While UFS has held much hype as being the successor to eMMC providing greater performance, it has yet to be seen if the mobile market as a collective whole has embraced this technology, according to Robert Bielby, senior director of Automotive at Micron Technology.A quick poll of the leading smartphone manufacturers yields an inconsistent story in regards to their propensity to move away from the existing, proven, eMMC technologies in favor of adopting UFS, Bielby wrote in an email response to a recent Digitimes Research article about the prospect of the UFS market.In practice most of the potential performance that is offered by UFS at the physical layer is obfuscated due to the immature, bloated software interfaces that ultimately affect overall system throughput, Bielby said.For the automotive market, the immaturity of the UFS will present unique challenges in embracing a technology that was targeted for a consumer application, missing the robustness and safety required for next generation automotive applications where safety, end to end data path protection is increasingly a non-negotiable capability, asserted the Micron executive.According to Bielby, with 35 years of industry adoption, the PCI standard is well proven across a broad spectrum of applications with well refined software stacks and system level integrity that is aligned to the needs of the most demanding markets.NVMe based storage solutions are now seeing increased consideration over UFS in situations when higher system level throughput, density, or system level integrity is essential.For lower density solutions, eMMC is still proving to have long-legs as extensions to the existing spec such as 5.2 continue to extend the longevity of this solution which has already proven to be the industry's workhorse.In short, while the UFS spec has seen a lot of hype in the past year, what is not clear is the ability for UFS to deliver to the promise of higher performance, Bielby continued. The broader mobile market is on the fence in adopting this technology, and is showing a trend to remaining with eMMC at lower densities, and embracing (PCIe based) NVME in applications that demand higher performance, higher density, or reliability that extends beyond and early generation consumer specification.At best, UFS will prove to be a distraction to the broader storage customer base for those who consider the technology, and most likely prove to be a dead-end strategy as OEMs will have to re-think their storage strategy in light of the existence of more robust, proven technologies, he added.Micron believes UFS will prove to be a distraction to the broader storage customer base for those who consider the technology.Photo: Digitimes file photo

Mini LEDs, with much higher yield rates than those for Micro LEDs, are expected to come into commercial production next year, with initial application expected to be backlighting of LCD automotive displays, according to Digitimes Research.Mini LEDs range from 100-200 microns in size, larger than sub-30-micron Micro LEDs. The former's much larger size makes them less difficult in mass transfer and mass inspection, offering higher yield rates, Digitimes Research indicated.As compared with conventional LEDs used in direct-type backlighting, Mini LEDs offfer better display quality, more flexibility for device design and curved-surface screens, less power consumption and visibility under strong sunlight.For direct-type backlighting of automotive displays, a single panel will use several thousand to over 10,000 Mini LED chips and there will be monthly global demand for an estimated 15.56 billion Mini LED chips in the first year of commercial use.The estimated global demand will need production capacity support of about 48 4-inch MOCVD sets for 200-micron x 200-micron equvilanets assuming yield rates of 85-90% and almost 30 MOCVD sets for 150-micron x 150-micron equivalents at yield rates of 80-85%.Prices for a conventional LED backlight unit (BLU) for a 10-inch or above LCD automotive display stand at US$2.7-3.6 currently. Without considering yield rates, prices for a BLU made of 150-micron x 150-micron and 200-micron x 200-micron LED chips are estimated at US$11.50 and US$19.40 respectively. The large price gap accounts for little of the total cost for a car but makes a big difference if Mini LED chips are used in displays of smart wearable devices, IT devices and LCD TVs, and this is why Mini LEDs are expected to be initially adopted for BLUs for automotive display.Mini LEDs offer more flexibility in display design.Photo: Digitimes file photo

