外文翻譯--數(shù)字信號(hào)處理器重新采納多核架構(gòu)

1、<p>　　http://www.eetasia.com/</p><p>　　Digital Signal Processing Chips reembrace multicore architecture</p><p>　　Posted: 03 Nov 2008 </p><p>　　Adding cores to a processor to ga

2、in a performance boost, while lowering power demand, has become standard practice in the computing and embedded processor industries. While a similar evolution seems inevitable for all types of high-performance processin

3、g, prior experience has made DSP vendors more selective in applying the multicore approach. DSPs are now beginning to reembrace multicore architectures, but mainly for specific applications possessing well-partitioned pr

4、ocessing tasks. </p><p>　　Perform partitioning</p><p>　　A DSP application often comprises only a few highly complex tasks, and system performance improvements depend on hastening task execution,

5、 not simply running more tasks. Instead of partitioning at the task level, this system often requires partitioning at the algorithm level. The overall task, such as compressing a video stream, must be broken into steps t

6、hat can run in parallel on separate cores. The task scheduler or OS cannot perform such partitioning; it must come during the software design.</p><p>　　Homogenous vs. heterogeneous</p><p>　　This

7、 doesn't mean that the multicore approach hasn't been tried with DSPs. PicoChip has long had its picoArray architecture that puts multiple, identical cores together for high-performance DSP. In most cases, howeve

8、r, multicore design offerings with DSP had not been homogeneous—having multiple copies of the same core. Instead, they integrated a DSP core with a RISC CPU core. Such heterogeneous DSPs, for instance, have been part of

9、multicore processor designs for a number of years in the handset</p><p>　　One exception was the Blackfin BF561 dual-core DSP from Analog Devices Inc. The device used cores that were designed to handle both t

10、ypes of tasks well, so there was no need to partition along task lines. Instead, developers could assign tasks according to their preference to balance the load among the cores. Most developers at the time, however, were

11、 inexperienced at partitioning software and automated tool support was lacking, so the homogeneous multicore DSP was not quickly adopted. </p><p>　　"The BF561 was an early entrant to the field," sa

12、id David Katz, ADI's Black fin applications manager, "and it was ahead of its time. It has taken a while for people to learn partitioning." </p><p>　　He noted that it has not introduced a homog

13、eneous multicore DSP design since the BF561, although "multicore is an important part of our roadmap strategy." </p><p>　　Other DSP vendors also view multicore as an inevitable trend for DSPs, for

14、the same reason it was adopted in computing: higher performance at lower power. What is now making homogeneous multicore DSP chip designs appear once more is a shift in the performance increases some systems require. In

15、some applications, the need for performance is moving from performing a single task faster to performing more tasks. This shift is simplifying DSP task partitioning, making it more like that of other embed</p><

16、;p>　　Processing change</p><p>　　With the rising demand for VoIP and video over IP, media processing has become one such shifting application. A media gateway design, for instance, must provide a number of

17、 voice, audio and video codecs, and handle multiple independent channels. This application structure is easily partitioned into independent tasks, making it a good fit for multicore DSP designs. </p><p>　　Th

18、e Octasic Vocallo addresses this application space.</p><p>　　Vocallo has 15 identical DSP cores, giving designers considerable flexibility in creating parallel and pipelined architectures that strike the rig

19、ht balance of channel capacity for different installations. </p><p>　　Another multicore DSP for communications is Texas Instruments Inc.'s TNETV3020 for high-density core networks. "What we are doin

20、g now is application-specific multiprocessing," said Ray Simar, manager of multicore solutions at TI. "In infrastructure applications you're doing the same tasks, so designs can gravitate to multiple copies

21、 of the same processor." </p><p>　　The TNETV3020 has six DSP cores, along with a switch fabric and a variety of serial I/O channels, allowing designers to configure the design for tasks such as channel

22、format conversion. </p><p>　　Communications is not the only DSP application changing its character. Audio processing has also grown to require high-performance handling of multiple tasks that are needed for

23、simple partitioning among multiple cores. According to Sujata Neidig, audio DSP product manager at Freescale Semiconductor Inc., the advent of high definition, Dolby and Blu-ray audio algorithms has increased performance

24、 demands on audio DSPs as much as five times, with rising complexity, data rates and numbers of channe</p><p>　　Recycling codes</p><p>　　Freescale's Symphony DSP56724 and DSP56725 DSPs offer

