Re: [expert] New Mobo Roundup--Dr Tom
On Thu, 2002-05-16 at 09:25, Tom Brinkman wrote: I'm with you. It's not just motherboard revisions, bios updates either. Chipsets and cpu steppings (production runs) get fixes usually within the first several months. A good example is the VIA kt133a. The ide bug surfaced during the first two steppings. Later it was discovered that many motherboard designs exacerbated the problem. By step 4 of the kt133a the problem had all but disappeared (as long as you didn't use an Abit kt7* ; Later it became apparent some high bandwith pci cards (specially the SB Live!) were largely part of the problem. So if you jumped on bandwagon when the KT7 first appeared, installed an SB Live!, you were pretty much SOL, 'cept for bios upgrades, and upping the Vcore to 1.82v, and dropping the HDD mode to = udma2. Dirty work arounds. Much the same scenario with cpu steppings. I think there's a little confusion here about what's a fix and what's a workaround. In my day, a bios upgrade that corrected problems was called a fix. Since I'm one of those KT7 owners, I can state with some authority that it's not hard to apply a bios flash, since I've done it with this particular board at least four times that I can remember. And I wasn't out to fix problems, because I did'nt really notice any; I was out to take advantage of new bios features. But even if I wasn't and I was in a worse case scenario with an SB card, I would not have been SOL; I would have just downloaded and flashed a new bios into the mainboard. For the record, there's been no problems with this particular Abit KT7. I also think staying away from the cuttin edge, first releases of boards and chipsets is very important for open source too. EG, the people who rushed out to by ATI 8500's and still don't have support for them yet. There's also work arounds for chipset/cpu errata that take a while. EG, I have a kt133a board, shortly after I built the system, months after kt133a boards were out, dmesg began to include the line Applying VIA southbridge workaround. Tho since my board (Soyo, AMD appr'vd) is the second (and final) revision, it's not an Abit, I don't have an SB Live, and both the chipset and the Tbird are 4th stepping, I doubt I need that kernel parameter ;) I agree with you as far as the ATI cards go. With the mainboards, and the roundup comparisons, I've never been led astray with regard to Linux yet (knock on wood) in relation to third party mainboard roundups. I respect the AMD approved list, but I don't completely trust it because I believe they are somewhat in a conflict of interests since they produce their own chipset. AFAICT, the third party sites doing reviews are not in that kind of a position and conflict. It's entirely possible that (and I do trust AMD much more than Intel) AMD would not put a motherboard on their approved list if by chance their marketing department deemed the mainboard a viable competition for some of AMD's own product. Not that they do, mind you; I just think it's better business to work from a third party site for information regarding mainboard evaluation. Then there's also performance regardless of OS. It's reported that DDR333 has negligible improvement over 266. In some tests it's actually slower. I attribute this to 266 havin matured, while 333 is still brand new and wet behind the ears. Not neccesarilly the ram's fault, but the motherboard's implementation of it. Y'allsMMV ;) -- Tom BrinkmanCorpus Christi, Texas The reference board comparison between 266 and 333 attempt to bear out what you are saying: http://www6.tomshardware.com/mainboard/02q1/020220/kt333-11.html Unfortunately, this isn't the whole story, because the new 333's are production boards, and as such perform much differently than the reference boards; they outperform the KT266A's. In the following graph I direct your attention to the green bars, which happen to be a KT266A and a KT133A, respectively. http://www6.tomshardware.com/mainboard/02q2/020509/kt333-29.html In comparison to the KT333's there is a substantial difference in performance. Both the older KT266A and the KT133A are close to the bottom of the chart in performance. The Epox, Enmic, and Gigabyte boards consistently perform close to the top of the roundup. In addition, I did a little research on the new Abit KX7-333R mobo and was pleasantly surprised; it outperformed the Epox 8k3a+ in another site's OpenGL benchmarks under Quake 3: http://www.tweakers.com.au/articles/motherboard/abit_kx7333r/page9.asp In theory then, this board would have been at or close to the top of the Dr Tom roundup in those benchmarks. Best Regards, LX -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°°
Re: [expert] New Mobo Roundup--Dr Tom
On 15 May 2002 13:02:09 -0400 Lyvim Xaphir [EMAIL PROTECTED] wrote: On Wed, 2002-05-15 at 01:18, Sridhar Dhanapalan wrote: How good are the integrated devices on these boards, compared to the ones you can buy as add-in PCI cards? Don't know yet, you've got me started on CPU utilization numbers. ;) I'll give you the bean spill when I get thru distilling this google page... Well, here's a bit I just found, regarding CPU utilization and IEEE 1394/USB 2.0: http://www.extremetech.com/article/0,3396,apn=7s=1005a=25038app=5ap=6,00.asp This test puts both buses under considerable stress as it is evaluating a portable hard drive capable of attaching to both buses. Therefore to the authors of the article it seemed an excellent opportunity for a bus to bus performance comparison. BTW, it may be just me, but that site seems slow. It's not just you their adds come from Doubleclick and it's sloww as molasas. + Tom's reviews never seem to give them much attention. How do they perform? Do they tax the CPU any more than a standalone card would? I am interested in getting a board with integrated sound and ethernet. I would like the board to have at least 4 IDE interfaces (not channels) so I can give an entire interface to each of my devices. How well do the features of these boards work in Linux? Based on these, can anyone make a board recommendation? I don't want to spend too much money, and that is the reason why I'm going for integrated peripherals. They need to be decent, though. In the link above, it's evident that USB 2.0 and IEEE 1394 are very very close in throughput; 1394 has the edge in CPU utilization at 11.6%. USB 2.0 rated at 15.4%. The authors of the article I'm looking at point out that drivers have a big impact on these numbers. I don't yet see a mention of the chipset that supports USB; I'm still looking for comparisons of that sort. With regard to IDE controllers, I can tell you that a HPT37X IDE controller is better than anything I've seen yet for IDE drives; including the Promise options. I've got my primary Raid 0 array on the integrated Highpoint controller. This frees up the vanilla IDE controller on the mainboard for such mundane stuff like CDrom or zip, or experimentation; keeping the Highpoint channels free and dedicated to soft Raid. Since you're board searching, check this out: http://www.enmic.de/www/produkte/boards/8ttx2+/8ttx2+.htm First board from a German company I've ever seen; impressive. In the Uncle Tom roundup OpenGL standings, here is where it stood: http://www6.tomshardware.com/mainboard/02q2/020509/kt333-29.html Originally I was leaning towards the Epox, but the german Enmic board is touted by Pabst as being supremely stable; by comparison he said there were some crashes with the Epox 8k3a+. This is interesting since externally the Epox and the Enmic are virtually indistinguishable. The stability statement doesn't bother me much, (the Epox) I take that with a grain of salt, since he got a weeks worth of benches out of this board, after it went thru the standard burn in process. If he had gotten some real trouble he would have raised holy hell. Or I should say unholy hell. What does bother me is the same thing that might be attractive to you, namely the onboard sound. I'd rather not have onboard sound, I think it's evil; I've already got two sound cards here that are exceptional. However, since I love the Epox and the Enmic boards so much in this review, I'm considering getting one of those anyway and then disabling the onboard sound. I looked in the features section of the TH review (there for anybody who bothered to read the article) and found what chipset the onboard sound is using, so you can evaluate it for yourself: http://www.realtek.com.tw/htm/products/cp/alc650.asp If by chance you are interested in perusing the board features, here is the link to the Enmic: http://www6.tomshardware.com/mainboard/02q2/020509/kt333-07.html HTH, Best Regards, L8r, LX :) -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
