Linley Newsletter: January 17, 2018

 weSRCH's Best of the Internet Award

Linley Newsletter

Please feel free to forward this to your colleagues

Issue #635

January 17, 2019

Independent Analysis of Microprocessors and the Semiconductor Industry

Editor: Tom R. Halfhill

Contributors: Linley Gwennap, Mike Demler, Bob Wheeler

In This Issue:

- Ryzen Mobile Rises to 12nm

- Achronix eFPGA Serves AI Two Ways

- Year in Review: Mobile Vendors Prepare for 5G

New Deep Learning Report From The Linley Group

Our report "A Guide to Processors for Deep Learning" has been completely revised to reflect the growth in this rapidly evolving market. AI has been dominating the news with dozens of new products and technologies announced in just the last year. Deep learning isn't limited to just data centers and self-driving cars, it is even moving into client devices such as smartphones and embedded (IoT) systems.

This completely revised report focuses on the hardware technologies and products driving this market. You'll find detailed technical coverage of announced chip products from AMD, Eta Compute, Graphcore, Gyrfalcon, Habana, Intel (including former Altera, Mobileye, Movidius, and Nervana technologies), Mythic, NXP, Nvidia (including Tegra and Tesla), Qualcomm, Wave Computing, and Xilinx. It also covers IP cores from AImotive, Arm, Cadence, Cambricon, Ceva, Imagination, Synopsys, Videantis, and the open-source NVDLA. Other chapters cover Google's TPU family of ASICs and Microsoft's Brainwave.

Order by January 31 to take advantage of the introductory offer. For more information on this new edition, visit our website:

Ryzen Mobile Rises to 12nm

By Tom R. Halfhill

Amid flat laptop shipments, PC vendors hope gaming notebooks will attract more users willing to pay higher prices for higher performance. AMD's new second-generation Ryzen Mobile processors attack Intel's dominance in this specialized segment while broadening the choices for mainstream users. Although they introduce no new features relative to the first generation, they nudge clock speeds upward and provide higher power/performance options for notebooks that can dissipate the additional heat.

AMD is exploiting Intel's long-delayed move to 10nm technology by manufacturing most of its new Ryzen Mobile products in a 12nm FinFET process that boosts CPU and GPU clock speeds compared with its existing 14nm chips. To gain even more clock-frequency headroom, the new H-series raises the TDP to 35W -- more than twice that of the U-series. And in a bid for bargain seekers, the new A-series drops the TDP to only 6W and uses trailing-edge 28nm CMOS technology to offer low-power, low-cost chips for Chromebooks. All told, the second-generation Ryzen Mobile family is a broad response to Intel's popular Core i3, Core i5, and Core i7 mobile processors, including the latest Whiskey Lake and Coffee Lake models.

In most respects, the new Ryzens are identical to their predecessors. They have the same first-generation Zen CPUs, Radeon Vega GPUs, integrated south-bridge functions, and I/O interfaces. The process shrink from 14nm to 12nm follows the same path as the second-generation Ryzen desktop models, boosting clock speeds by about 5-10%. Although the enhancements are minor, they show that AMD continues to execute on an annual cadence without serious hiccups. The improved execution bodes well for future transitions to the Zen 2 core and 7nm node, which will make these products even better.

Microprocessor Report subscribers can access the full article:

Achronix eFPGA Serves AI Two Ways

By Bob Wheeler

It's a classic chip-design conundrum. By the time you freeze your specification and then design, verify, tape out, and sample the silicon, customer requirements have changed. To make matters worse, in the world of AI, neural-network models are evolving at lightning speed. If your deep-learning accelerator can't efficiently map the newest models, someone else's probably can. A great challenge in accelerator design is therefore to find the sweet spot between flexibility and efficiency. Presently, FPGAs occupy that middle ground between general-purpose processors and purpose-built accelerators.

To address this challenge, Achronix recently introduced the Speedcore 7t, its 7nm embedded-FPGA (eFPGA) intellectual property (IP). This fourth-generation (Gen4) architecture adds a machine-learning-processor (MLP) block to the menu from which customers configure their core. The MLP is essentially a DSP block optimized for neural-network matrix math. The Speedcore 7t arranges blocks in homogeneous columns of logic (LUT6), block RAM (BRAM), logic RAM (LRAM), and MLPs. On the basis of customer specifications, the company lays out a block with X columns of Y height.

In addition to the new MLP block, Achronix enhanced its Gen4 architecture relative to the prior-generation Speedcore it delivered in 2016. It has already released Gen4 support -- excepting MLP -- in its Ace design tools. Ace support for MLP blocks is due in 1Q19. All Speedcore 7t IP is available now, and Achronix plans to release a 16nm Gen4 design in mid-2019. More surprisingly, it plans to introduce 7nm Speedster FPGA chips (not just IP) in 1H19. Achronix is first to deliver eFPGA IP for 7nm technology, and customers in all segments will benefit from Gen4 enhancements.

Microprocessor Report subscribers can access the full article:

Year in Review: Mobile Vendors Prepare for 5G

By Linley Gwennap

The mobile industry is tremendously excited -- perhaps too excited -- about the launch of next-generation 5G cellular technology in 2019. Despite the widespread interest, however, 5G will be less than 1% of phone shipments in 2019, and perhaps a few percent in 2020. Instead, most phone buyers will continue to purchase lower-cost LTE models.

In the premium segment, three top suppliers -- Apple, Samsung, and Huawei -- develop their own processor chips to help differentiate their phones. Vendors such as OnePlus, Oppo, and Xiaomi instead employ Qualcomm silicon to deliver competitive products with similar capabilities. These premium processors typically offer the highest-performance CPUs and graphic units, powerful AI engines, and LTE downloads at 1.0Gbps or faster.

For customers who don't want to pay $400 and up for a premium smartphone, phone makers have established a new "mid-premium" segment in the $250-$400 range. These phones offer some premium features, such as high-speed CPUs and AI engines, but they lag in other areas, such as GPU performance and LTE throughput. Although MediaTek hasn't broken into the premium tier, it's been successful in the mid-premium tier, most recently with its Helio P60. To counter, Qualcomm has created the new 700-series for mid-premium phones.

The growing mid-premium segment is a bright spot, but total smartphone shipments were flat in 2018 for the second straight year. This stagnation increases the pressure on mobile-chip vendors to move into other markets. Wearables shipments remain a small fraction of phone shipments, but Qualcomm hopes to goose sales with a trio of new products for everything from traditional smartwatches to smart headsets. At the other end of the power scale, the company competes head-on against Intel for 2-in-1 PCs.

Microprocessor Report subscribers can access the full article:

About Linley Newsletter

Linley Newsletter is a free electronic newsletter that reports and analyzes advances in microprocessors, networking chips, and mobile-communications chips. It is published by The Linley Group. To subscribe, please visit:

Domain: Electronics
Category: Semiconductors

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