25、 a dual-core architecture that allows developers to split the processing burden while reusing their existing code. Multicore DSPs that target video or mixed audio and video processing are also appearing. Samples are the

26、CT3616 from Cradle Technologies Inc., the Voyageur from Gennum Corp. and multicore DSPs for audio from Cirrus Logic Inc. </p><p>　　The trend toward multicore DSP chip designs may ultimately move the partitio

27、ning task out of the developers' hands into the chip vendors'. An example is the PC302, recently introduced by PicoChip. The company used its general-purpose picoArray architecture to create a device that impleme

28、nts a complete femtocell access point on a single chip. The company handled all the partitioning and put the core system software in on-chip memory, limiting developers' efforts to add custom functionality. </p>

29、;<p>　　Such specialized devices may be the near-term future for multicore DSP designs, but long term the multicore approach will envelope general-purpose DSP designs as well. "As we look down the road we can

30、see that multicore for DSP is not a one-trick pony," said Simar. "It will become more common." </p><p>　　Along with that shift will come a growing need for developers to learn how to partition

31、 their designs to effectively utilize homogenous multicore DSPs. "A number of people want a compiler to handle this," said Simar, "but that's not going to happen for a while. We will need to think abou

32、t things differently in order to apply these devices." </p><p>　　Richard Quinnell</p><p><b>　　EE Times</b></p><p>　　數(shù)字信號(hào)處理器重新采納多核架構(gòu)</p><p>　　為了提升性能同時(shí)降低

33、功率要求，在處理器中增加內(nèi)核已經(jīng)成為計(jì)算和嵌入式處理器產(chǎn)業(yè)的標(biāo)準(zhǔn)作法。雖然同樣的演變對(duì)各種高性能處理來(lái)說(shuō)似乎是不可避免的，以往的經(jīng)驗(yàn)使得數(shù)字信號(hào)處理器(DSP)供應(yīng)商更愿意選擇多核方式。DSP正在開(kāi)始重新采納多核架構(gòu)，不過(guò)主要是針對(duì)那些處理任務(wù)可以得到妥善劃分的特殊應(yīng)用。 </p><p><b>　　操作分解</b></p><p>　　DSP應(yīng)用通常只包含少量高度

34、復(fù)雜的任務(wù)，系統(tǒng)性能的提升依賴于加快任務(wù)執(zhí)行速度，而不是簡(jiǎn)單地運(yùn)行更多的任務(wù)。與任務(wù)級(jí)劃分不同，DSP系統(tǒng)通常要求在算法級(jí)劃分任務(wù)。整個(gè)任務(wù)，比如壓縮一個(gè)視頻流，必須被分解成可以在單獨(dú)內(nèi)核上并行運(yùn)行的多個(gè)步驟。任務(wù)調(diào)度器或操作系統(tǒng)無(wú)法完成這種劃分，這種劃分必須在軟件設(shè)計(jì)過(guò)程完成。許多DSP應(yīng)用開(kāi)發(fā)人員正因?yàn)樗惴▌澐掷щy而回避多內(nèi)核方法。也有一些任務(wù)(如加密)不支持并行運(yùn)算。 </p><p><b>　

35、　比較相同與不同</b></p><p>　　這并不是說(shuō)多核方法未嘗試與DSP親近過(guò)。PicoChip公司很早前推出的picoArray架構(gòu)就整合了多個(gè)相同的內(nèi)核來(lái)支持高性能DSP。然而在大多數(shù)情況下，采用DSP的多核設(shè)計(jì)中的內(nèi)核不是同一類的。相反，它們采用一個(gè)DSP內(nèi)核和一個(gè)RISC CPU內(nèi)核。多年來(lái)這種異質(zhì)DSP一直被用于蜂窩電話和通信行業(yè)中的多核處理器設(shè)計(jì)。這些處理器的目標(biāo)應(yīng)用可以被很好地劃分

36、為適合DSP的信號(hào)處理任務(wù)和適合RISC CPU的控制任務(wù)，從而使得劃分相當(dāng)簡(jiǎn)單。 </p><p>　　一個(gè)例外是ADI的Blackfin BF561雙核DSP。該器件使用的內(nèi)核可以很好地執(zhí)行兩種任務(wù)，因此無(wú)需對(duì)任務(wù)組進(jìn)行劃分。相反，開(kāi)發(fā)人員可能隨意分配任務(wù)來(lái)平衡內(nèi)核之間的負(fù)載。不過(guò)，大多數(shù)開(kāi)發(fā)人員缺少劃分軟件的經(jīng)驗(yàn)，而自動(dòng)化工具支持也非常缺乏，因此同類多核DSP無(wú)法得到迅速普及。 </p>&l