The question I'd have on this thread is: For none high end applications is it worth it to be on the edge? 1.0 of anything is usually not good. I also think about things like nics (I've fried a few) etc. Whereas a Linksys or Netgear can be bought for 15 bucks in the US it scares me to think that I could lose a mobo cause my nic went down. For personal use and for High reliability, I really like to stay one step behind the curve. Wait till the bug fix version comes out. (very important if you are buying closed source software!) Am I alone in this attitude? James On Thu, 16 May 2002 14:20:09 +1000 Sridhar Dhanapalan [EMAIL PROTECTED] wrote: On 15 May 2002 13:02:09 -0400, Lyvim Xaphir [EMAIL PROTECTED] wrote: On Wed, 2002-05-15 at 01:18, Sridhar Dhanapalan wrote: With regard to IDE controllers, I can tell you that a HPT37X IDE controller is better than anything I've seen yet for IDE drives; including the Promise options. I've got my primary Raid 0 array on the integrated Highpoint controller. This frees up the vanilla IDE controller on the mainboard for such mundane stuff like CDrom or zip, or experimentation; keeping the Highpoint channels free and dedicated to soft Raid. Since you're board searching, check this out: http://www.enmic.de/www/produkte/boards/8ttx2+/8ttx2+.htm First board from a German company I've ever seen; impressive. In the Uncle Tom roundup OpenGL standings, here is where it stood: http://www6.tomshardware.com/mainboard/02q2/020509/kt333-29.html Originally I was leaning towards the Epox, but the german Enmic board is touted by Pabst as being supremely stable; by comparison he said there were some crashes with the Epox 8k3a+. This is interesting since externally the Epox and the Enmic are virtually indistinguishable. The stability statement doesn't bother me much, (the Epox) I take that with a grain of salt, since he got a weeks worth of benches out of this board, after it went thru the standard burn in process. If he had gotten some real trouble he would have raised holy hell. Or I should say unholy hell. I must've overlooked the Enmic the first time I read the review. The Gigabyte board (the review winner) initially looked the most attractive to me. It has a good price, great performance and a decent feature set. If i can find an Enmic supplier here is Australia, I might buy one of those instead. What does bother me is the same thing that might be attractive to you, namely the onboard sound. I'd rather not have onboard sound, I think it's evil; I've already got two sound cards here that are exceptional. This is the way I see it. Sound hardware has been reasonably decent for almost ten years now, and advances in sound hardware since the Sound Blaster 16 have relatively been minor in comparison to advances in components like CPUs and video cards. I don't need Audigy-quality sound, but I would like something nice and affordable. Integrated sound seems to fit the bill well. My brother has an older machine (circa 2000) with integrated audio. The sound tends to distort under high CPU loads (e.g. when playing a game), and this has made me wary of integrated solutions. Today, it seems as if manufacturers have gotten around this problem (otherwise, I suppose, reviewers would complain about it), and I find myself again considering integrated audio. The Gigabyte board, for example, uses the same chipset as the Sound Blaster PCI 128, which isn't too shabby. Another feature of the Gigabyte board is integrated ethernet. Most of the boards in the review with integrated ethernet use the VIA chipset for networking. I am wary of these -- I get the feeling that these rely on the CPU just like a winmodem (I'm only speculating, though). The Gigabyte board, on the other hand, uses a Realtek 8100BL, which is of the same family as the 8139 (which I've been using for the past few years without any problem). Thanks for the help! -- Sridhar Dhanapalan If you want to travel around the world and be invited to speak at a lot of different places, just write a Unix operating system. -- Linus Torvalds Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Wed, 15 May 2002 23:27:11 -0700 James [EMAIL PROTECTED] wrote: The question I'd have on this thread is: For none high end applications is it worth it to be on the edge? 1.0 of anything is usually not good. I also think about things like nics (I've fried a few) etc. Whereas a Linksys or Netgear can be bought for 15 bucks in the US it scares me to think that I could lose a mobo cause my nic went down. For personal use and for High reliability, I really like to stay one step behind the curve. Wait till the bug fix version comes out. (very important if you are buying closed source software!) Am I alone in this attitude? I have not read the reviews of the MOBOs from this thread so I am speaking only of integrated sound and NIC. I have a SOYO Dragon+ which has both integrated sound and NIC. As Sridhar made note in his last post in most cases this means VIA, which is the case with the SOYO. It Is straight hardware and taxes CPU no more than any other NIC. The sound is CMI 5.1, and this too is also the case with the other brands of upper end MOBOs, though there are a couple that use versions of the SBLive. Sound quality and features are excellent. Both the sound and the NIC can be disabled in the BIOS, so replacing or choosing to use a different component is hassle free and causes no difficulty. In todays market if you are looking at any of the upperend MOBOs onboard sound and NIC are a Plus. You still want to shy away from them on the lower end budget boards and I personally would 'never again' use any board that has onboard video, but that's another story. Charles Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Thursday 16 May 2002 01:27 am, James wrote: The question I'd have on this thread is: For none high end applications is it worth it to be on the edge? 1.0 of anything is usually not good. I also think about things like nics (I've fried a few) etc. Whereas a Linksys or Netgear can be bought for 15 bucks in the US it scares me to think that I could lose a mobo cause my nic went down. For personal use and for High reliability, I really like to stay one step behind the curve. Wait till the bug fix version comes out. (very important if you are buying closed source software!) Am I alone in this attitude? I'm with you. It's not just motherboard revisions, bios updates either. Chipsets and cpu steppings (production runs) get fixes usually within the first several months. A good example is the VIA kt133a. The ide bug surfaced during the first two steppings. Later it was discovered that many motherboard designs exacerbated the problem. By step 4 of the kt133a the problem had all but disappeared (as long as you didn't use an Abit kt7* ; Later it became apparent some high bandwith pci cards (specially the SB Live!) were largely part of the problem. So if you jumped on bandwagon when the