According to the Transparency Market Research, the global connected car market is expected to reach US$131.9 billion by 2019 continuing to maintain its steady growth. Businesses are scrambling for a share of the automotive electronics market in recent years. Lelon Electronics, Taiwan's No. 1 manufacturer of aluminum electrolytic capacitors, entered the market at an opportune time and have secured their place in the OEM supply chain of American and European automakers. As the automotive electronics market continues to grow, Lelon is confident in sustaining its business model and market share in the second half of the year.Jimmy Wu, General Manager of Lelon Electronics, noted that automotive LCD panels used to be installed only in luxury cars. However, due to increased importance being placed on vehicle safety, even entry-level cars are now coming equipped with a variety of sensors to collect information, then processed and displayed on LCD monitors. Lelon is one of the select few non-Japanese manufacturers of aluminum electrolytic capacitors that are capable of being used in automotive applications. Their products now recognized by leading international automotive electronics companies. The automotive electronics market accounted for 9% of Lelon's 2016 revenue and is expected to increase to 10% in 2017. Lelon plans to introduce aluminum electrolytic capacitors for automobile transmission systems in the future to expand Lelon's automotive applications product line.Focusing on the development of high-end capacitors to build up the company's positionFounded in 1976, Lelon has engaged in the R&D, manufacturing and sales of aluminum electrolytic capacitors for over four decades. In order to provide comprehensive services for their customers, Lelon expanded into upstream material production of formed aluminum foil and automated production equipment. Through vertical integration of upstream and downstream manufacturing, Lelon is strengthening its competitive edge in technology, quality, cost, and service. Lelon now has complete offerings of aluminum electrolytic capacitors, including aluminum electrolytic capacitors (radial, SMD, snap-in, and screw terminal), organic conductive polymer capacitors, and conductive polymer hybrid aluminum electrolytic capacitors, which are widely used in automotive electronics, telecommunication equipment, cloud computing, and power electronics. Its quality products and services are highly regarded by their customers.Last year, Lelon's major revenue generators included power electronics, telecommunication equipment, automotive electronics, and consumer electronics. With the company transitioning from mid-range to high-end markets, it plans to reduce its share of products used in consumer electronics due to the fiercely competitive pricing environment. Instead focusing its efforts on aluminum electrolytic capacitors for high-end electronics, which are used in automobiles, telecommunication, healthcare, green energy, smart grid, and cloud computing.Wu further commented that the rising demand for consumer electronics, especially flat-panel TVs, smartphones, and tablets, drew a large number of companies to expand into the production of aluminum electrolytic capacitors. However, as prices of consumer electronics continue to fall, passive component markets have become "red oceans" where many vendors have resorted to price-cutting and have sacrificed quality for cost. If Lelon were to use its resources on low-end products with little margin, it would eventually impact its overall profitability and competitiveness. Lelon has always insisted on delivering premium quality products since its establishment.Actively engaging in new technology development and vertical integrationLelon has been able to come out ahead in the competitive market of aluminum electrolytic capacitors due to their long-term devotion to the R&D of advanced technologies. In addition to their own professional R&D team, they have also hired technical consultants from Japan to help refine their technology and quality systems. Furthermore, Lelon has been collaborating with Industrial Technology Research Institute to jointly research critical technologies and participate in the research of material suppliers and customers in order to develop products that cater to market needs.Lelon has generated impressive results in terms of its patents. It has obtained multiple patents in Taiwan, China and Japan, including ones for a V-chip vibration-proof back plate, a vibration-proof capacitor, a capacitor electrolytic solution with heat resistance and high spark voltage, electrolytic capacitors with optimum cooling, and horizontal capacitors. With growing public awareness of the need to protect the environment and an increasingly strict enforcement of environmental protection laws, the supply of aluminum foil for aluminum electrolytic capacitors has become rather limited. Wu pointed out that in an attempt to minimize the impact of market instability and maintain a steady supply for its customers, Lelon had already begun to search for new material suppliers prior to the Japan 3/11 earthquake and conducted testing on a variety of materials in order to understand the characteristics of the materials and the type of capacitors they were suitable for.Demand for polymer capacitors is rising with exciting developments in quick chargingA growing number of mobile devices such as smartphones and tablets now feature