With double-digit attendee growth to the 60,000 level, MWC'11 has to be labeled as a big success. Moreover, the new products introduced and the overall optimism exhibited bodes well for continued wireless market growth in 2011 and beyond. Of course, many LTE-related products were introduced along with a myriad of new software and service offerings. I'm sure that you have seen many of the news releases, so I'll only highlight a few things from MWC and some beyond MWC that I think will be of interest to you.
Icera Demo's Its First Smartphone Socket
In my February newsletter I mentioned Icera's new cellphone chipset based on its latest baseband and RF technology. At MWC, the company was displaying an HSPA+ i-Mobile smartphone based on Icera's new chipset. The 21 Mbps Android device is said to be software upgradeable to 28 Mbps. It was real, and I was able to call my wife in the U.S. with it and the voice quality was excellent. Although the company's cellular chips are in 45 (mostly USB) products, and the company has shipped millions, this is the first handset socket for the company.
Icera has implemented its baseband products based on its DXP software baseband, employing both vector and scalar versions of the DSP core, employing no hardware accelerators...which is said to simplify software upgrades to LTE (a capability they demonstrated at MWC'10). They feel that their baseband design gives them a "massive" cost advantage over Qualcomm, Samsung and Infineon. As the only independent cellular modem company, we fully expect them to be acquired within the next 12 months (maybe by Samsung, who also bid against Intel for Infineon).
Maxim Surprises with first MWC Appearance
Maxim Integrated Products, a powerhouse in analog and mixed-signal chips, had a big booth in the premium Hall 8, directly across from Texas Instruments' booth (making a market positioning statement?). Maxim claims to be the #1 supplier of power management chips (with TI as their main competitor, followed by Dialog Semiconductor). We've seen Maxim power management chips paired with Qualcomm cellphone chips, so they have legs. Maxim also claims to be #1 in touch screen controllers (but Atmel makes the same claim, but it's clear that they are a significant player). Clearly, the wireless market is a main thrust for the resurgent chip supplier.
MIPS Finally in Cellphones
MIPS Technologies Inc. has been best known as dominating the segment of the RISC market that plugs in the wall, vs. portable products, the market segment dominated by ARM. Ingenic Semiconductor, a Beijing-based client with a MIPS32 architectural license, has fielded an application processor chip for a $100 Android smartphone (paired with a Spreadtrum modem) for the China market. Moreover, new LTE chip players like Sequans and Altair are employing MIPS cores in their products while both 4M Wireless and SySDSoft have demonstrated their L2+ LTE stacks on MIPS32 architecture.
Renesas Reveals LTE Modem at MWC
Renesas Mobile Corporation actually has two LTE modem solutions. Although the company did not provide many details, the first LTE modem (SP2531) is based on a Nokia-originated design which employs a hard-wired DSP augmented by 3 ARM cores (2 dedicated to the modem function). That modem has long been used to test and tune LTE networks. It supports TD-LTE and FDD-LTE cat 3 along with HSPA+ for data rates of up to 42Mbps (DL). It is paired with a 65nm multi-mode LTE transceiver and multi-mode multi-band power amplifier module. The second LTE modem (not announced at MWC) employs DoCoMo technology (presumably, a Fujitsu baseband design based on Tensilica DSP cores). Further details will have to wait for official product announcements.
Intel CID pushes Cloud Cellular Infrastructure
At MWC, Intel's Communications Infrastructure Division (CID) was pitching its Cloud computing infrastructure that it is developing in conjunction with China Mobile. The basic idea of the proposed centralized radio access network (CRAN) concept is for the RF subsystems to be atop the base station towers, with the digital (I/Q) signals passed via optical cable to the cloud computing center where baseband signal processing will take place for many base stations. Needless to say, this centralized baseband architecture is a radically different approach to current base station deployment, but ties in nicely with Ubidyne's active antenna array design. If the Intel approach proves to be feasible, I expect that it will be several years before it can garner general industry acceptance.
Nokia & Microsoft Marriage Observations
My first thought was two old ladies helping each other cross the street. Then it occurred to me that it's something that both companies have to do, even though there will be blood.
Actually, the Microsoft-Nokia deal makes sense to me. First of all, Nokia is in trouble with an overloaded R&D budget and aging software and Microsoft badly needs better traction for Windows Phone 7. Although Windows Phone 7 is late and badly trailing iOS and Android, the early reviews have been generally very good. But Windows Phone 7 is on very few handsets, and Nokia fills the expansion needs for MS. Nokia badly needs a non-Symbian O/S, and adopting Android would make them simply another me-too house. Since MeeGo has not been proven, I think their only quick alternatives were WebOS (now owned by HP) or Windows Phone 7. Microsoft's Bing browser would "go along for the ride." Bing is OK, but it could also keep Google, Opera, Safari and others from dominating the mobile browser market. The near-term chip beneficiary of this marriage is Qualcomm, since currently only their chipset supports the DirectX multimedia interface of Windows Phone 7. Other smartphone chipsets tend to use the OpenGL interface. Of course, those other vendors will eventually embrace DirectX. Finally, both Nokia and Microsoft have lots of resources to throw at the problem, and now both finally have realistic roadmaps for their respective flagship wireless products. But, there will be blood.