KT7 first appeared, installed an SB Live!, you were pretty much SOL, 'cept for bios upgrades, and upping the Vcore to 1.82v, and dropping the HDD mode to = udma2. Dirty work arounds. Much the same scenario with cpu steppings. I also think staying away from the cuttin edge, first releases of boards and chipsets is very important for open source too. EG, the people who rushed out to by ATI 8500's and still don't have support for them yet. There's also work arounds for chipset/cpu errata that take a while. EG, I have a kt133a board, shortly after I built the system, months after kt133a boards were out, dmesg began to include the line Applying VIA southbridge workaround. Tho since my board (Soyo, AMD appr'vd) is the second (and final) revision, it's not an Abit, I don't have an SB Live, and both the chipset and the Tbird are 4th stepping, I doubt I need that kernel parameter ;) Then there's also performance regardless of OS. It's reported that DDR333 has negligible improvement over 266. In some tests it's actually slower. I attribute this to 266 havin matured, while 333 is still brand new and wet behind the ears. Not neccesarilly the ram's fault, but the motherboard's implementation of it. Y'allsMMV ;) -- Tom BrinkmanCorpus Christi, Texas Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
Tom Brinkman wrote: On Thursday 16 May 2002 01:27 am, James wrote: The question I'd have on this thread is: For none high end applications is it worth it to be on the edge? 1.0 of anything is usually not good. I also think about things like nics (I've fried a few) etc. Whereas a Linksys or Netgear can be bought for 15 bucks in the US it scares me to think that I could lose a mobo cause my nic went down. For personal use and for High reliability, I really like to stay one step behind the curve. Wait till the bug fix version comes out. (very important if you are buying closed source software!) Am I alone in this attitude? I'm with you. It's not just motherboard revisions, bios updates either. Chipsets and cpu steppings (production runs) get fixes usually within the first several months. A good example is the VIA kt133a. The ide bug surfaced during the first two steppings. Later it was discovered that many motherboard designs exacerbated the problem. By step 4 of the kt133a the problem had all but disappeared (as long as you didn't use an Abit kt7* ; Later it became apparent some high bandwith pci cards (specially the SB Live!) were largely part of the problem. So if you jumped on bandwagon when the KT7 first appeared, installed an SB Live!, you were pretty much SOL, 'cept for bios upgrades, and upping the Vcore to 1.82v, and dropping the HDD mode to = udma2. Dirty work arounds. Much the same scenario with cpu steppings. I also think staying away from the cuttin edge, first releases of boards and chipsets is very important for open source too. EG, the people who rushed out to by ATI 8500's and still don't have support for them yet. There's also work arounds for chipset/cpu errata that take a while. EG, I have a kt133a board, shortly after I built the system, months after kt133a boards were out, dmesg began to include the line Applying VIA southbridge workaround. Tho since my board (Soyo, AMD appr'vd) is the second (and final) revision, it's not an Abit, I don't have an SB Live, and both the chipset and the Tbird are 4th stepping, I doubt I need that kernel parameter ;) Then there's also performance regardless of OS. It's reported that DDR333 has negligible improvement over 266. In some tests it's actually slower. I attribute this to 266 havin matured, while 333 is still brand new and wet behind the ears. Not neccesarilly the ram's fault, but the motherboard's implementation of it. Y'allsMMV ;) -- Tom BrinkmanCorpus Christi, Texas Just to chime in with total agreement with both James and Tom. You can bleed a bit with cutting edge *software*, and if things get too bad, there will soon be patches or you can go back a build version for some stability. If you start bleeding with cutting edge *hardware*, there just might not be any Band-Aids to stop the bleeding. On the other hand, it must be said that pioneers, like the LX_MAN, are the reason why the hardware bugs are found in the first place, enabling the later development of a fix for said bugs. The people, such as Tom, James, and I, then have the opportunity to get the later hardware revisions... Keep on truckin, LX Dr John, The Night Tripper -- J. Craig Woods UNIX/NT Network/System Administration http://www.trismegistus.net Character is built upon the debris of despair --Emerson Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tue, 2002-05-14 at 16:35, David Rankin wrote: Great heads up!, I'm in the market for a board and love Abit. Keep of posted if you get info in the Abit review. David, here is a comparison of the Abit mobo to the Epox 8k3a+ board, which was reviewed in the Dr Tom roundup. So by looking at this comparison, you can get an idea of where the Kx7-333R would have been in the roundup. There's also feature and stability information. Features and photos: http://www.tweakers.com.au/articles/motherboard/abit_kx7333r/page2.asp OpenGL/Quake3 benchmarks: http://www.tweakers.com.au/articles/motherboard/abit_kx7333r/page9.asp Stability and conclusions: http://www.tweakers.com.au/articles/motherboard/abit_kx7333r/page10.asp -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
Yes, and so are the Promise controllers; however, the soft RAID is a better setup, and since either way you are using the CPU... Somewhere, I saw a comparison technically speaking between a Promise card and Linux Soft RAID. SoftRaid was far superior in all aspects. :) Finally, from what I understand, the Highpoint controllers are now usable under Linux in their native RAID modes, via (no pun intended) the latest kernel module code. Not that I'm interested meself; I'm staying with soft raid. Still, the chance is there for anyone interested to run the numbers. Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Wednesday 15 May 2002 6:18 am, Sridhar Dhanapalan wrote: How good are the integrated devices on these boards, compared to the ones you can buy as add-in PCI cards? Tom's reviews never seem to give them much attention. How do they perform? Do they tax the CPU any more than a standalone card would? I am interested in getting a board with integrated sound and ethernet. I would like the board to have at least 4 IDE interfaces (not channels) so I can give an entire interface to each of my devices. How well do the features of these boards work in Linux? As mentioned somewhere else, my borrowed machine has an Intel 815E-based motherboard with integrated sound, video and Ethernet. 8.2 installed no problem and the inteegrated parts were identified as: Ethernet: 3Com Corporation 3C905-TX Sound: Intel Corp 820 (Camino 2) Chipset AC97 Audio Controller Video: Intel Corp 815 CGC [Chipset Graphics Controller] There is no obvious extra CPU load and the first two work fine; the third is a problem, as there are artifacts and slight breaking-up when really driving things. However, there's an AGP 2x slot supplied and I'll be using it soon :) Alastair - -- Alastair Scott (London, United Kingdom) http://www.unmetered.org.uk/ -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: For info see http://www.gnupg.org iD8DBQE84hwnCv59vFiSU4YRAggKAJ9KCSNZMRkeNP41s8FsrQD9FxCYdgCgyb1a 38JrsuJSK0/a7lAap+36TzU= =7Khw -END PGP SIGNATURE- Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Wed, 15 May 2002 09:28:18 +0100 Alastair Scott [EMAIL PROTECTED] wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Wednesday 15 May 2002 6:18 am, Sridhar Dhanapalan wrote: How good are the integrated devices on these boards, compared to the ones you can buy as add-in PCI cards? Tom's reviews never seem to give them much attention. How do they perform? Do they tax the CPU any more than a standalone card would? I am interested in getting a board with integrated sound and ethernet. I would like the board to have at least 4 IDE interfaces (not channels) so I can give an entire interface to each of my devices. How well do the features of these boards work in Linux? As mentioned somewhere else, my borrowed machine has an Intel 815E-based motherboard with integrated sound, video and Ethernet. 