC specialists in automotive access control about developing secure communications between its keyless-entry radio subsystems, individual cars, and, potentially, wireless automotive networks.

The potential business for RTLS alone, as well as broader IR-UWB, could be enough to sustain several players. But for now, three small vendors with focused business plans can help develop a market still in its infancy.

Networking Report subscribers can access the full article here:

Qualcomm’s GigaDSL Uses FDD
By Loring Wirbel

Qualcomm has introduced the first new products from its 2015 acquisition of DSL specialist Ikanos. The latter company designed both the QCO5700 for multiple-dwelling units (MDUs) and the QCM5720 for customer premises equipment (CPE) before completion of the deal in 3Q15. Branded GigaDSL, the chips upgrade VDSL networks using a frequency-division-duplex (FDD) version of

The two devices were under development at Ikanos before the International Telecommunication Union (ITU) approved G.9701 ( in 2013. The standard specifies time-division duplexing (TDD), causing some inherent incompatibility between the pedestal and existing VDSL deployments. Several Korea- and Japan-based operators and OEMs have pushed for an FDD alternative. Japan’s Miti intended to approve GigaDSL as an alternative to TDD in 2Q16 or 3Q16; it formed a liaison with the ITU to consider an annex standard for gigabit DSL as part of

Qualcomm pledges to follow the TDD standard outside Korea and Japan, but it believes GigaDSL offers a faster upgrade from VDSL in some regions. The company gained immediate benefits from its Ikanos acquisition by working with Asia-based OEMs and operators to bring an FDD product to market quickly. To date, however, it’s the only chip supplier backing ITU standardization of GigaDSL.

Deployment realities of existing fiber-to-the-home (FTTH) and GPON pedestals will guarantee a market for GigaDSL over the next two to three years, but Qualcomm should avoid delaying development of a standard TDD-based portfolio. The company will likely face GigaDSL competitors, particularly Broadcom and perhaps Huawei’s HiSilicon. But GigaDSL remains a transitional technology, and we expect traditional TDD-based to dominate larger post-VDSL growth.

Networking Report subscribers can access the full article here:

Broadwell Accelerates the DPDK
By Tom R. Halfhill

Intel is quickening its march into networking with new acceleration features in the latest Broadwell Xeon chips. These features speed up common tasks such as cryptography, packet I/O, forwarding, and virtualization. As usual, though, the company prefers to execute most networking tasks in software running on its powerful x86 CPUs instead of using specialized hardware engines. Other processor vendors prefer the latter approach.

Fortunately for OEMs and their customers, these differences are partly mitigated by using the Data Plane Development Kit—originally an Intel invention that the industry has adopted as a BSD-licensed open-source standard through the community. All the major vendors of networking-oriented RISC SoCs have embraced the DPDK as well.

The latest release enables four quality-of-service techniques that Intel collectively calls Resource Director Technology (RDT). These techniques include cache-allocation technology, code/data prioritization, cache-monitoring technology, and memory-bandwidth monitoring. Together, they keep transient packets from clogging the caches and displacing important program code and data.

Although several Haswell Xeon chips implemented some RDT features, new Broadwell Xeons implement all of them. These features are supported in DPDK 16.04, which released in April. Further enhancements are coming in DPDK 16.07, which is due in July. ( has replaced the previous version numbering with a Linux-like scheme in which the first two digits represent the year and the last two represent the month.)

Yet another DPDK release is scheduled for November, and four are slated for next year. The rapid cadence of new releases is transforming the DPDK into a more capable API that’s democratizing network software.

Networking Report subscribers can access the full article here:

ConnectX-5 Targets HPC Co-Design
By Loring Wirbel

Mellanox has followed its switch chips with adapter families optimized for specific InfiniBand (IB) or Ethernet applications. The new ConnectX-5 may be optimized for IB, but it targets HPC co-design, which involves development of clustered multicore hardware in close association with simulated HPC applications known as mini-apps. In co-design models, NICs function as coprocessors as data moves through the network.

Consequently, ConnectX-5 must offer greater performance for exascale computing compared with previous IB adapter generations. Its 100Gbps throughput is much less important than the 600ns end-to-end latency. Many HPC topologies prefer IB, but the Mellanox adapter must support Ethernet to address emerging university and consortia HPC co-design projects based on the latter protocol.

The greatest change to adapters as co-design enters the server-cluster mainstream is offloading of some message-passing-interface (MPI) processing from the host. Mellanox has always supported MPI as a user-space API, but it began offloading this function in SwitchIB-2 and ConnectX-4 by implementing collectives processing. ConnectX-5 delivers 200 million messages per second and adds a small on-chip memory of less than a megabyte.

Given its role as the last chip- and card-level supporter of true IB, Mellanox only faces a challenge from Intel’s Omni-Path Architecture (OPA), which displaces the x86 giant’s IB products with a new fabric combining elements of the Cray Aries switch and former QLogic TrueScale IB products.

Networking Report subscribers can access the full article here:

About Linley Wire

Linley Wire is a free electronic newsletter published by The Linley Group, a technology analysis and strategic consulting firm. Linley Wire will present our analysis of recent news on semiconductors for networking and communications. Articles are posted weekly to our web site and sent monthly via email. To access the web content directly, visit our web site.

Domain: Electronics
Category: Semiconductors
weSRCH App on Apple

Recent Newsletters

Linley Newsletter: August 8, 2019

Linley Newsletter Please feel free to forward this to your colleagues Issue #664 August 8, 2019 Independent Analysis of Microprocessors and the Semiconductor Industry E

08 August, 2019

Linley Newsletter: August 1, 2019

Linley Newsletter Please feel free to forward this to your colleagues Issue #663 August 1, 2019 Independent Analysis of Microprocessors and the Semiconductor Industry E

01 August, 2019

Linley Newsletter: July 25, 2019

Linley Newsletter Please feel free to forward this to your colleagues Issue #662 July 25, 2019 Independent Analysis of Microprocessors and the Semiconductor Industry

25 July, 2019