Sign in with
Sign up | Sign in
Your question

My First Build - Pls Rate this Rig!! (part II)

Last response: in CPUs
Share
February 14, 2002 2:26:57 PM

OK I've made several revisions based on suggestions from the forum. Thanks to everyone that replied!


Athlon XP 1700+
Volcano 6cu+ fan 7000 rpm 38 CFM (and Actic Silver II)
Leadtek 7350KDA (SiS735 chipset) w/ onboard LAN and sound
512 meg DDR from Kingston

HD - Maxtor 7200rpm - 30gig or 40 gig
CDRW ??? I dont care about write speed really, only READ speed (maybe the Aopen 16/10/40 ?)
3d card - Gainward GF3 Ti200
FDD - Mitsumi FDD (does it really matter? lol)
Modem - Creative Modem Blaster 56k
Case - Maxtop ATX case and 350W PSU and 3 case fans
Logitech Optical Mouse
Windows XP

Using my old speakers, keyboard, cable modem and monitor. This is for gaming and internet, and mild overclocking (thats my next project!). I am trying to get the best price/performance as well as easy to install.

Comments/suggestions/flames etc are all welcome!

thanks!

----
A newbie is only a newbie for as long as you allow him to be.
-Anonymous Veteran
February 14, 2002 2:35:51 PM

Quote:
Leadtek 7350KDA (SiS735 chipset) w/ onboard LAN and sound

im not too sure about that motherboard (in general that is)...but its probably a good one.
Quote:
512 meg DDR from Kingston

you sure you dont want crucial ram?...kingston is good and all...but its not the best for overclocking....but if you're not an overclocker...then kingston is fine...
Quote:
CDRW ??? I dont care about write speed really, only READ speed (maybe the Aopen 16/10/40 ?)

you will care about the write speed...plextor's or lite on's are good, cheap burners.
everything else seems fine...

<A HREF="http://gamershq.madonion.com/compare2k1.shtml?2649487" target="_new">P4 NW + DDR</A> = <font color=blue>Not Bad</font color=blue>
February 14, 2002 3:11:11 PM

Have you read any posts on overclocking XP's? you might want to try a different motherboard (unless you are just doing FSB). I heard the Epox 8kha+ should be ok for small overclocking projectsi agree plextor would be the best way to go with the burner, a little more expensive, but should be worth it
Related resources
February 14, 2002 4:38:37 PM

yeah, I know not many people have/are familiar with the Leadtek board, but everyone that I have talked to that does have it, loves it. Theres a Leadtek forum at ocworkbench and anadtech too I think, and they all say its fast, rock stable, oc-able, and easy to install. Faster than some KT266A boards, and only 1-3% slower than the best ones that cost more. I was getting that EPOX board all the way, up until about a week ago, but bumped it for the Leadtek. I think both are a good choice, but I keep hearing about the mysterious "VIA bugs" so I said screw it, get the SiS from Leadtek. There is too much info for me to weed through as it is, being a newb and all.

As for overclocking, no I haven't read too much about it yet. Just too much info to digest as it is. I just want the option in the future, and the Leadtek (with the latest BIOS) is said to be the most overclockable SiS735 board. The Gainward is said to be EASILY overclockable, and the kingston memory, well I just learned its *not* very overclockable thanks to pr497. Thanks man! I think I might be ok with that though, since better memory is being released almost every month nowadays. DDR333 (or pc2700? whaddevadahell you call it, why must DDR names be so confusing? Isn't pc2100 really pc166 DDR SDRAM?) Anyway, I might wait for that to drop in price and upgrade in a year when half a gig isn't enough for me anymore anyway, lol. And I think the Leadtek board supports DDR333 too, albeit unofficially.

----
A newbie is only a newbie for as long as you allow him to be.
-Anonymous Veteran
February 14, 2002 4:57:02 PM

No sound card?

You might look for Kingmax RAM, I've heard unbelieveable stories about their PC2100 OCing ability.

Oh, and once again I recommend against that HSF. This time, because it's not the best performer. Since you're overclocking, look for a bare Alpha Pal8045 ($30?), and possibly a Panaflo fan.

<font color=orange>Quarter</font color=orange> <font color=blue>Pounder</font color=blue> <font color=orange>Inside</font color=orange><P ID="edit"><FONT SIZE=-1><EM>Edited by FatBurger on 02/14/02 10:58 AM.</EM></FONT></P>
February 14, 2002 5:15:58 PM

Just one comment:

Get Crucial RAM! Go to www.crucial.com and check out there price and order some. I have heard there is a 15% discount for anandtech users also... might be worth checking it out.