8.2 installed no problem and the inteegrated parts were identified as: Ethernet: 3Com Corporation 3C905-TX Sound: Intel Corp 820 (Camino 2) Chipset AC97 Audio Controller Video: Intel Corp 815 CGC [Chipset Graphics Controller] There is no obvious extra CPU load and the first two work fine; the third is a problem, as there are artifacts and slight breaking-up when really driving things. However, there's an AGP 2x slot supplied and I'll be using it soon :) Actually on mine I've noticed certian programs break it up... like sylphheds menu bar. One point to note. Running glxgears I'm getting 3175 root 16 0 3020 3008 1688 R72.6 0.7 1:00 glxgears out of top where the 72.6 is the cpu load percentage. And the frame rate is 694 frames in 5.0 seconds = 138.800 FPS 697 frames in 5.0 seconds = 139.400 FPS 699 frames in 5.0 seconds = 139.800 FPS 698 frames in 5.0 seconds = 139.600 FPS 700 frames in 5.0 seconds = 140.000 FPS 703 frames in 5.0 seconds = 140.600 FPS Alhtough it does seem to nice around it it's still a heavy cpu user. James Alastair - -- Alastair Scott (London, United Kingdom) http://www.unmetered.org.uk/ -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.6 (GNU/Linux) Comment: For info see http://www.gnupg.org iD8DBQE84hwnCv59vFiSU4YRAggKAJ9KCSNZMRkeNP41s8FsrQD9FxCYdgCgyb1a 38JrsuJSK0/a7lAap+36TzU= =7Khw -END PGP SIGNATURE- Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tuesday 14 May 2002 10:26 pm, Sridhar Dhanapalan wrote: On Tue, 14 May 2002 18:02:12 -0500, Tom Brinkman [EMAIL PROTECTED] wrote: And ... I believe it's not quite time to upgrade an AMD system till the Thorobreds come out. Are the Thoroughbreds really that good? From what I have read, they are not much more than Palomino cores shrunk down to 0.09 microns. If you want a cache improvement, you'll have to wait for Barton, which is due to be released (according to AMD's roadmap) very close to Hammer-time (sorry -- Opteron-time). Clock speeds won't increase by very much, but I'm sure the shrunken die will make overclocking a bit safer. I don't believe they'll be that much better either, but they are comin soon, and will support (and have motherboard support) for internal monitoring of the cpu core temperature. A major item that has been sorely lacking from AMD cpu's. June 10th is the date I've heard. I probly should'a said 'not quite time to upgrade an _athlon_ system. It's long been time to upgrade lesser AMD or Intel systems. Specially since a 1.4 Tbird can be had for as little as $75, and can be run just fine on an AMD apprv'd $70 motherboard usin any old sdram (at 1.53 gig and outperform a P4 at 2 gig). The thinner .13 micron T-bred die (XP's are .18) should make for lower temps under load. OC'ing athlons has been difficult with the XP versions, probly even more so with the T-breds due to more intricate locking by AMD and the Tbirds/XP's run so damn hot. Last good oc'rs were the 1. 1.4 Tbirds. Instructions per clock (P4=6, XP=9) and on die cache have been and will continue to be superior to Intel's P4. It takes a P4 at 2.53 gig to out bench an XP2100+ (1.73 gig). So I believe the days of measuring cpu's primarily on clock speed are long over. -- Tom BrinkmanCorpus Christi, Texas Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
Lyvim Xaphir wrote: So you've got a mouse that communicates at 150 bytes/sec on a bus that has been RL evaluated at 5.7 Megabytes/sec; that's 38,000 times more bandwidth than the mouse requires. This is an invalid evaluation however, because I still don't know how fast you can type. ;) But, it wouldn't surprise me to find out that, for example, each byte requires a context switch in the processor, that requires, for example, putting 1000 bytes on the stack before processing each byte, and taking those 1000 bytes off the stack afterwards, etc., etc., etc. I'm not saying that's the case -- I don't know. It's just that I have been surprised sometimes at how slow things can be despite a fast processor, and finding out some of the reasons for the slowness. Enquiring minds want to know (more)! ;-) Randy Kramer Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On 15 May 2002 01:07:11 -0400, Lyvim Xaphir [EMAIL PROTECTED] wrote: On Tue, 2002-05-14 at 23:19, Sridhar Dhanapalan wrote: How is the CPU usage of USB compared to the 'legacy' ports (serial, parallel, PS/2, etc.)? The legacy ports were designed for older systems, so they cannot suck up too much juice. It is interesting that you bring up the CPU usage aspect of USB, as I've not seen any numbers regarding that. The resource benefits I was referring to was directed towards IRQ usage, because I have a real need for extra IRQ's (and I suspect others with legacy hardware do also), so in that regard the USB bus was a real benefit. There's a lot you can do with three extra IRQ's. I am interested in CPU utilization numbers, if you've got some URL's... I don't have any URLs; I'm just heard/read anecdotes about this. It just may be that USB devices do more than their PS/2 and serial versions. For instance, I read a review on the USB optical Intellimouse a few months ago. The reviewer said that the mouse was far more accurate and responsive in USB mode, but when he tried to use it in Quake 3 it slowed his machine so that he couldn't play it properly. He switched to PS/2 mode using the supplied adaptor, and the mouse became less responsive but also less-taxing on the CPU. I hear that USB, on the other hand, is a real pig in this regard (no surprise that Intel supports it). If that is the case, is it really worth using USB peripherals on a PC when legacy types would suffice? That begs the question of some numbers, does it not? Such as, what exactly is required for the device in question? (is it an isdn modem, or) Not only that, but in order to answer your question properly you also need some performance numbers in regard to the chipset platform that is giving you the USB functionality. Quite naturally, this is going to be giving different performance results depending on which chipset platform you choose; i.e., Intel or Via. Performance in this context meaning both CPU utilization and USB bus throughput. I'd love to see a performance comparison of the different buses. Does anyone on the list know of any? -- Sridhar Dhanapalan I once preached peaceful coexistence with Windows. You may laugh at my expense -- I deserve it. -- Jean-Louis Gassée, founder of BeOS Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Wed, 2002-05-15 at 01:18, Sridhar Dhanapalan wrote: How good are the integrated devices on these boards, compared to the ones you can buy as add-in PCI cards? Don't know yet, you've got me started on CPU utilization numbers. ;) I'll give you the bean spill when I get thru distilling this google page... Well, here's a bit I just found, regarding CPU utilization and IEEE 1394/USB 2.0: http://www.extremetech.com/article/0,3396,apn=7s=1005a=25038app=5ap=6,00.asp This test puts both buses under considerable stress as it is evaluating a portable hard drive capable of attaching to both buses. Therefore to the authors of the article it seemed an excellent opportunity for a bus to bus performance comparison. BTW, it may be just me, but that site seems slow. Tom's reviews never seem to give them much attention. How do they perform? Do they tax the CPU any more than a standalone card would? I am interested in getting