I-Appliance BBS The Official Source for Internet Appliance Upgrades and Mods BBS Main List | Sign In | Sign Up | Search | Help | Linux-Hacker.net Reply to Thread | Printer |

Home / Other I-Appliances / Gateway/AOL Connected Touch Pad Updated testing results and new website

Updated testing results and new website (modified 0 times) yellow1 Profile | Email Not a very happy camper...something's wrong here 128megs of RAM barely help. see my new website for details http://orchycam.hypermart.net - Choose Data and then My Guides/GWCP 05-06-2001 20:11:35

RE:Updated testing results and new website (modified 0 times) yellow1 Profile | Email I wonder if Win95 wouldn't be faster ! Guess the drivers might work too. 05-06-2001 20:59:28

RE:Updated testing results and new website (modified 0 times) SuperRob Profile | Email Yellow, UI is almost entirely dependent on VIDEO, and somewhat on CPU, not RAM. Adding RAM will never help drawspeed, unless video memory is shared. The video chip looks to be a real dog. Your speed issues are probably coming from badly optimized video drivers. However, I leave open the possibility that the Transmeta chip simply doesn't handle video instructions as well as other types of instructions. Anyway, I just want to make sure that when people are troubleshooting problems with these units, that they are starting in the right places. Adding RAM to solve video issues is the _last_ thing you should do. Tracking down better drivers should be first. 05-06-2001 21:45:08

RE:Updated testing results and new website (modified 0 times) Ricko73 Profile | Email At least Gateway/AOL figured this one out. Microsoft anything is a resource hog and is crash prone. When you say 128mb...how did you do that? Are you including the 32mb of compact flash, bc that doesn't work like memory. That's a storage device. They kinda trick you...when they say it comes with 64 mb ram. It really comes with 32, plus the 32 mb of storage. Adding 32 to the memory slot gives you 64mb. Couple that with the fact that the Crusoe processor used (Transmeta3200) does all the floating point calulations in software rather than hardware, it will be slow as molasses. I would guess probably about the same speed as a 233Mhz Pentium (one) processor running with 64mb ram. The video device is also short on memeory. you may gain some increases in speed by using 16 colors or 256 colors instead of as many as you can force it to use. I don't mean to condemn you for wanting to use Windoze...it is the most used OS right now at least. Just don't expect wonderful things. Ricko 05-06-2001 21:49:13

RE:Updated testing results and new website (modified 0 times) yellow1 Profile | Email Actually I replaced the 32mb SODIM by a 128mb one. Only $80. Anyway, the good news is that under WinME, this device is MUCh zippier, see my new post http://www.linux-hacker.net/cgi-bin/UltraBoard/UltraBoard.pl?Action=ShowPost&Board=aolgw&Post=25&Idle=0&Sort=0&Order=Descend&Page=0&Session=yellow1.98922235858499 05-07-2001 01:01:41

RE:Updated testing results and new website (modified 0 times) ckbone Profile I'm curious how this thing benchmarks...You can get Sandra2001 free here under power tools: http://www.nonags.com/nonags/ The cpu benchmark most closely relates to how fast the whole system will be. Reading between the lines here, it sounds like this chip is about the same as the RISE266 chip in an I-Opener. The first cpu benchmark with a RISE chip on Win98 is about 425. Just curious here...I was thinking about getting one. These latest revelations give me second thoughts. 05-07-2001 10:47:13

RE:Updated testing results and new website (modified 0 times) yellow1 Profile | Email I'll take a look at Sandra, but why don't you run iBench with your iOpener and send in the results, no downloads required and it will probably take you about 10 minutes to run the HTML Load, that is if the iOpener is as fast as the GWCT under WinME The CPU benchmark is a synthetic benchmark that might not be very adapted to the Cruose processor. You're better off testing real life performance of the whole unit, i.e. CPU/Video, etc... 05-07-2001 12:28:16

RE:Updated testing results and new website (modified 0 times) ckbone Profile Since many people may be new to Windows ME, you might try this to really speed-up your system....Turn off Restore. You can find it in control panel/system/performance (I think). Anyway do it...it helps low-powered systems, my I-Opener included. I'd still like to see a Sandra cpu benchmark. You could tell what you really have....A K6400 on an I-Opener scores about 750. 05-08-2001 19:14:03

