Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: rsi_boy who wrote (14047)	10/15/2000 9:13:30 PM
From: dunnerjohn	Respond to of 275872

I want to echo the opinion expressed by rsi boy regarding voice recognition as an application that will drive the demand for ever faster and more powerful PCs. However, I was surprised that rsi boy has been experiencing such a slowdown in his system performance when running Dragon NaturallySpeaking. I would therefore like to recount my own experience for his benefit and that of anyone else who is interested in voice recognition or its potential for increasing the demand for ever faster computers. I have been running a new Athlon Thunderbird 900 MHz system under Windows 2000 for about eight weeks now and I have been delighted with its performance and stability. I have not had a single crash or freeze. I was previously running a Pentium II 400 MHz system under Windows 98 and was experiencing lots of crashes and freezes. In spite of the fact that I was running the Pentium II 400 MHz system with 384 megabytes of RAM I still found that it was intolerably slow and simply didn't have the required horsepower to run Dragon NaturallySpeaking effectively, particularly if I wanted to have multiple programs open at the same time. I have been attributing most of the improvement in stability to Windows 2000 but it may be that the Athlon has been making some contribution as well. The system specifications are set out below. You will see that I have been attempting to optimize the system to get the best possible performance of Dragon NaturallySpeaking voice recognition dictation software. The new system has made a huge difference in speed and performance. With my new Athlon Thunderbird 900 MHz system I am able to achieve nearly real-time performance in spite of the fact that I normally have several programs open simultaneously with Dragon NaturallySpeaking, including several instances of Internet Explorer, Outlook, a lawyer's PIM known as Amicus Attorney and either Word 97 or WordPerfect. As I said at the beginning I see voice recognition as one application that will drive the demand for ever faster processors and memory. I can now achieve about 98 percent accuracy when I am careful to speak clearly and distinctly to Dragon and by times I can achieve as much as 99 percent accuracy. Until now I believe that Dragon and other makers of voice recognition software have been designing their products to be able to be run optimally on a system such as a Pentium III 500 MHz. Although I don't have any particular technical expertise to base this opinion on, I would bet that if they knew that they would be able to count on a large market of potential purchasers who have much faster systems they would be able to implement improvements to the software to take account of the increased computing power. I believe that there will be a tremendous market for a high-end voice recognition dictation system which is created on the assumption that it will be used on a very fast system such as will be possible in the next year or so with a Mustang running at 1.5 GHz and fast memory or even better with a dual processor Mustang system or even a Sledgehammer system. With a top end system and dictation software designed to take advantage of such a system's capabilities I believe it will be possible to achieve 99.5 percent and perhaps better accuracy in real-time. This should be achievable for a cost for hardware and software in the range of $5000 to $7500 which is a drop in the bucket for a doctor or lawyer compared with the cost of employing a person to transcribe dictation. If anyone is interested in exploring voice recognition I would be glad to provide the benefit of my experience. John Dunn My current system specifications: Athlon Thunderbird 900 MHz CPU Asus A7V Socket A main board with Promise ATA 100 IDE controller 384 MB PCI 100 RAM Memory Twin Maxtor 30.2 GB ATA 100 2MB 7200 rpm hard disks. The second hard drive is being used solely for the Windows swap file which is currently set for a fixed size of 100 MB minimum and maximum. ATI Expert 2000 32 MB AGP Video Card Creative Labs Sound Blaster Live! Platinum (used for playback only) Buddy II modified Telex USB device with Philips chip from Parmod Gandhi (InSync Software) used for recording speech Sennheiser MD 431 II hand-held microphone Pioneer 16X CD/DVD ROM Enlite Mid-tower case with Antec 300 Watt PS Extra case fan Microsoft Windows 2000 Professional Dragon NaturallySpeaking (DNS) Professional version 4.1 Preferred word processor: WordPerfect 9 (also use Word 97, however)