a board with integrated sound and ethernet. I would like the board to have at least 4 IDE interfaces (not channels) so I can give an entire interface to each of my devices. How well do the features of these boards work in Linux? Based on these, can anyone make a board recommendation? I don't want to spend too much money, and that is the reason why I'm going for integrated peripherals. They need to be decent, though. In the link above, it's evident that USB 2.0 and IEEE 1394 are very very close in throughput; 1394 has the edge in CPU utilization at 11.6%. USB 2.0 rated at 15.4%. The authors of the article I'm looking at point out that drivers have a big impact on these numbers. I don't yet see a mention of the chipset that supports USB; I'm still looking for comparisons of that sort. With regard to IDE controllers, I can tell you that a HPT37X IDE controller is better than anything I've seen yet for IDE drives; including the Promise options. I've got my primary Raid 0 array on the integrated Highpoint controller. This frees up the vanilla IDE controller on the mainboard for such mundane stuff like CDrom or zip, or experimentation; keeping the Highpoint channels free and dedicated to soft Raid. Since you're board searching, check this out: http://www.enmic.de/www/produkte/boards/8ttx2+/8ttx2+.htm First board from a German company I've ever seen; impressive. In the Uncle Tom roundup OpenGL standings, here is where it stood: http://www6.tomshardware.com/mainboard/02q2/020509/kt333-29.html Originally I was leaning towards the Epox, but the german Enmic board is touted by Pabst as being supremely stable; by comparison he said there were some crashes with the Epox 8k3a+. This is interesting since externally the Epox and the Enmic are virtually indistinguishable. The stability statement doesn't bother me much, (the Epox) I take that with a grain of salt, since he got a weeks worth of benches out of this board, after it went thru the standard burn in process. If he had gotten some real trouble he would have raised holy hell. Or I should say unholy hell. What does bother me is the same thing that might be attractive to you, namely the onboard sound. I'd rather not have onboard sound, I think it's evil; I've already got two sound cards here that are exceptional. However, since I love the Epox and the Enmic boards so much in this review, I'm considering getting one of those anyway and then disabling the onboard sound. I looked in the features section of the TH review (there for anybody who bothered to read the article) and found what chipset the onboard sound is using, so you can evaluate it for yourself: http://www.realtek.com.tw/htm/products/cp/alc650.asp If by chance you are interested in perusing the board features, here is the link to the Enmic: http://www6.tomshardware.com/mainboard/02q2/020509/kt333-07.html HTH, Best Regards, L8r, LX :) -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Wed, 2002-05-15 at 10:52, Sridhar Dhanapalan wrote: I don't have any URLs; I'm just heard/read anecdotes about this. It just may be that USB devices do more than their PS/2 and serial versions. For instance, I read a review on the USB optical Intellimouse a few months ago. The reviewer said that the mouse was far more accurate and responsive in USB mode, but when he tried to use it in Quake 3 it slowed his machine so that he couldn't play it properly. He switched to PS/2 mode using the supplied adaptor, and the mouse became less responsive but also less-taxing on the CPU. This isn't a bandwidth problem, there's no way it could be. Therefore it must be one of three other things: driver bugs (immaturity), peripheral device bugs, or USB chipset bugs. Usually I've seen chipset bugs worked around with driver fixes, many times before in the past; I suspect the trend will continue successfully. also need some performance numbers in regard to the chipset platform that is giving you the USB functionality. Quite naturally, this is going to be giving different performance results depending on which chipset platform you choose; i.e., Intel or Via. Performance in this context meaning both CPU utilization and USB bus throughput. I'd love to see a performance comparison of the different buses. Does anyone on the list know of any? I did'nt see this note until now; I posted another reply to one of your posts including a URL I located after you raised the topic of USB CPU utilization last night; check it out. :) Plus I posted some more mainboard bits. LX -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On 15 May 2002 13:02:09 -0400, Lyvim Xaphir [EMAIL PROTECTED] wrote: On Wed, 2002-05-15 at 01:18, Sridhar Dhanapalan wrote: With regard to IDE controllers, I can tell you that a HPT37X IDE controller is better than anything I've seen yet for IDE drives; including the Promise options. I've got my primary Raid 0 array on the integrated Highpoint controller. This frees up the vanilla IDE controller on the mainboard for such mundane stuff like CDrom or zip, or experimentation; keeping the Highpoint channels free and dedicated to soft Raid. Since you're board searching, check this out: http://www.enmic.de/www/produkte/boards/8ttx2+/8ttx2+.htm First board from a German company I've ever seen; impressive. In the Uncle Tom roundup OpenGL standings, here is where it stood: http://www6.tomshardware.com/mainboard/02q2/020509/kt333-29.html Originally I was leaning towards the Epox, but the german Enmic board is touted by Pabst as being supremely stable; by comparison he said there were some crashes with the Epox 8k3a+. This is interesting since externally the Epox and the Enmic are virtually indistinguishable. The stability statement doesn't bother me much, (the Epox) I take that with a grain of salt, since he got a weeks worth of benches out of this board, after it went thru the standard burn in process. If he had gotten some real trouble he would have raised holy hell. Or I should say unholy hell. I must've overlooked the Enmic the first time I read the review. The Gigabyte board (the review winner) initially looked the most attractive to me. It has a good price, great performance and a decent feature set. If i can find an Enmic supplier here is Australia, I might buy one of those instead. What does bother me is the same thing that might be attractive to you, namely the onboard sound. I'd rather not have onboard sound, I think it's evil; I've already got two sound cards here that are exceptional. This is the way I see it. Sound hardware has been reasonably decent for almost ten years now, and advances in sound hardware since the Sound Blaster 16 have relatively been minor in comparison to advances in components like CPUs and video cards. I don't need Audigy-quality sound, but I would like something nice and affordable. Integrated sound seems to fit the bill well. My brother has an older machine (circa 2000) with integrated audio. The sound tends to distort under high CPU loads (e.g. when playing a game), and this has made me wary of integrated solutions. Today, it seems as if manufacturers have gotten around this problem (otherwise, I suppose, reviewers would complain about it), and I find myself again considering integrated audio. The Gigabyte board, for example, uses the same chipset as the Sound Blaster PCI 128, which isn't too shabby. Another feature of the Gigabyte board is integrated ethernet. Most of the boards in the review with integrated ethernet use the VIA chipset for networking. I am wary of these -- I get the feeling that these rely on the CPU just like a winmodem (I'm only speculating, though). The Gigabyte board, on the other hand, uses a Realtek 8100BL, which is of the same family as the 8139 (which I've been using for the past few years without any problem). Thanks for the help! -- Sridhar Dhanapalan If you want to travel around