Crucial memory is very overclockable, and I have never heard of any problems when people increase the memory timing to 2-2-2 with it.
Anonymous
a b à CPUs
February 14, 2002 7:31:33 PM

CDRW:
get smt with 8mb buffer.
FDD:
they are all the same
February 14, 2002 7:33:31 PM

hmm, can't find the discount at anandtech, anybody else hear of this?

fatburger - easy solution on the sound card - build it for now and later if I hate the onboard sound I'll get a sound card (and I know who to ask ;)  otherwise stick with what I got.

sad to hear on the hsf :(  I'm sure that Alpha is great, but I bet its 45-50 dBA too (I think, kinda guessing on that). I was hoping for a happy medium between the cheap stock hsf's and the jet turbine Alpha/Swiftech's. The Volcano 6cu+ pushes 38 CFM at 39 dBA. Is 38 CFM enough? I don't need to OC like a madman.

----
A newbie is only a newbie for as long as you allow him to be.
-Anonymous Veteran
February 14, 2002 8:00:29 PM

actually if u buy from crucial.com you automatically receive 15% off..and about the alpha8045, its just the heatsink..what makes the noise level go up is the fan..you can chose which fan to throw on top of that. i have the alpha myself, great heatsink and on top of that i have a delta fan..very loud

didnt have one of em electronic pens so ill just type my name,<i>CoOoLMaNX</i>
February 14, 2002 8:53:34 PM

Crucial does rule. If you are not OCing the Kingston is excellent also, never got a bad stick from either company and at stock speeds never had problems....

Jesus saves, but Mario scores!!!
Anonymous
a b à CPUs
February 14, 2002 9:28:17 PM

Quote:

(...) I keep hearing about the mysterious "VIA bugs"

Please tell me more, I was just going for a mb w/ via kt266a. I've read some posting about it in mb/chipset, but that got refuted/flamed;) so I didn't take any notice.
Where did you read about this?


<A HREF="http://uk.indymedia.org/front.php3?article_id=6047" target="_new">my pacifist summer holiday</A> :frown:
February 14, 2002 10:36:14 PM

That's why I'm saying you get the BARE heatsink (no fan), and get a Panaflo fan to go on it.

<font color=orange>Quarter</font color=orange> <font color=blue>Pounder</font color=blue> <font color=orange>Inside</font color=orange>
February 14, 2002 10:58:32 PM

Quote:
:(  I'm sure that Alpha is great, but I bet its 45-50 dBA too (I think, kinda guessing on that).



No, the alphapal8045 is an 80mm heatsink, and it uses 80mm fans, which are much quieter than 60mm fans, I just bought one, and with an 80mm sunon on their, temps are frosty and its nice and quiet.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 14, 2002 11:00:30 PM

Quote:
Please tell me more, I was just going for a mb w/ via kt266a. I've read some posting about it in mb/chipset, but that got refuted/flamed;) so I didn't take any notice.
Where did you read about this?



The kt266a chipset is stable so far.

When people say they dislike via, or distrust via, it is due to the bugs which occured with the kt133 and kt133a chipsets. Via's current platform dosent have any bugs that I am aware of, but old mistrust dies hard.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 15, 2002 12:23:54 AM

as far as i can see the 7350kda is great for overclocking as it has the 1/5 and 1/6 dividers for high fsb/ram speeds.

i want one too, but impossible to find here
plus, im kinda broke :( 

the idea is you unlock your processor, reduce the multiplier, set the fsb to 166 then install some PC2700 ram and let rip!
out of all the types of PC2700 corsiar looks the best i think, tested and rated for Cas2

Overclocked athlon 1200C @ 8.5 x 166FSB + PC2700 = GOOD! :smile:
February 15, 2002 3:53:54 AM

Quote:
Via's current platform dosent have any bugs that I am aware of, but old mistrust dies hard.

Have you not heard of the pci bus limitations that all via chipsets have? with faster hard drives and raid controllers becoming the norm, this is becoming a real issue.

It's not what they tell you, its what they don't tell you!
February 15, 2002 5:34:56 AM

Oh, you mean the same bug found in this chipset?

<A HREF="http://www.theinquirer.net/17010203.htm" target="_new">http://www.theinquirer.net/17010203.htm&lt;/A>


SO I suppose its ok never to buy intel motherboards either, because of some bugs?


BTW.