37、t;p>　　“BF561是較早進(jìn)入這一領(lǐng)域的產(chǎn)品?！盇DI公司Blackfin應(yīng)用經(jīng)理David Katz表示，“它是領(lǐng)先于時(shí)代的產(chǎn)品。我們需要給開(kāi)發(fā)員一定的時(shí)間來(lái)學(xué)習(xí)劃分技術(shù)?！?</p><p>　　Katz指出，ADI公司在BF561后就沒(méi)有推出過(guò)同類多核DSP設(shè)計(jì)，雖然“多內(nèi)核是我們發(fā)展策略中的重要組成部分”。 </p><p>　　其它DSP供應(yīng)商也將多核看作是DSP不可避

38、免的發(fā)展趨勢(shì)，這與它被計(jì)算應(yīng)用廣泛采納的理由是相同的：以更低的功率提供更高的性能。再次出現(xiàn)同類多核DSP芯片設(shè)計(jì)的驅(qū)動(dòng)因素，是一些系統(tǒng)要求的性能越來(lái)越高。在一些關(guān)鍵應(yīng)用中，對(duì)性能的要求正在從更快地執(zhí)行單一任務(wù)轉(zhuǎn)變?yōu)閳?zhí)行更多的任務(wù)。這種變化將簡(jiǎn)化DSP的任務(wù)劃分，使得它更像是其它嵌入式應(yīng)用的任務(wù)劃分，而DSP供應(yīng)商也在抓住這個(gè)難得的機(jī)會(huì)積極開(kāi)發(fā)產(chǎn)品。 </p><p><b>　　處理器的改變</b

39、></p><p>　　隨著對(duì)語(yǔ)音和VoIP需求的增長(zhǎng)，媒體處理已經(jīng)成為一種符合上述變化趨勢(shì)的應(yīng)用。例如，媒體網(wǎng)關(guān)設(shè)計(jì)必須提供各種語(yǔ)音、音頻和視頻編解碼功能，并處理多個(gè)獨(dú)立的通道。這種應(yīng)用結(jié)構(gòu)很容易被劃分成獨(dú)立的任務(wù)，因而非常適合多核DSP設(shè)計(jì)。 </p><p>　　Octasic公司的Vocallo就是滿足這一應(yīng)用領(lǐng)域要求的最新產(chǎn)品。 </p><p>　

40、　Vocallo有15個(gè)相同的DSP內(nèi)核，因此設(shè)計(jì)師具有相當(dāng)大的靈活性來(lái)創(chuàng)建并行和管線架構(gòu)，從而在不同工程的通道容量和性能間取得最佳平衡。 </p><p>　　另外一個(gè)最近推出的用于通信的多核DSP來(lái)自TI公司，型號(hào)是TNETV3020，主要用于高密度核心網(wǎng)絡(luò)。“我們目前正在做的工作是特殊應(yīng)用多處理?！盩I公司多核解決方案部經(jīng)理Ray Simar表示，“在基礎(chǔ)架構(gòu)應(yīng)用中，我們正在執(zhí)行大量相同的任務(wù)，因此設(shè)計(jì)可以

41、轉(zhuǎn)向多個(gè)相同的處理器?！?</p><p>　　TNETV3020采用了6個(gè)DSP內(nèi)核、1個(gè)開(kāi)關(guān)矩陣和多種串行I/O通道，允許設(shè)計(jì)師針對(duì)通道格式轉(zhuǎn)換等任務(wù)對(duì)設(shè)計(jì)進(jìn)行配置。 </p><p>　　同樣，音頻處理也需要對(duì)多任務(wù)實(shí)現(xiàn)高性能處理，這些任務(wù)適合在多個(gè)內(nèi)核間作簡(jiǎn)單的劃分。飛思卡爾半導(dǎo)體公司音頻DSP產(chǎn)品經(jīng)理Sujata Neidig指出，高清、杜比和藍(lán)光音頻算法的發(fā)明對(duì)音頻DSP的性能