the world and be invited to speak at a lot of different places, just write a Unix operating system. -- Linus Torvalds Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tuesday 14 May 2002 01:29 am, you wrote: Guys, if you're in the market for some new technology for under your hood, I strongly suggest that you check out the 18 mainboard lineup that just came out as of May 9th on Toms: http://www.tomshardware.com/mainboard/02q2/020509/index.html These boards are using the new VIA Apollo KT333 chipset; its the one I've been waiting for, so I could upgrade. There is one board that evidently came out too late for their tests, which took over two weeks. It's one I'm personally watching: http://www.abit.com.tw/abitweb/webjsp/english/products_shortcut.jsp?pPRODUC T_TYPE=MotherBoardpMODEL_NAME=KX7-333R However, no matter how good the Abit board looks on Abit's home site, I'm not going to buy it until it's been reviewed in a roundup. I'm in the market for a mobo upgrade. As it sits, judging from the roundup, I'm looking strongly at either the Epox 8k3a+, or the Enmic 8TTX2+. The Gigabyte got a strong review, but if a mobo doesn't have at least 7 slots, it's not interesting. Overclocking features are important too, which both the Epox and Enmic have. I wish he'd go ahead and update this roundup with the Abit KX7-333R, so I could finalize a decision. L8r, LX LX: Skimming through the article, I noticed that some of these boards have gone to a USB-only configuration -- no serial or parallel ports. Given that these boards are aimed at end-users and not OEM's, this does not sound like a Really Good Idea to me. In my case, it would mean adding the cost of a new printer and a new modem to the overall project cost, and those are not trivial costs. Besides, my present modem and printer work just fine, thank you. -- cmg Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
Great heads up!, I'm in the market for a board and love Abit. Keep of posted if you get info in the Abit review. Lyvim Xaphir wrote: Guys, if you're in the market for some new technology for under your hood, I strongly suggest that you check out the 18 mainboard lineup that just came out as of May 9th on Toms: http://www.tomshardware.com/mainboard/02q2/020509/index.html These boards are using the new VIA Apollo KT333 chipset; its the one I've been waiting for, so I could upgrade. There is one board that evidently came out too late for their tests, which took over two weeks. It's one I'm personally watching: http://www.abit.com.tw/abitweb/webjsp/english/products_shortcut.jsp?pPRODUCT_TYPE=MotherBoardpMODEL_NAME=KX7-333R However, no matter how good the Abit board looks on Abit's home site, I'm not going to buy it until it's been reviewed in a roundup. I'm in the market for a mobo upgrade. As it sits, judging from the roundup, I'm looking strongly at either the Epox 8k3a+, or the Enmic 8TTX2+. The Gigabyte got a strong review, but if a mobo doesn't have at least 7 slots, it's not interesting. Overclocking features are important too, which both the Epox and Enmic have. I wish he'd go ahead and update this roundup with the Abit KX7-333R, so I could finalize a decision. L8r, LX -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com -- David C. Rankin, J.D., P.E. RANKIN * BERTIN, PLLC 1329 N. University, Suite D4 Nacogdoches, Texas 75961 (936) 715-9333 (936) 715-9339 fax Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tue, 2002-05-14 at 10:29, Carroll Grigsby wrote: LX: Skimming through the article, I noticed that some of these boards have gone to a USB-only configuration -- no serial or parallel ports. Given that these boards are aimed at end-users and not OEM's, this does not sound like a Really Good Idea to me. In my case, it would mean adding the cost of a new printer and a new modem to the overall project cost, and those are not trivial costs. Besides, my present modem and printer work just fine, thank you. -- cmg Ahhh...Carroll, my friend. :) That is the beauty part. If indeed you do contemplate an upgrade, all you need do is get a usb serial/parallel adapter for a modest sum, and presto, your printer and modem are back in business. In addition, having done that you can do as I did and use IRQ's 7, 3, and 4 for things other than being tied to onerous dated and obsoleted onboard ports. Obsoleted in the sense that their functions are now duplicated on a less resource intensive external bus. Best Regards, LX -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tuesday 14 May 2002 03:35 pm, David Rankin wrote: Great heads up!, I'm in the market for a board and love Abit. Keep of posted if you get info in the Abit review. Lyvim Xaphir wrote: Guys, if you're in the market for some new technology for under your hood, I strongly suggest that you check out the 18 mainboard lineup that just came out as of May 9th on Toms: http://www.tomshardware.com/mainboard/02q2/020509/index.html One Abit board has finally made it back to the AMD approved list after long being a vendor to avoid. I wouldn't buy a board not AMD approved. And ... I believe it's not quite time to upgrade an AMD system till the Thorobreds come out. More importantly tho, while I think Tom's is a good site. it's still primarily a Windoze review site. Heck even Dell's sometimes sort'a work OK with Linux ; Just 'cause a board garners rave reviews with Windoze, there's no guarantee (MOF, just the opposite more'n more lately) it'll work as well or be as supported by the OS, as it is by the Win-tel/nvidia/ ati/creative/dell, etc. gang. EG, most all of Tom's boards incorporate Win-fake-raid. Also, preliminary reports I've seen are that ddr333 offers none, to very little improvement over 266 (much the same as with ata/133 over /100). One advantage tho is if you get marginal ddr333, it might run as ddr266 at cas2 (bettr'n 333 at CL2.5). So buy quality (micron, infineon) ddr333 now while it's cheap, wait for the (Linux) dust to settle on the rest of the cuttin' edge hardware. -- Tom BrinkmanCorpus Christi, Texas Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On May 14, 2002 02:35 pm, David Rankin wrote: Great heads up!, I'm in the market for a board and love Abit. Keep of posted if you get info in the Abit review. Lyvim Xaphir wrote: Guys, if you're in the market for some new technology for under your hood, I strongly suggest that you check out the 18 mainboard lineup that just came out as of May 9th on Toms: http://www.tomshardware.com/mainboard/02q2/020509/index.html These boards are using the new VIA Apollo KT333 chipset; its the one I've been waiting for, so I could upgrade. There is one board that evidently came out too late for their tests, which took over two weeks. It's one I'm personally watching: http://www.abit.com.tw/abitweb/webjsp/english/products_shortcut.jsp?pPROD UCT_TYPE=MotherBoardpMODEL_NAME=KX7-333R However, no matter how good the Abit board looks on Abit's home site, I'm not going to buy it until it's been reviewed in a roundup. I'm in the market for a mobo upgrade. As it sits, judging from the roundup, I'm looking strongly at either the Epox 8k3a+, or the Enmic 8TTX2+. The Gigabyte got a strong review, but if a mobo doesn't have at least 7 slots, it's not interesting. Overclocking features are important too, which both the Epox and Enmic have. I wish he'd go ahead and update this roundup with the Abit KX7-333R, so I could finalize a decision. L8r, LX -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com Is this review OK? http://www.legionhardware.com/html/doc.php?id=168 They also did a test/review along the same lines as Dr. Pabst if you'd like to do a direct comparison of all of the boards. Skipping ahead to the conclusion page and the last paragraph (yeah yeah, I read the end of the book before I buy it to see if I'll like the ending ;-) ) If you require the very best