<A HREF="http://www.theinquirer.net/01020217.htm" target="_new">http://www.theinquirer.net/01020217.htm&lt;/A>


The bug you mentioned has been patched and repaired...so, as I said before, the current kt266a chipset has no bugs that I am aware of.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 15, 2002 6:09:47 AM

The fan on the <A HREF="http://www.dansdata.com/coolercomp.htm" target="_new">6cu+</A> (note the plus) just makes more noise. It's no better than the 6cu, which I have and really recommend.

<font color=red>God</font color=red> <font color=blue>Bless</font color=blue> <font color=red>America!</font color=red>
February 15, 2002 10:06:30 AM

The bug has existed ever since the mvp3 chipset. Not until Tec channel posted conclusive evidence that such a bug existed did VIA own up to and "patch" the bug. Can you say quality control? How can you muff up something as standard as a PCI bus and not find it for years? Remember the creative incidence? ( this is when a pci bandwith hog meats a pci bandwith limited chipset). at least Intel posted there bug from the get go in their white pages. supposedly VIA didn't even know about it? What else don't they know about?

It's not what they tell you, its what they don't tell you!
February 15, 2002 10:14:12 AM

LoL, like I said, via's current chips are bug free as far as I know. I am not defending via, but its funny how people villify them for things which even intel does.

Via has a bad history, but their CURRENT chip does not have any unpatched bugs that I am aware of.

Thats all I said, and you claimed there was a bug, I showed there was a patch, and threw the intel thing in to show that the issue was not via specific.

Its ok not to like via, but not to like their current chipsets based on their past chipsets problems dosent make much sense to me.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 15, 2002 12:43:31 PM

LOL, you see? this is why I am getting the Leadtek. You guys all make good arguments and god knows I can't refute any it cuz my brain is just too full as it is to try to make sense of it all.

Back to hsf's, so the panaflo is quiet? k, didn't know that, where can I get one? Also, what do other people use? Any other suggestions?

BTW, the link dhlucke posted is great. Its a monster article on every hsf imaginable. Thanks man. I'm reading through it right now.

----
A newbie is only a newbie for as long as you allow him to be.
-Anonymous Veteran
February 15, 2002 1:28:48 PM

All,

That article unfortunately doesn't provide a convenient summary of all the hsf's they tested, other than a long alphabetical list down the side, so I copied and pasted that into excel, isolated the thermal resistance numbers and sorted it.

Here are the results. Lower numbers are better.