RE:Updated testing results and new website (modified 0 times) yellow1 Profile | Email Yep restore is off and a lot of the clutter went away with 98lite too. Ok I'll run Sandra tonight ! Did you run iBench ? 05-08-2001 19:58:39

RE:Updated testing results and new website (modified 0 times) ckbone Profile Yellow1.....I'm still looking for a Sandra CPU benchmark. Look, there's no reason to keep it secret. Nobody's opinion of this appliance will probably be changed if you publish the results. I'm still trying to convince myself to get one. I'd like to know what I dealing with, before I plunk down the cash...... My web connection is normally so poor, any results I could give using an I-Opener, would be meaningless. 05-11-2001 17:47:14

RE:Updated testing results and new website (modified 0 times) ckbone Profile FWIW...I put up some screen shots of the Sandra benchmarks of three I-Openers, and two other computers, on my Yahoo Photo page. The last few versions of Sandra return identical benchmarks. I also have the benchmarks of these I-Openers with three different CPUs (no screen shots). Here's the breakdown: Winchip @210 245ALU Winchip2 @210 305ALU Pentium200mmx @210 448ALU Rise266 @210 425ALU K6III-333 @210 530ALU K62+ 400 @420 740ALU I am really interested in how this Transmedia 400 fits into this list. From descriptions of performance posted on this board, my guess would be a benchmark of about 300ALU. Won't know of course, unless Yellow1 or somebody else posts results. This is no contest...just being curious. Here's the screen shots...in the album SISoft Sandra CPU Benchmarks. http://photos.yahoo.com/ckbone63123 05-12-2001 18:51:19

RE:Updated testing results and new website (modified 0 times) yellow1 Profile | Email As I've explained several times, iBench's test is independent (except for the first load) from your connection speed. I'm also very interested in your results with the iOpener. 05-12-2001 20:08:59

RE:Updated testing results and new website (modified 0 times) smee Profile My first shot at Sandra CPU Benchmark: I've never run this before so I don't know if I'm missing any options, etc. Gateway Connected TouchPad Transmeta Crusue 3200 400MHz Windows ME (normal install, no 98lite) Video and audio drivers loaded Stock memory 10G Hitachi 9.5" drive HPNA version with Linksys USB network Dhrystone ALU 498 MIPS Whetstone FPU 139 MFLOPS 05-13-2001 09:38:16

RE:Updated testing results and new website (modified 0 times) wordsmith Profile Here is where I become ignorent. Could you explain what the difference between those two are, and if they are good bench marks? Thanks! Also, Nice Work! * WordsmitH 05-13-2001 11:37:56

RE:Updated testing results and new website (modified 0 times) ckbone Profile Thanks smee....I'm surprised at the score. I wonder why Yellow1 has trouble playing mp3s and doing other things on this computer with that score. With an I-Opener any score over 300 or so, will let you play mp3s and surf or whatever. I wonder what the hang-up is. It could be a BIOS setting...maybe a system configuration. Maybe just needs more tweaking.....thanks again. 05-13-2001 11:39:46

RE:Updated testing results and new website (modified 0 times) ckbone Profile FWIW....ALU and FPU usually follow each other...when ALU goes up so does FPU. In this case, FPU does not seem to be as high as it should. Somebody else will have to explain this....this may be a reason for a percieved lack of performance. The I-Opener does not have any L2 cache on the board...this slows it down quite a bit. I'm not sure about the GCT. I'm certain there are other people on this BBS who could diagnose these numbers. Maybe a few more Sandra scores from other people would help. This score should make for a reasonably fast computer....about the speed of a pentium 233. 05-13-2001 12:02:02

RE:Updated testing results and new website (modified 0 times) smee Profile I've been able to play mp3s with no problem. I'll have to check the spec's on what I was playing, but they sounded good. I think yellow1 (or maybe it was somebody else) said that mp3s played, just that the cpu usage was high - you couldn't do anything else. I was using the standard WinME media player and grabbed them from over my local network. I haven't done anything yet to try and improve performance. 05-13-2001 12:19:34

RE:Updated testing results and new website (modified 0 times) ckbone Profile Windows new media player is a real cpu hog.....with any media. Mp3s play much better with WinAmp. Depending on how they were recorded, different mp3s will definately play better(no skipping). Playing them straight off the hard drive is the easiest on the cpu. I wonder if this GCT can be over-clocked? 05-14-2001 10:05:19