performance then the KX7-333R is your board, however if your not fussed about performance and second best is good enough, then the AT7 MAX may suit your needs better. That is if you're after loads of features, no legacy support and aren't scared to pay through your nose for a motherboard. I can't express how impressed I am with the performance of the KX7-333R, so I wont go on and on about it. To conclude this review I would just like to say the KX7-333R is the fastest KT333 board I have reviewed to date and possibly the fastest on the market! HTH -- Charlie, Edmonton, AB, Canada Mandrake 8.2 Registered Linux user 244963, http://counter.li.org Arguments are extremely vulgar, for everyone in good society holds exactly the same opinion. -- Oscar Wilde Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tue, 2002-05-14 at 19:02, Tom Brinkman wrote: On Tuesday 14 May 2002 03:35 pm, David Rankin wrote: Great heads up!, I'm in the market for a board and love Abit. Keep of posted if you get info in the Abit review. Lyvim Xaphir wrote: Guys, if you're in the market for some new technology for under your hood, I strongly suggest that you check out the 18 mainboard lineup that just came out as of May 9th on Toms: http://www.tomshardware.com/mainboard/02q2/020509/index.html One Abit board has finally made it back to the AMD approved list after long being a vendor to avoid. I wouldn't buy a board not AMD approved. And ... I believe it's not quite time to upgrade an AMD system till the Thorobreds come out. More importantly tho, while I think Tom's is a good site. it's still primarily a Windoze review site. Heck even Dell's sometimes sort'a work OK with Linux ; It is probably a safe choice to select from AMD's approved list. The main reason I don't is simply because I have never, as yet, selected a board from a roundup review that was either anti-athlon or anti-linux; and also because I have in the past found AMD's list a little too restrictive for my taste; somewhat unusually so. One restriction case in point was what you yourself brought up, Tomthe Abits. I've been running this particular Abit for over two years now, and I'm fairly certain that it never did make the approved list. Normally if a board passes diags and runs a full set of long term week long benches in a third party shop roundup exam, chances are it's going to run linux OK. It's the nontested, noncompared mobos you usually have to worry about. Just 'cause a board garners rave reviews with Windoze, there's no guarantee (MOF, just the opposite more'n more lately) it'll work as well or be as supported by the OS, as it is by the Win-tel/nvidia/ ati/creative/dell, etc. gang. EG, most all of Tom's boards incorporate Win-fake-raid. While this is true, it is also true that the HPT37X IDE controllers have in large part gotten a bad rap for nothing, for several reasons. First, the HPT37X is an IDE controller, first and foremost, and needs to be examined in that light before anything else. Second, most Linux peeps know that anyone who wants to run Raid needs to be on soft raid anyway, no matter what whiz-bang dedicated RAID hardware they have. Third, in all the benchmarks I've seen that compare conventional IDE channels and the HPT37X IDE channels, the Highpoint controllers have always been superior in performance. (AS an IDE controller.) Fourth, the HPT37X controller, while it can be criticized for being a raid faker, cannot be criticized for not providing better performance under winblows with it's, eh, pseudo raid (as compared to single drive performance). The CPU utilization is lower and the throughput is higher. Bios/driver driven pseudo raid, maybe; but the benches are not arguable. And finally, from a personal standpoint, since running this Abit board, it's my opinion that a Linux user can get better performance from soft raid running from an HPT37X controller than they can get from the standard IDE controllers. That's why I've been getting nothing but mobos with HPT37X-RAID controllers on them since they came out; not because of the superior winblows performance they provide, which doesn't mean a hill of beans to me. But because the HPT37X IDE controllers are just better IDE controllers. Finally, from what I understand, the Highpoint controllers are now usable under Linux in their native RAID modes, via (no pun intended) the latest kernel module code. Not that I'm interested meself; I'm staying with soft raid. Still, the chance is there for anyone interested to run the numbers. Also, preliminary reports I've seen are that ddr333 offers none, to very little improvement over 266 (much the same as with ata/133 over /100). One advantage tho is if you get marginal ddr333, it might run as ddr266 at cas2 (bettr'n 333 at CL2.5). So buy quality (micron, infineon) ddr333 now while it's cheap, wait for the (Linux) dust to settle on the rest of the cuttin' edge hardware. -- Tom BrinkmanCorpus Christi, Texas Has someone done any roundup reviews of mainboards with Linux, I wonder? I haven't come across any when I've been cruising for comparative mobo evaluations; I'd be glad for any info regarding that! TIA, LX -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On 14 May 2002 18:52:53 -0400, Lyvim Xaphir [EMAIL PROTECTED] wrote: On Tue, 2002-05-14 at 10:29, Carroll Grigsby wrote: LX: Skimming through the article, I noticed that some of these boards have gone to a USB-only configuration -- no serial or parallel ports. Given that these boards are aimed at end-users and not OEM's, this does not sound like a Really Good Idea to me. In my case, it would mean adding the cost of a new printer and a new modem to the overall project cost, and those are not trivial costs. Besides, my present modem and printer work just fine, thank you. -- cmg Ahhh...Carroll, my friend. :) That is the beauty part. If indeed you do contemplate an upgrade, all you need do is get a usb serial/parallel adapter for a modest sum, and presto, your printer and modem are back in business. In addition, having done that you can do as I did and use IRQ's 7, 3, and 4 for things other than being tied to onerous dated and obsoleted onboard ports. Obsoleted in the sense that their functions are now duplicated on a less resource intensive external bus. How is the CPU usage of USB compared to the 'legacy' ports (serial, parallel, PS/2, etc.)? The legacy ports were designed for older systems, so they cannot suck up too much juice. I hear that USB, on the other hand, is a real pig in this regard (no surprise that Intel supports it). If that is the case, is it really worth using USB peripherals on a PC when legacy types would suffice? Here, I am referring to simple components that have little to gain from USB, like keyboards and mice. I don't want my keyboard and mouse slowing down my system :) -- Sridhar Dhanapalan Never over-design. Never think Hmm, maybe somebody would find this useful. Start from what you know people _have_ to have, and try to make that set smaller. When you can make it no smaller, you've reached one point. That's a good point to start from - use that for some real implementation. -- Linus Torvalds Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tue, 14 May 2002 18:02:12 -0500, Tom Brinkman [EMAIL PROTECTED] wrote: And ... I believe it's not quite time to upgrade an AMD system till the Thorobreds come out. Are the Thoroughbreds really that good? From what I have read, they are not much more than Palomino cores shrunk down to 0.09 microns. If you want a cache improvement, you'll have to wait for Barton, which is due to be released (according to AMD's roadmap) very close to Hammer-time (sorry -- Opteron-time). Clock speeds won't increase by very much, but I'm sure the shrunken die will make overclocking a bit safer. -- Sridhar Dhanapalan Jennifer Lopez = JLo (pron: Jay-Low). Bill Gates = BGa (pron: Be-Gay). Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tuesday 14 May 2002 11:19 pm, Sridhar Dhanapalan wrote: On 14 May 2002 18:52:53 -0400, Lyvim Xaphir [EMAIL PROTECTED] wrote: On Tue, 2002-05-14 at 10:29, Carroll Grigsby wrote: LX: Skimming through the article, I noticed that some of these boards have gone to a USB-only configuration -- no serial or parallel ports. Given that these boards are aimed at end-users and not OEM's, this does not sound like a Really Good Idea to me. In my case, it would mean adding the cost of a new printer and a new modem to the overall project cost, and those are not trivial costs. Besides, my present modem and printer work just fine, thank you. -- cmg Ahhh...Carroll, my friend. :) That is the beauty part. If indeed you do contemplate an upgrade, all you need do is get a usb serial/parallel adapter for a modest sum, and presto, your printer and modem are back in business. In addition, having done that you can do as I did and use IRQ's 7, 3, and 4 for things other than being tied to onerous dated and obsoleted onboard ports. Obsoleted in the sense that their functions are now duplicated on a less resource intensive external bus. How is the CPU usage of USB compared to the 'legacy' ports (serial, parallel, PS/2, etc.)? The legacy ports were designed for older systems, so they cannot suck up too much juice. I hear that USB, on the other hand, is a real pig in this regard (no surprise that Intel supports it). If that is the case, is it really worth using USB peripherals on a PC when legacy types would suffice? Here, I am referring to simple components that have little to gain from USB, like keyboards and mice. I don't want my keyboard and mouse slowing down my system :) Thanks for the very interesting replies. I didn't realize that there such things as USB to serial/parallel adapters. Most of my experience with I/O cards goes back to the Real Old Days, when everything was done with add-in ISA cards. (IIRC, the original IBM PC required a separate card for each function.) I suppose that there are also PCI cards that could be used, too, but that gets us back to the IRQ mess, doesn't it? Gotta think about this one, but there's a lot of time. First, I'll have to convince my wife that she needs to start delivering newspapers so that I can buy a new rig. Predictions: Lots of postings to this and other mail lists about configuration problems; much more severe on Windows-oriented lists. Pigs sighted flying over Raleigh. -- cmg Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
On Tue, 2002-05-14 at 23:19, Sridhar Dhanapalan wrote: Ahhh...Carroll, my friend. :) That is the beauty part. If indeed you do contemplate an upgrade, all you need do is get a usb serial/parallel adapter for a modest sum, and presto, your printer and modem are back in business. In addition, having done that you can do as I did and use IRQ's 7, 3, and 4 for things other than being tied to onerous dated and obsoleted onboard ports. Obsoleted in the sense that their functions are now duplicated on a less resource intensive external bus. How is the CPU usage of USB compared to the 'legacy' ports (serial, parallel, PS/2, etc.)? The legacy ports were designed for older systems, so they cannot suck up too much juice. It is interesting that you bring up the CPU usage aspect of USB, as I've not seen any numbers regarding that. The resource benefits I was referring to was directed towards IRQ usage, because I have a real need for extra IRQ's (and I suspect others with legacy hardware do also), so in that regard the USB bus was a real benefit. There's a lot you can do with three extra IRQ's. I am interested in CPU utilization numbers, if you've got some URL's... I hear that USB, on the other hand, is a real pig in this regard (no surprise that Intel supports it). If that is the case, is it really worth using USB peripherals on a PC when legacy types would suffice? That begs the question of some numbers, does it not? Such as, what exactly is required for the device in question? (is it an isdn modem, or) Not only that, but in order to answer your question properly you also need some performance numbers in regard to the chipset platform that is giving you the USB functionality. Quite naturally, this is going to be giving different performance results depending on which chipset platform you choose; i.e., Intel or Via. Performance in this context meaning both CPU utilization and USB bus throughput. Here, I am referring to simple components that have little to gain from USB, like keyboards and mice. I don't want my keyboard and mouse slowing down my system :) -- Sridhar Dhanapalan I've never been an advocate of moving primary user IO off the legacy ports. You are removing redundancy and putting all your eggs in one basket, cause if the USB bus goes you lose it all. I know you've got a little tongue in cheek here with the mouse/keyboard thing, but just for the heck of it I'll run the numbers on it. There's not much of a chance that a 1200 bits/sec mouse serial stream and a character stream from some human hands at the macro level can influence a bus that's been benched at up to 5.7 MB/sec (BTW, the USB bus has been *marketed* as a 12 MB/sec bus. 5.7 is the best I've seen in RL, tho). So you've got a mouse that communicates at 150 bytes/sec on a bus that has been RL evaluated at 5.7 Megabytes/sec; that's 38,000 times more bandwidth than the mouse requires. This is an invalid evaluation however, because I still don't know how fast you can type. ;) L8r, LX -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] New Mobo Roundup--Dr Tom
How good are the integrated devices on these boards, compared to the ones you can buy as add-in PCI cards? Tom's reviews never seem to give them much attention. How do they perform? Do they tax the CPU any more than a standalone card would? I am interested in getting a board with integrated sound and ethernet. I would like the board to have at least 4 IDE interfaces (not channels) so I can give an entire interface to each of my devices. How well do the features of these boards work in Linux? Based on these, can anyone make a board recommendation? I don't want to spend too much money, and that is the reason why I'm going for integrated peripherals. They need to be decent, though. -- Sridhar Dhanapalan The only intuitive interface is a nipple. After that, it's all learned. Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
[expert] New Mobo Roundup--Dr Tom
Guys, if you're in the market for some new technology for under your hood, I strongly suggest that you check out the 18 mainboard lineup that just came out as of May 9th on Toms: http://www.tomshardware.com/mainboard/02q2/020509/index.html These boards are using the new VIA Apollo KT333 chipset; its the one I've been waiting for, so I could upgrade. There is one board that evidently came out too late for their tests, which took over two weeks. It's one I'm personally watching: http://www.abit.com.tw/abitweb/webjsp/english/products_shortcut.jsp?pPRODUCT_TYPE=MotherBoardpMODEL_NAME=KX7-333R However, no matter how good the Abit board looks on Abit's home site, I'm not going to buy it until it's been reviewed in a roundup. I'm in the market for a mobo upgrade. As it sits, judging from the roundup, I'm looking strongly at either the Epox 8k3a+, or the Enmic 8TTX2+. The Gigabyte got a strong review, but if a mobo doesn't have at least 7 slots, it's not interesting. Overclocking features are important too, which both the Epox and Enmic have. I wish he'd go ahead and update this roundup with the Abit KX7-333R, so I could finalize a decision. L8r, LX -- °°° Kernel 2.4.8-26mdk Mandrake Linux 8.1 Enlightenment 0.16.5Evolution 1.02 Registered Linux User #268899 http://counter.li.org/ °°° Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com