42、要求提高了5倍，而且復(fù)雜性、數(shù)據(jù)速率和通道數(shù)量也在不斷增加。另外，Neidig表示，像自動(dòng)音量控制等更多功能也在向音頻處理集中。</p><p>　　許多DSP應(yīng)用開(kāi)發(fā)人員由于算法劃分困難而回避多核方法。</p><p><b>　　代碼復(fù)用 </b></p><p>　　飛思卡爾公司新款Symphony DSP56724和DSP56725 D

43、SP采用了一種雙核架構(gòu)，允許開(kāi)發(fā)人員分割處理任務(wù)，同時(shí)復(fù)用現(xiàn)有的代碼。用于視頻或混合音頻與視頻處理的多核DSP也已出現(xiàn)，例如Cradle Technologies公司的CT3616，Gennum公司的Voyageur以及Cirrus Logic公司的音頻用多核DSP。 </p><p>　　向多核DSP芯片設(shè)計(jì)發(fā)展的趨勢(shì)，最終會(huì)將劃分任務(wù)的工作從開(kāi)發(fā)人員手中轉(zhuǎn)移到芯片廠商。PicoChip公司最近推出的PC302

44、就是一個(gè)很好的例證。該公司使用其通用的picoArray架構(gòu)成功的在單個(gè)芯片上實(shí)現(xiàn)了完整的毫微微蜂窩接入點(diǎn)。PicoChip公司完成了所有的劃分工作，并將內(nèi)核系統(tǒng)軟件下載進(jìn)了片上存儲(chǔ)器，因此開(kāi)發(fā)人員只需增加自定義的功能。 </p><p>　　這種特殊器件可能是多核DSP設(shè)計(jì)的近期目標(biāo)，但長(zhǎng)期來(lái)看多核方法還將包含通用的DSP設(shè)計(jì)。“展望未來(lái)，我們可以發(fā)現(xiàn)，針對(duì)DSP的多核技術(shù)不會(huì)只是小把戲，”TI公司的Simar

45、表示，“它將變得越來(lái)越普及?！?</p><p>　　隨之而來(lái)的是，開(kāi)發(fā)人員需要學(xué)會(huì)如何劃分設(shè)計(jì)以便高效地使用同類多核DSP?！霸S多人希望編譯器能完成分割任務(wù)?！盨imar表示，“但這不是一蹴而就的事。為了用好這些器件，我們需要學(xué)會(huì)多角度的思考?！?lt;/p><p><b>　　專業(yè)詞匯：</b></p><p>　　chips ： 芯片<

46、/p><p><b>　　core ： 內(nèi)核</b></p><p>　　multicore：多芯的 、多核心</p><p>　　processor　： 處理器,處理機(jī)</p><p>　　power ：功率</p><p>　　DSP ： 數(shù)字信號(hào)處理器</p><p>

47、　　Blu-ray：Blu-ray Disc 藍(lán)色激光</p><p>　　video stream 視頻流</p><p>　　task scheduler 任務(wù)調(diào)度程序</p><p>　　dual-core : 雙核</p><p>　　automated tool：自動(dòng)化工具</p><p>　　VoIP

48、 ：網(wǎng)絡(luò)電話 、 IP電話</p><p>　　media processing：媒體處理</p><p><b>　　audio：音頻</b></p><p>　　codecs：碼器 編解碼</p><p>　　pipelined architecture　 廉線結(jié)構(gòu)</p><p>　　s

49、witch fabric：交換結(jié)構(gòu) 、 交換核心 、 交換光纖</p><p>　　serial I / O：串行輸入/輸出</p><p>　　format conversion：格式轉(zhuǎn)換 、 格式轉(zhuǎn)換 </p><p>　　high definition：高清晰度 、 高清 、 高精度</p><p>　　Dolby：杜比，杜比功能 、

50、杜比降噪 、 降低噪聲電路、道爾貝、杜比 、</p><p>　　不僅具有高清晰畫(huà)質(zhì)、高保真音質(zhì)</p><p>　　Blu-ray：Blu-ray Disc 藍(lán)色激光</p><p>　　automatic volume control：自動(dòng)音量控制</p><p><b>　　code：代碼 </b></p&

51、gt;<p>　　mixed audio：混合音頻</p><p>　　femtocell：基站 、 家用型</p><p>　　compiler：編譯程序 、 編譯器</p><p>　　EDA/IC Design</p><p>　　Adobe-ARM team up boosts ARM-powered devices&l