quick charging, which requires the use of matching quick chargers. Quick chargers must meet more stringent requirements than conventional chargers and need polymer capacitors to match product design specifications. Polymer capacitors are also used in cloud computing equipment, telecommunication equipment, and power supplies. Lelon has accumulated years of experience in manufacturing polymer capacitors and will be able to produce close to 100 million polymer capacitors per month. This number includes the monthly capacity of 60 million polymer capacitors produced on the equipment acquired from Matsuki Polymer last year, and Lelon's original monthly capacity of 35 million polymer capacitors. After the completion of this expansion, which is expected to happen this year, Lelon will become a leading manufacturer of SMD polymer capacitors made outside of Japan. Benefiting from an increasing market demand, polymer capacitors will become a major revenue growth driver for Lelon this year.Wu remarked that Lelon has adopted a flexible production approach. It is able to quickly adjust the production line according to order changes and thereby meet their customers' delivery schedules. The purchase of Matsuki Polymer's production equipment of polymer capacitors was completed efficiently and economically. This deal will assist in moving up the company's original scheduled date to expand its production capacity and upgrade its production equipment. It will also further enhance the company's competitiveness.Lelon has made remarkable progress in the market of aluminum electrolytic capacitors and has become a chosen partner of multiple global companies by maintaining its goal of providing premium quality aluminum electrolytic capacitors. As it looks forward to a promising future, Lelon will devote additional resources to the development of high-end aluminum electrolytic capacitors, focusing its efforts on the high-end electronics markets.Jimmy Wu, General Manager of Lelon ElectronicsLelon focuses on the development of high-end aluminum electrolytic capacitors, and on actively expanding its high-end electronics customer base.

Within one month, both Intel and AMD announced massive, long overdue refreshes of their server processor families, with EPYC and Scalable Xeon (yes, the credit card-style precious metal ranked) CPUs out. Both of these required brand new sockets, boards, chipset, even the memory and I/O configurations. As their arrival also significantly changes the competitive performance standing between the two X86 rivals, which way should the Taiwanese vendors go towards?Firstly, after both launches, the obvious impact is that the AMD platform has, in both benchmarks and actual application performance, come close enough to - and in quite a few cases, matching and exceeding - Intel, that it deserves far more space than before. Add to this 33% higher memory bandwidth and capacity per socket, plus equally higher PCIe scaling in the most prevalent 2-socket configurations, and there's meat to grind here - even including AMD's usual well proven ability to shoot itself in the foot when it comes to winning the market battle. Intel may have brought in AVX512 floating point into the picture, but it requires software recompilation and, having it enabled at full throughput only on some of the new Xeons (the pricey ones), it may not gain much traction until the next refresh.Secondly, both platforms have massively increased board size and power requirements per socket, making the dual-CPU ATX-sized server or workstation mainboard, a mark of integration excellence we last saw from Gigabyte and Asus in the Broadwell-EP Xeon generation, a virtually impossible thing to do. AMD may still have a theoretical chance here, as its CPUs don't necessitate a chipset (the minimum USB 3 ports can run from few of the CPU's PCIe lanes) and, using 8 pcs of ECC SO-DIMMs per socket (plenty of them shown by Taiwan firms at Computex), squeeze quarter of terabyte of memory and two sockets into the ATX constraint. Otherwise, it's big and bigger boards only, with correspondingly beefier power supplies, liquid cooling and so on.Third, the above increased actual board and system costs (to the makers too looking at the BOM) are coupled with the increased average and peak CPU prices themselves. As expected, Intel has broken into five-digit US$ per socket stickers with few of its "Platinum" Xeon SKUs, getting itself into the pricing range of the still architecturally superior IBM POWER9 realm which, after all, can claim true enterprise mainframe legacy, rather than PC descendance. Knowing the general customers more or less stagnant per-unit server budgets, this could put further margin squeeze on the Taiwan vendor share of the pie, nudging them to go there where lower CPU price allows for more margin space. Or, aim towards the very top end, where, again, the antique X86 architecture isn't seen as a plus – watch POWER9 & Volta likely claim the next TOP1 system with USA "Summit" this November, setting