Not well promoted by Nokia, I use Nokia maps on my (jail-broken) Nokia E71x cellphone from AT&T and it works worldwide, without incurring data charges (unless you want the premium voice directions). I've used it extensively in the U.S., U.K., Spain, Germany, Austria, Switzerland, Italy, Turkey and (last week) in Jordan. I actually prefer the Nokia map presentation over the map display that I get on my $400 Garmin car GPS. BTW, my wife and son use iPhones, so I'm not wedded to Nokia.
Sandbridge Acquired by Wuxi DSP Technologies
Sandbridge Technologies Inc., a designer of cellular baseband chips has been acquired by Wuxi DSP Technologies, based in Wuxi (pronounced "wu-hsi") in Jiangsu Province, just West of Shanghai. The original Sandbridge office in the U.S. has become the company's R&D center, called Optimum Semiconductor Technologies: www.optimumsemi-tech.com . The company's SB3500 chip features the Sandblaster® DSP for execution of baseband operations in software, including the physical layer. It has a programmable RF interface, with the capability to capture raw data at 240 M samples/sec. It includes interfaces to LCD, keypad, USIM, SmartCard, Audio codec, IrDA, plus emerging 'critical' features such as add-on memory cards, camera interface, and USB.
As a side note, Wuxi is the city where most of China's original analog TV receiver chips were fabricated, back in the early '80s. Historically, the city is noted as the favored location for the Emperor's concubines, since the women there are said to be very beautiful (I didn't check this out, since it was only a train stop for my visit).
Smartphone Definition gets Fuzzy
In the past, smartphones were distinguished from the lesser "feature phones" by the fact that they employed a high-level operating system, like Symbian, Windows Mobile, iOS or, most recently, Android. And smartphones were clearly more expensive than feature phones that could provide at least limited multimedia capabilities. However, now with $100 (un-subsidized) phones running Android due out this year, the old definition based on the high-level O/S no longer seems appropriate. The net effect is that the feature phone market is disappearing while the smartphone market now addresses a wider spectrum of handsets. Moreover, the lowest-end smartphones are not 3G-capable. Clearly, the $100 smartphone (one is running at 400MHz) will not be able to reach the performance level (2x1.2GHz) of a $600 smartphone, so I solicit your opinions on new nomenclature for distinguishing "expensive" from "cheap" smartphones. And, like the poor, even cheaper budget phones will always be with us.
Icera Demo's Its First Smartphone Socket
In my February newsletter I mentioned Icera's new cellphone chipset based on its latest baseband and RF technology. At MWC, the company was displaying an HSPA+ i-Mobile smartphone based on Icera's new chipset. The 21 Mbps Android device is said to be software upgradeable to 28 Mbps. It was real, and I was able to call my wife in the U.S. with it and the voice quality was excellent. Although the company's cellular chips are in 45 (mostly USB) products, and the company has shipped millions, this is the first handset socket for the company.
Icera has implemented its baseband products based on its DXP software baseband, employing both vector and scalar versions of the DSP core, employing no hardware accelerators...which is said to simplify software upgrades to LTE (a capability they demonstrated at MWC'10). They feel that their baseband design gives them a "massive" cost advantage over Qualcomm, Samsung and Infineon. As the only independent cellular modem company, we fully expect them to be acquired within the next 12 months (maybe by Samsung, who also bid against Intel for Infineon).
Maxim Surprises with first MWC Appearance
Maxim Integrated Products, a powerhouse in analog and mixed-signal chips, had a big booth in the premium Hall 8, directly across from Texas Instruments' booth (making a market positioning statement?). Maxim claims to be the #1 supplier of power management chips (with TI as their main competitor, followed by Dialog Semiconductor). We've seen Maxim power management chips paired with Qualcomm cellphone chips, so they have legs. Maxim also claims to be #1 in touch screen controllers (but Atmel makes the same claim, but it's clear that they are a significant player). Clearly, the wireless market is a main thrust for the resurgent chip supplier.
MIPS Finally in Cellphones
MIPS Technologies Inc. has been best known as dominating the segment of the RISC market that plugs in the wall, vs. portable products, the market segment dominated by ARM. Ingenic Semiconductor, a Beijing-based client with a MIPS32 architectural license, has fielded an application processor chip for a $100 Android smartphone (paired with a Spreadtrum modem) for the China market. Moreover, new LTE chip players like Sequans and Altair are employing MIPS cores in their products while both 4M Wireless and SySDSoft have demonstrated their L2+ LTE stacks on MIPS32 architecture.