52、t;/p><p>　　Posted: 20 Nov 2008 </p><p>　　Adobe Systems Inc. and ARM have announced a technology collaboration to optimize and enable Adobe Flash Player 10 and Adobe AIR for ARM-powered devices, ran

53、ging from mobile phones to set-top boxes, mobile Internet devices, TVs, automotive platforms, personal media players and other mobile computing devices. The collaboration is expected to accelerate mobile graphics and vid

54、eo capabilities on ARM platforms to bring rich Internet applications and Web services to mobile devices and consumer electro</p><p>　　The joint technology optimization is targeted for the ARMv6 and ARMv7arch

55、itectures used in the ARM11 family and the Cortex-A series of processors and is expected to be available in 2H 09. The partnership stems from the Open Screen Project, a broad Adobe sponsored initiative of industry leader

56、s, including ARM to deliver a consistent runtime environment across multiple devices by taking advantage of Adobe Flash Player and, in the future, Adobe AIR. The initiative is set to address the challenges o</p>&

57、lt;p>　　Gary Kovacs, general manager and VP, mobile and devices at Adobe said, "We are pleased to work with ARM and the other industry leaders in the Open Screen Project, to make browsing and applications as rich

58、and powerful in mobile as they are on the desktop." </p><p>　　"ARM believes this partnership will develop optimized Adobe Flash and AIR implementations that will run on billions of devices from our

59、 Partners such as pocket-sized mobile devices, mobile computing platforms, set-top boxes, digital TVs and automotive infotainment," said Ian Drew, VP marketing at ARM. </p><p>　　Broad support from indus

60、try</p><p>　　"Mobile users deserve a richer, faster Web experience with extended battery life. ARM technology, together with Broadcom's low power Video Core multimedia processor will turbo charge Ad

61、obe Flash Player 10 and Adobe AIR performance using OpenGL ES 2.0 and other open standards," said Mark Casey, VP and general manager of Broadcom's mobile multimedia business unit. "The Adobe ARM initiative

62、is an exciting opportunity for Broadcom to enhance the consumer experience on mobile and consumer devices." </p><p>　　"The ARM architecture is poised for explosive growth among mobile Internet-cent

63、ric devices and into the consumer net book market," said Glen Burchers, director of global consumer marketing for Freescale, "The addition of Flash Player 10 and Adobe AIR optimized to ARM with support for HW a

64、cceleration is a significant move toward this end. Freescale intends to fully support this exciting combination." </p><p>　　"Nvidia is working with ARM and Adobe to ensure Adobe Flash takes full ad

65、vantage of Nvidia Tegra computer-on-a-chip solutions through open standards such as OpenGL ES 2.0," said Michael Ray field, general manager of NVIDIA's mobile business. "ARM CPU technology, tightly integrat

66、ed with NVIDIA's ultra-low-power GeForce GPUs and media acceleration, enhances the ability of Adobe Flash to provide the full Web experience and compelling user interfaces in the palm of your hand." </p>

67、<p>　　"Samsung looks forward to having Flash Player 10 and AIR fully supported on our range of optimized ARM technology-based Application Processors and SoC products," said Yiwan Wong, VP of marketing, Sam

68、sung Electronics' System LSI Division. "This initiative will enable Flash Player 10 contents to be accessible by any ARM technology-based consumer devices with a screen and connectivity, and at very low power co

69、nsumption, offering consumers the ultimate mobile Internet experience beyond what is avail</p><p>　　"Texas Instruments is excited to support Flash Player 10 which will deliver advanced features such as

70、no-compromise Internet experience on OMAP 3 processors," said Raj Talluri, general manager, Texas Instruments. "The addition of Flash Player 10 to TI's portfolio builds on its history of creating platforms

71、that provide exceptional performance running Flash technology. Leveraging the OMAP 3 platform's SoC architecture and foundation software enables TI's customers to develop a variety of differen</p><p>

72、;　　The agreement will enable ARM and Adobe to deliver an optimized Adobe Flash Player 10 for the ARM architecture as well as industry-standard API support for GPUs and hardware accelerators. The collaboration is also exp

73、ected to lower power consumption for mobile devices running Flash Player 10 and AIR content. </p><p>　　Adobe Flash Player 10 for ARMv6 and ARMv7 architecture-based hardware is expected to be available royalt

74、y-free to Partners in the Open Screen Project. Flash Player 10 for ARM processor-based devices will be made available to OEMs by Adobe.</p><p><b>　　EAM/IC設(shè)計(jì)</b></p><p>　　ADOBE和ARM一起提