the bar for Taiwanese vendors, like Wistron, to sell aggressively priced baby 2U server versions of this globally.In integer data-dominated commercial apps, a 2-socket 64-core AMD EPYC solution performance would roughly equal that of 2-socket 56-core Intel Xeon - but at 3x lower CPU price yet 33% more memory and I/O. That extra memory and I/O can justify the board vendor charge somewhat more for the board itself, with the related margin benefit to boot. AMD wins here.Floating point-bound jobs, like HPC and to some extent AI, would give Intel a bit of an advantage if using AVX512 extensions, but again that is limited to certain Intel SKUs only - Intel's selective disabling of major features across different SKUs of same product has its drawbacks too. It doesn't help that these extensions are just slightly different from those used on the last generation of Xeon Phi, which Intel is expected to retire after the last Phi, the Knights Hill, appears next year - another recompile needed? So, the HPC users - an increasing customer base for Taiwan board makers - could decide to wait for Intel's next refresh where AVX512 will be more prevalent, to jump on board the new platform.In both cases, the new AMD and Intel platforms are kind of a pilot run for their next year's immediate plug-in successors in denser manufacturing processes, which should enable now-missing features like PCIe V4, faster DDR4 and full AVX512 on both vendors CPUs. AMD definitely offers better bang-for-the-buck with their new line; however it has to prove its manufacturing rollout will match the pent-up industry demand for a true high-end X86 competitor to Intel.I feel the best bet for Taiwan vendors X86 server strategy right now to give equal attention to both vendors in the product line spread, and watch how AMD actually performs over the course of this calendar year. Intel has exhibited obvious long time quasi-monopoly fatigue with its high-end products, with need for a radical new CPU core way overdue after over a decade since its Israel office created what would become the "Core" - and saved Intel's then-CPU business from what was then-superior AMD architecture. We come back full circle, don't we?

While today's mobile devices use either UFS (Universal Flash Storage) or eMMC (embedded multi-media card)-based NAND flash solutions to store information, eMMC5.1 chips are becoming progressively unable to meet the data transmission speed demands of high-end mobile devices, according to Digitimes Research.UFS2.1-based NAND flash chips boast the highest data transmission speed of up to 11.6Gbps (1200MB/s) compared to 3.2Gbps (400MB/s) attainable by eMMC5.1 chips. The difference will make UFS2.1 NAND flash chips the mainstream storage technology for smartphones in 2017, Digitimes Research commented.Samsung Electronics, SK Hynix and Toshiba are currently the major suppliers of UFS-based chips, while Micron Technology is expected to start ramping up its output in 2017. But the supply of UFS chips from these big players has been unable to cope with demand from the high-end mobile device segment in the first half of 2017 as most chipmakers were busy shifting their capacity to 2D/3D NAND flash products. Rising demand from other end-market applications such as SSDs also squeezed supply.With chipmakers planning to ramp up their capacity for UFS chips in the second half of 2017, increased 3D NAND flash capacity coupled with improving yield rates will be instrumental to raising the penetration rate of UFS chips in latter half of 2017. This will pave the way for UFS to spread to the entry-level and mid-tier mobile device segments in 2018.

As dual-lens cameras are being increasingly adopted in high-end smartphones, the proportion of smartphones with dual-lens cameras is expected to rise from 16% in 2017 to 30% in 2018 and possibly to 50% in 2019, according to Digitimes Research.For some dual-lens models, the two lenses have the same specifications, while other models have differing specifications such as resolution level or focal length, Digitimes Research indicated.Dual-lens model come in two structural types:, independent lenses at separate places in the smartphone, or two lenses integrated on a circuit board.Due to technological difficulty, prices for dual-lens models are much higher than for a single-lens, standing at about US$20 versus US$8.Taiwan-based Largan Precision and Genius Electronic Optical, China-based Sunny Optical Technology and Japan-based Kantatsu are the four main makers of smartphone-use optical lenses.Laragn has the largest monthly production capacity at 150 million lenses and will expand capacity to 190 million units in the fourth quarter of 2017, followed by Genius Electronic 70 million, Sunny Optical 50 million and Kantatsu 37.5 million units.In addition to production capacity, Largan has seen the highest gross profitability, with gross margin rising from 66.36% in the second quarter of 2016 to 70.66% in the fourth quarter, and further to 70.94% in the first quarter of 2017.