RE:Updated testing results and new website (modified 0 times) ttn1 Profile | Email I have read reports of the Transmeta chips doing much better in benchmarks after a few runs. The code morphing takes a run or two to come up to speed. Try running sandra two or three times and see if the bench doesn't get higher. In fact from what I've read, most things will get faster after you run it a few times. I haven't had a chance to try out a transmeta chip, this is just what I've heard. ttn1 05-14-2001 10:41:47

RE:Updated testing results and new website (modified 0 times) yellow1 Profile | Email That's also the hype I read, but haven't experienced first hand with the GWCP over the past two weeks I've had it up and running. A synthetic benchmark like Sandra is probably not a very good test to run on devices like that anyway, because they're so tightly integrated (video/audio, etc...)that just testing the power of the CPU isn't going to tell you much. Going by Sandra, the SC3200 at 400mhz has the power of a Pentium Pro 200...which is ridiculous at best after you've experienced the respective "zippiness" of such systems... Using something iBench is really the only way to go for a websurfing appliance like that. I'm surprised no one realizes that... 05-14-2001 12:37:17

RE:Updated testing results and new website (modified 0 times) ckbone Profile Yellow1....you have to realize these devices....I-Opener, GCT, Websurfer, and others... are being used for many other things besides surfing. How do you surf with one in your car, for instance? Also, since the I-Opener can support multiple kinds of CPUs, and I've tried six sifferent ones, I can tell you with certainty, the Sandra benchmarks increase as you try faster CPUs. Whether a Sandra benchmark on an I-Opener can be compared to a GCT benchmark...I guess it's debatable. But I've used this benchmark on many computers....never once did the benchmarks lie. If your computer has a higher CPU benchmark.....it's a faster computer, no matter what the other benchmarks are. What did your CPU benchmark say? 05-14-2001 19:01:08

RE:Updated testing results and new website (modified 0 times) yellow1 Profile | Email Same as what was posted above. Again I don't think synthetic benchmarks are of much use. Sandra tells me that the SC3200 at 400mhz compares to a Pentium Pro at 200mhz. This may be true in raw integer/fpu calculations but once you realize the PPro had a 200mhz 512L2 cache and the SC3200 has none, that's pretty much the end of that. Do me a favor and run iBench's HTML load, it will take you 10 minutes and will provide a meaningful comparison. iBench tests the speed at which a PC RENDERS web pages (regardless of connection speed when it draws from the cache) and believe me, that's a very good real life test on how your CuIA will perform across the board (mp3 playing, GPS, etc...). I've run Sandra and it's obvious that it's a flawed mehtod with the PPro 200 example. Hope that helps. 05-14-2001 19:40:38

Reply to Thread | Printer | All times are PST Powered by UltraBoard v1.62

I-Appliance BBS General - NEWS, Breaking NEWS - linux-hacker.net BBS on CD Questions - User Groups in Your Town - E-Bay, Yahoo Auction LINKS ONLY (MODERATED) - Jokes & Humor - usdtv box I-Opener Areas - I-Opener General Posts - I-Opener Technical Stuff - I-Opener Technical Info Reference (MODERATED) WebSurfer Areas - WebSurfer General Posts - WebSurfer Technical Stuff - WebSurfer Technical Info Reference (MODERATED) Other I-Appliances - Virgin Webplayer - WebPal - MailStation - 3Com Audrey - ThinkNIC / Oracle NIC - Gateway/AOL Connected Touch Pad - Philips/AOL-TV - ACER NT-150 STB - MSN Companion - Compaq IA - Emachines IA - IBM NetVista IA - Intel dot.station - MSNTV2 Network Computers - Fujitsu-Siemens Scovery Linux Net Computers - GallantWeb Server Consoles - Dreamcast - xbox hacking Web Pads - Epods One General Post - Panasonic CF-01 - Fujitsu Stylistic 1200 - Cybiko Thin Clients - Thin Clients DigitalVideoRecorders - DTV Recorders, Replay, TiVo, Etc... - Tivo Hacks Linux - Midori Linux General Board - ITX MotherBoards Single Board Computers - Earth Computers mARMalade Elite Member Area MISC Areas - The Swap Meet WE TAKE NO RESPONSIBILITY FOR ANY TRANSACTION HERE AND NO COMMERCIAL POST PERIOD !!! - Wireless Net - Cameras - Robotics - Routers/Gateways - Mattel JuiceBox