75、高了ARM的性能</p><p>　　Adobe Systems Incorporated 和ARM公司共同宣布了一項(xiàng)技術(shù)合作協(xié)議，將對(duì)Adobe Flash Player 10和Adobe AIR技術(shù)進(jìn)行優(yōu)化，使其能夠應(yīng)用于基于ARM的產(chǎn)品，包括從手機(jī)到機(jī)頂盒、移動(dòng)互聯(lián)網(wǎng)設(shè)備、電視機(jī)、汽車電子平臺(tái)、個(gè)人媒體播放器和其它移動(dòng)計(jì)算設(shè)備。這一合作的目的是加速ARM平臺(tái)上的移動(dòng)圖像和視頻處理能力，從而為全球范圍內(nèi)的移動(dòng)

76、設(shè)備和消費(fèi)電子產(chǎn)品帶去豐富的Internet應(yīng)用和Web服務(wù)。 </p><p>　　這一合作技術(shù)優(yōu)化針對(duì)用于ARM11系列和Cortex-A系列處理器的ARMv6和ARMv7架構(gòu)，計(jì)劃將于2009年下半年完成。該合作源于Open Screen Project，這是一項(xiàng)廣泛的、由Adobe贊助的、涉及包括ARM在內(nèi)的眾多行業(yè)領(lǐng)先者的倡議，旨在通過(guò)充分利用Adobe Flash Player以及（未來(lái)的）Adobe

77、 AIR，提供一個(gè)跨設(shè)備的一致的運(yùn)行環(huán)境。該倡議的目的是應(yīng)對(duì)在眾多不同的屏幕上進(jìn)行網(wǎng)頁(yè)瀏覽所帶來(lái)的挑戰(zhàn)，同時(shí)消除在不同屏幕間無(wú)縫地發(fā)布內(nèi)容和應(yīng)用的阻礙。</p><p>　　Adobe移動(dòng)和設(shè)備總經(jīng)理兼副總裁Gary Kovacs表示：“我們非常高興與ARM和其它行業(yè)領(lǐng)導(dǎo)者在Open Screen Project中的合作，這將為消費(fèi)者帶來(lái)與臺(tái)式電腦一樣豐富、強(qiáng)大的瀏覽和應(yīng)用體驗(yàn)。” </p><

78、;p>　　ARM市場(chǎng)營(yíng)銷副總裁Ian Drew表示：“ARM相信這一合作將開(kāi)發(fā)出優(yōu)化的Adobe Flash和AIR技術(shù)，使其能夠在數(shù)以十億計(jì)的由我們的合作伙伴所提供的設(shè)備上運(yùn)行。這些設(shè)備包括口袋大小的移動(dòng)設(shè)備、移動(dòng)計(jì)算平臺(tái)、機(jī)頂盒、數(shù)字電視和汽車信息娛樂(lè)系統(tǒng)。</p><p>　　業(yè)界領(lǐng)先企業(yè)的廣泛支持 </p><p>　　Broadcom移動(dòng)多媒體業(yè)務(wù)部副總裁兼總經(jīng)理Mark

79、Casey表示：“移動(dòng)用戶應(yīng)該得到更豐富、更快的Web體驗(yàn)和更長(zhǎng)的電池壽命。ARM技術(shù)同Broadcom的低功耗VideoCore多媒體處理器一起，將推動(dòng)使用OpenGL ES 2.0和其它開(kāi)放標(biāo)準(zhǔn)的Adobe Flash Player 10和Adobe AIR的性能。Adobe與ARM合作對(duì)于Broadcom提供移動(dòng)和消費(fèi)電子設(shè)備上的消費(fèi)者體驗(yàn)而言是一個(gè)激動(dòng)人心的機(jī)遇?！?</p><p>　　NVIDIA移動(dòng)

80、業(yè)務(wù)總經(jīng)理Michael Rayfield 表示：“NVIDIA正在同ARM和Adobe合作，通過(guò)OpenGL ES 2.0等開(kāi)放標(biāo)準(zhǔn)，重復(fù)利用NVIDIA Tegra片上計(jì)算機(jī)解決方案（computer-on-a-chip solution）在ARM架構(gòu)上實(shí)現(xiàn)Adobe Flash。將ARM CPU技術(shù)與NVIDIA的超低功耗GeForce GPU和媒體加速技術(shù)整合在一起，將提高Adobe Flash的能力，為消費(fèi)者手中的掌上設(shè)備提供