At LED Expo held in Taipei last month, Taiwan-based LED and laser diode makers demonstrated a strong focus on automotive applications.Lextar Electronics showcased an LED automotive dual-lens headlight module with integration of the high beam and low beam, and a laser automotive headlight module, with the former featuring a wide lighting angle and maximum luminous intensity of 70,000cd (candela) for the high beam and the latter featuring a lighting range of 600 meters. Lite-On Technology showcased LED in-car scenario-based lights and automotive sensing devices, while Genesis Photonics exhibited CSP (chip scale packaging) LED devices and dual-color-temperature LED devices for use in automotive direction indicators and daytime running lights.Ultimems showcased an automotive head-up display, a scanning laser projection module through combining in-house-developed MEMS mirror and a RGB laser lighting source. HPB Optoelectronics exhibited an image detecting system for car speed detection, with the system consisting of two laser devices in wavelength of 830nm, a compact camera module and algorithm software.Government-sponsored Automotive Research & Testing Center showcased international regulations concerning automotive head lighting for safety and related core technologies, according to Digitimes Research.

Half a month in late-June European heatwave may make Computex-time Taipei feel actually comfortable; however there was a good reason this time to endure the heat – the ISC 2017 in Frankfurt, held on 18 to 22 June, as well as some interesting post-expo disclosures, made it more than worthwhile.Frankfurt-based ISC, the "International" sister event of the US-based SC annual supercomputer conference, has a higher proportion of both European and Asian vendors, speakers and visitors alike, including those from Greater China – easier visa situation for the EU certainly helps there. The hardware shown at the ISC (and of course SC) is the very top, cream-of-the-crop, aimed at HPC, datacenters, AI and similar server- and workstation-rich applications. Since supercomputing is the bleeding-edge showcase of high end computing, you'll usually see never-heard-of high end CPU and system platforms, as well as interesting integration, cooling, packaging and other approaches normally not seen even at Computex overclocker competitions.For instance, you'll see the Japanese vendors such as ExaScaler/PEZY, with fully immersed (in Fluorinert liquid, out of all) ultradense batteries of Xeon CPUs, Nvidia GPUs, FPGAs and even ExaScaler own SC2 ultraparallel MIPS processor with on-chip 3-D stacked memory and FP performance akin to that of Nvidia Pascal GPU. There are also our usual Chinese friends like Shenwei from Wuxi showing the first concept mass-market 2-CPU liquid-cooled server boards using high end PC-like water cooling from EK, or FPGA accelerator boards trying to get into the overheated coin mining market. I won't even mention the usual chip (Intel, Nvidia, AMD, IBM) and system (IBM again, Lenovo, NEC, Fujitsu, Dell, HPE and so on) vendors.How about the Taiwanese? Guess what, there are few, yet valuable, shows of presence – while Asus wasn't there, the cousin, Asrock, was there, exhibiting the new dense HPC-class server offerings for the X86 platforms. These half-width and full-width nodes may not be anything revolutionary yet, but they are there to meet the needs of general purpose HPC and datacenter clusters. Tyan showed a bit more, including the upcoming Skylake Xeon Platinum (Purley) as well as AMD EPYC servers. It seems that, thanks to the latter one's abundance of PCIe lanes per socket, there'll be many EPYC boards for I/O intensive computing like storage servers for fast PCIe SSD arrays, for instance. Their 2-socket EPYC board wasn't ready for the show though – seems that's due to AMD CPU issues.Gigabyte had similar server stuff on display, except one interesting thing – a tad over cubic-foot sized immersed liquid multi-node computing box in a transparent enclosure. While far from any production schedule, the attempt justifies some praise as just getting an immersed solution to work right, yet isolate the outside world connection for the box from that same liquid, is some work to do.Finally, Wistron's new 2U high compute server box was shown on the fliers – but the specs had nothing in common with the other boxes around there. Dual IBM POWER9 processors, up to 24 cores each, coupled over NVLink 2.0 shared memory links with dual Nvidia Volta compute GPUs on SXM2 cards, with plenty of spare PCIe Gen4 lanes and up to 4TB memory, makes for some 16 TFLOPs double precision performance, and internal board bandwidth numbers far ahead of that seen in the new X86 CPUs this year. As mentioned in my Computex enterprise story, these new platforms were about to come out, and now they are entering the market – bringing the much-needed diversity. Add some better cooling (liquid cooling was everywhere at the exhibition) for higher density, and you may have that better-margin, more unique set of offerings right here from Taiwan.However, they'll have to learn how to create a whole system offering including service and support – in that respect, Chinese vendors like Lenovo, Inspur and Sugon far outweigh them, with multiple Europe installations and full scale support and partner network there, making money out of every step of the process. I've seen some of these systems, like Sugon ARCTURUS in Slovenia, after the expo, and the solution delivery was first-class.It's a good time to move forward now, as EU is contemplating its own high-end processor design for the next-generation large machines, and Taiwan vendors should endear themselves to the powers that be to provide the board and system hardware for such new platform, this being Taiwan's core competency for decades.In summary, the top-notch supercomputing hardware shouldn't be beyond Taiwan vendors' capabilities, as even China is using its No.1 position to spread its wings around. Yes, the most recent Taiwan order for a Fujitsu machine to be the first petascale national supercomputer may feel a little disappointing; however local vendors should be more emboldened with this market rapid expansion and increase their presence – hope the next ISC has twice as many of them, at least.