Renesas Reveals LTE Modem at MWC
Renesas Mobile Corporation actually has two LTE modem solutions. Although the company did not provide many details, the first LTE modem (SP2531) is based on a Nokia-originated design which employs a hard-wired DSP augmented by 3 ARM cores (2 dedicated to the modem function). That modem has long been used to test and tune LTE networks. It supports TD-LTE and FDD-LTE cat 3 along with HSPA+ for data rates of up to 42Mbps (DL). It is paired with a 65nm multi-mode LTE transceiver and multi-mode multi-band power amplifier module. The second LTE modem (not announced at MWC) employs DoCoMo technology (presumably, a Fujitsu baseband design based on Tensilica DSP cores). Further details will have to wait for official product announcements.
Intel CID pushes Cloud Cellular Infrastructure
At MWC, Intel's Communications Infrastructure Division (CID) was pitching its Cloud computing infrastructure that it is developing in conjunction with China Mobile. The basic idea of the proposed centralized radio access network (CRAN) concept is for the RF subsystems to be atop the base station towers, with the digital (I/Q) signals passed via optical cable to the cloud computing center where baseband signal processing will take place for many base stations. Needless to say, this centralized baseband architecture is a radically different approach to current base station deployment, but ties in nicely with Ubidyne's active antenna array design. If the Intel approach proves to be feasible, I expect that it will be several years before it can garner general industry acceptance.
Nokia & Microsoft Marriage Observations
My first thought was two old ladies helping each other cross the street. Then it occurred to me that it's something that both companies have to do, even though there will be blood.
Actually, the Microsoft-Nokia deal makes sense to me. First of all, Nokia is in trouble with an overloaded R&D budget and aging software and Microsoft badly needs better traction for Windows Phone 7. Although Windows Phone 7 is late and badly trailing iOS and Android, the early reviews have been generally very good. But Windows Phone 7 is on very few handsets, and Nokia fills the expansion needs for MS. Nokia badly needs a non-Symbian O/S, and adopting Android would make them simply another me-too house. Since MeeGo has not been proven, I think their only quick alternatives were WebOS (now owned by HP) or Windows Phone 7. Microsoft's Bing browser would "go along for the ride." Bing is OK, but it could also keep Google, Opera, Safari and others from dominating the mobile browser market. The near-term chip beneficiary of this marriage is Qualcomm, since currently only their chipset supports the DirectX multimedia interface of Windows Phone 7. Other smartphone chipsets tend to use the OpenGL interface. Of course, those other vendors will eventually embrace DirectX. Finally, both Nokia and Microsoft have lots of resources to throw at the problem, and now both finally have realistic roadmaps for their respective flagship wireless products. But, there will be blood.
Not well promoted by Nokia, I use Nokia maps on my (jail-broken) Nokia E71x cellphone from AT&T and it works worldwide, without incurring data charges (unless you want the premium voice directions). I've used it extensively in the U.S., U.K., Spain, Germany, Austria, Switzerland, Italy, Turkey and (last week) in Jordan. I actually prefer the Nokia map presentation over the map display that I get on my $400 Garmin car GPS. BTW, my wife and son use iPhones, so I'm not wedded to Nokia.
Sandbridge Acquired by Wuxi DSP Technologies
Sandbridge Technologies Inc., a designer of cellular baseband chips has been acquired by Wuxi DSP Technologies, based in Wuxi (pronounced "wu-hsi") in Jiangsu Province, just West of Shanghai. The original Sandbridge office in the U.S. has become the company's R&D center, called Optimum Semiconductor Technologies: www.optimumsemi-tech.com . The company's SB3500 chip features the Sandblaster® DSP for execution of baseband operations in software, including the physical layer. It has a programmable RF interface, with the capability to capture raw data at 240 M samples/sec. It includes interfaces to LCD, keypad, USIM, SmartCard, Audio codec, IrDA, plus emerging 'critical' features such as add-on memory cards, camera interface, and USB.
As a side note, Wuxi is the city where most of China's original analog TV receiver chips were fabricated, back in the early '80s. Historically, the city is noted as the favored location for the Emperor's concubines, since the women there are said to be very beautiful (I didn't check this out, since it was only a train stop for my visit).
Smartphone Definition gets Fuzzy
In the past, smartphones were distinguished from the lesser "feature phones" by the fact that they employed a high-level operating system, like Symbian, Windows Mobile, iOS or, most recently, Android. And smartphones were clearly more expensive than feature phones that could provide at least limited multimedia capabilities. However, now with $100 (un-subsidized) phones running Android due out this year, the old definition based on the high-level O/S no longer seems appropriate. The net effect is that the feature phone market is disappearing while the smartphone market now addresses a wider spectrum of handsets. Moreover, the lowest-end smartphones are not 3G-capable. Clearly, the $100 smartphone (one is running at 400MHz) will not be able to reach the performance level (2x1.2GHz) of a $600 smartphone, so I solicit your opinions on new nomenclature for distinguishing "expensive" from "cheap" smartphones. And, like the poor, even cheaper budget phones will always be with us.