81、完整的Web體驗(yàn)和出眾的用戶界面?！?</p><p>　　三星電子系統(tǒng)LSI部門市場(chǎng)營(yíng)銷副總裁Yiwan Wong表示：“三星期待著能夠在我們眾多優(yōu)化的、基于ARM技術(shù)的應(yīng)用處理器和片上系統(tǒng)產(chǎn)品上得到Flash Player 10和AIR的完整支持。這一合作將使得Flash Player 10內(nèi)容能夠通過(guò)任何具有屏幕和連接功能的、基于ARM技術(shù)的消費(fèi)產(chǎn)品獲得，同時(shí)功耗非常低，使得消費(fèi)者能夠突破當(dāng)今標(biāo)準(zhǔn)PC平臺(tái)，

82、而獲得真正的移動(dòng)Internet體驗(yàn)?！?</p><p>　　德州儀器總經(jīng)理Raj Talluri表示：“德州儀器將非常高興地支持Flash Player 10，它將在OMAP3處理器上提供先進(jìn)的功能，例如無(wú)妥協(xié)的Internet體驗(yàn)。在德州儀器多年的創(chuàng)建能夠提供卓越的Flash技術(shù)運(yùn)行性能的平臺(tái)的基礎(chǔ)之上，我們?cè)谖覀兊漠a(chǎn)品中加入了對(duì)Flash Player 10的支持。通過(guò)利用OMAP3平臺(tái)的SoC架構(gòu)以及基

83、礎(chǔ)軟件，德州儀器的客戶將能夠開(kāi)發(fā)眾多差異化的產(chǎn)品，滿足不同市場(chǎng)的需求，例如智能手機(jī)、移動(dòng)互聯(lián)網(wǎng)設(shè)備、網(wǎng)絡(luò)筆記本和其它便攜式消費(fèi)電子設(shè)備，能夠提供與筆記本電腦相似的功能，但無(wú)需在功耗、性能或成本方面做出犧牲?！?</p><p>　　這一協(xié)議將使得ARM和Adobe能夠?yàn)锳RM架構(gòu)提供優(yōu)化的Adobe Flash Player 10，并為GPU和硬件加速器提供行業(yè)標(biāo)準(zhǔn)API支持。這一合作還將降低運(yùn)行Flash Pl

84、ayer 10和AIR內(nèi)容的移動(dòng)設(shè)備的功耗。 </p><p>　　用于基于ARMv7架構(gòu)的硬件的Adobe Flash Player 10預(yù)計(jì)將向Open Screen Project中的合作伙伴以免版稅的形式提供。用于基于ARM處理器的設(shè)備的Flash Player 10將由Adobe向OEM廠商提供。</p><p><b>　　專業(yè)詞匯：</b></p&

85、gt;<p>　　EAD：放電涂覆處理</p><p><b>　　IC：集成電路</b></p><p><b>　　arm.：電樞</b></p><p>　　powered：上電狀態(tài) 、 有動(dòng)力的</p><p>　　set-top box：機(jī)頂盒</p><

86、p>　　mobile Internet device：移動(dòng)互聯(lián)網(wǎng)設(shè)備</p><p>　　automotive：汽車制造的 、 汽車的 、 自動(dòng)的</p><p>　　media player：媒體播放器</p><p>　　mobile computing device：移動(dòng)計(jì)算設(shè)備</p><p>　　mobile graphics

87、：移動(dòng)圖形</p><p>　　architectures：架構(gòu)</p><p>　　mobile device：移動(dòng)裝置</p><p>　　Digital TV：移動(dòng)數(shù)字電視</p><p><b>　　power：功率</b></p><p>　　multimedia processor：處

88、理器 、 多媒體處理器</p><p>　　turbo charge：渦輪增壓</p><p>　　PC: (personal computer)個(gè)人計(jì)算機(jī)</p><p><b>　　SoC：片上系統(tǒng)</b></p><p>　　smart phones：智能手機(jī) 、智慧型手機(jī) 、智能電話 、 智慧型電話</p&

89、gt;<p>　　Flash Player：播放器 、 多媒體播放機(jī)</p><p><b>　　GPU：圖形處理器</b></p><p>　　GPUs：處理器 、 單元 、 圖形芯片</p><p>　　hardware accelerator：硬件加速器</p><p>　　API：應(yīng)用程序編程接口&