June 2017 average retail pricing for 7W LED light bulbs (equivalent to 40W incandescents) in the China market increased 13.6% from April 2017 to CNY26.7 (US$3.9), while 9W models (equivalent to 60W incandescents) rose 1.6% to CNY32.2, according to Digitimes Research.Average retail prices for 40W- and 60W-equivalent LED light bulbs in Japan in June stood at JPY1,417 (US$13.0) and JPY1,834 respectively, growing 9.0% from April and slipping 6.1%, Digitimes Research indicated.June average retail prices for 40W- and 60W-equivalent LED light bulbs in other markets were: KRW7,590 (US$6.8, up 1.1% from April) and KRW12,480 (up 1.8%) in South Korea; US$17.20 (up 8.2%) and US$19 (down 1.0%) in the US; EUR6.20 (US$7.0, down 12.7%) and EUR8.30 (down 3.5%) in Europe.Philips 40W-equivalent LED light bulbs and Osram 60W-equivalent models for sale in the South Korea market had the highest average lumen-price ratios of 140.2lm/US$ and 240.5lm/US$ respectively in June. In terms of luminous efficiency, Mitsubishi Electric 40W-equivalent LED light bulbs and 60W-equivalent models available in the Japan market had the highest average levels of 95.5lm/W and 108.2lm/W respectively in June.

There have been diverse opinions about Cisco Systems' continued acquisition policy for over the past 25 years, with one of the opinion groups even arguing that the operating benefits derived from these acquisition projects have been limited, or less sucessful than expected. But for Cisco, this is simply not true.A close look at the acquisition projects and follow-up integration strategies implemented by Cisco over the past two years, clearly indicate that Cisco boasts the ability to synergize and transform target companies in its post-merger integration processes.Cisco has also justified its policy as saying that it has not been eyeing the growth potential of the core businesses of target companies, but rather the synergistic effects between the core technologies of target companies and Cisco's existing technologies and product lines.The takeovers of Meraki and Jasper Technologies mark two recently successful merger cases. The two companies have not only seen their original product lines expand significantly, but have also seen their products venture into other segments after integrating with Cisco's solutions.Cisco's established acquisition policy continued in the first half of 2017 with four deals reached in the six-month period, including the takeover of APP Dynamics for US$3.7 billion, as well as SD-WAN solution provider Viptela, data analyzing startup Saggezza and AI innovation startup MindMeld.Cisco's next-generation network solutions, including Intuitive Network and Intent Based Network, are mixed with a number of technologies such as machine learning, AI, data analysis and security protection, some of which were secured via acquisition projects in recent years.Cisco's collaboration solution, Spark, is being rolled out following the acquisition and integration with Worklife. Additionally, a portion of MindMeld technology will be also incorporated into Spark in order to expand the application scope of Cisco's mobile team communication tools.To sum up, Susie Wee, vice president and CTO of DevNet Innovations at Cisco Systems, said Cisco's existing teams are always excited and eagerly looking forward to working together with new corporate teams brought into Cisco through mergers, particularly as the market is facing increasing demand for innovations to support ever diversifying technology applications, which require new technologies and resources.