Swiftech MCX462 0.52°C/W
OCZ Dominator 0.53°C/W
Alpha PAL8045 (variable speed fan, maximum power) 0.56°C/W
Global Win CAK38 0.56°C/W
Global Win CAK38 (updated) 0.56°C/W
Thermalright SK-6 with 7000RPM fan 0.56°C/W
Bitspower NP60CD 0.57°C/W
Bitspower NP80D 0.57°C/W
OCZ Gladiator 0.57°C/W
OCZ Goliath with very high power fan 0.57°C/W
Swiftech MCXC-370 0.57°C/W
ThermoSonic ThermoEngine with YS Tech fan 0.57°C/W
Cooler Master HHC-001 0.58°C/W
Swiftech MCX370-OA 0.58°C/W
Thermal Integration TI-V77L with high power 70mm fan 0.58°C/W
Vantec FCE-6030D 0.58°C/W
OCZ silver plated Gladiator 0.59°C/W
Thermaltake Volcano 6Cu+ 0.59°C/W
TS Heatronics Zen CPU Radiator 0.59°C/W
Vantec CCK-6035D 0.60°C/W
Global Win FOP38 0.61°C/W
OCZ Dominator 2 (Y.S. Tech fan) 0.61°C/W
Sibak Tech AE-01-625B 0.61°C/W
Thermalright CB-6L (Y.S. Tech 7.2W fan) 0.61°C/W
Thermaltake Volcano 6Cu 0.61°C/W
Thermaltake Volcano 7 (full power) 0.61°C/W
Cooler Master EP5-6ID2 0.62°C/W
Power Cooler PCH137 0.62°C/W
Thermal Integration TI-V77L with 80mm fan 0.62°C/W
Thermalright SK-6 with 4200RPM fan 0.62°C/W
Thermaltake Volcano 5 0.62°C/W
ThermoSonic ThermoEngine with Sunon 80mm fan 0.62°C/W
Zalman CNPS5000-Plus 0.62°C/W
Ideal Elethermal CBCF-AMD-3 0.63°C/W
Tiger Cooler (Sunon 4.3W fan) 0.63°C/W
Bitspower NP60CS 0.64°C/W
Global Win WBK38 0.64°C/W
Sibak Tech AC-02-625B 0.64°C/W
Bitspower NP61D 0.65°C/W
TaiSol CGK760 0.65°C/W
Thermaltake Dragon Orb 3 0.65°C/W
Alpha PAL6035 0.66°C/W
Bitspower NP60D 0.66°C/W
Vantec FCE-6254OD 0.66°C/W
Ideal Elethermal CBCF-AMD-1 0.67°C/W
Power Cooler PCH075 0.68°C/W
Alpha PEP66 0.7°C/W
Kanie Hedgehog 238M 0.7°C/W
"FYH" cooler 0.70°C/W
"Galaxy" cooler 0.70°C/W
Global Win FOP32-1 0.70°C/W
Power Cooler PCH113 0.70°C/W
Power Cooler PCH123 0.70°C/W
Thermal Integration TI-V77L 0.70°C/W
Spire 5E34B3 0.71°C/W
Thermal Integration TI-V86 0.71°C/W
Thermalright CB-6L, stock fan 0.71°C/W
Thermaltake Super Orb 0.71°C/W
Tiger Cooler (temperature controlled fan, maximum power) 0.71°C/W
Bitspower NP60E 0.72°C/W
Power Cooler PCH141 0.72°C/W
Thermaltake Mini Copper Orb 0.72°C/W
Spire 5E32B3 0.73°C/W
Zalman CNPS-6000Cu (full power) 0.73°C/W
Spire 5P53B3 0.74°C/W
Thermal Integration TI-S86 0.74°C/W
Thermaltake Dragon Orb 1 0.74°C/W
Thermaltake Volcano II 0.74°C/W
Zalman CNPS3100G 0.76°C/W
Zalman CNPS-6000AlCu (full power) 0.77 °C/W
ThermoSonic ThermoEngine 0.77°C/W
Glacial Tech Igloo 2300 0.78°C/W
Just Cooler 80mm cooler 0.78°C/W
Cooler Master EP5-6151-A1 0.79°C/W
OCZ Glacier2 0.79°C/W
Thermaltake Golden Orb (twist clip) 0.8°C/W
Global Win FNP-50 0.80°C/W
Just Cooler P-5500 0.80°C/W
Thermaltake Chrome Orb 0.81°C/W
Arctic Cooling Super Silent 2500 0.82°C/W
Cooler Master DP5-6H51 (new) 0.82°C/W
Thermaltake Mini Super Orb 0.82°C/W
Cooler Master DP5-6H51 (old) 0.86°C/W
Just Cooler P-1000 0.87°C/W
"COOC" cooler 0.88°C/W
Sibak Tech AC-01-625B 0.88°C/W
Arctic Cooling Super Silent 2000 0.90°C/W
Bitspower NP15S 0.92°C/W
Bitspower NP13D 0.94°C/W
Just Cooler P-925 0.94°C/W
Alpha FC-PAL15 0.96°C/W
Zalman CNPS3100G, "silent mode" 0.98°C/W
Thermaltake TFCF009 "Harp" 1.13°C/W
Thermaltake Mini Golden Orb 1.17°C/W


I can clean that up in table format if someone can remind me how, I used to know :(  but forgot. Its like ]PI[ or something (brackets facing the other way of course).



----
A newbie is only a newbie for as long as you allow him to be.
-Anonymous Veteran
February 15, 2002 2:28:48 PM

Ok, let's get this straight. The heatsink/fan is made up of two parts, the <b>heatsink</b> and the <b>fan</b>. Now, this is the important part: YOU CAN BUY THE TWO PARTS SEPARATELY. Dammit, I can't even spell today :tongue:

Anyhow, get the Alpha PAL8045 (<font color=red>with no fan</font color=red>), and then buy a fan to put on it. Panaflo makes the best fans.

<font color=orange>Quarter</font color=orange> <font color=blue>Pounder</font color=blue> <font color=orange>Inside</font color=orange>
February 15, 2002 4:40:20 PM

Quote:

YOU CAN BUY THE TWO PARTS SEPARATELY

I know, I was asking where I can buy a panaflow fan? Newegg has the alpha heatsink, but I don't see a panaflow fan on their site.

edit: nm, found it. its under Panasonic.

----
A newbie is only a newbie for as long as you allow him to be.
-Anonymous Veteran <P ID="edit"><FONT SIZE=-1><EM>Edited by endless_n00b on 02/15/02 02:13 PM.</EM></FONT></P>
February 15, 2002 5:11:51 PM

Try <A HREF="http://plycon.com" target="_new">plycon.com</A>, they have a pretty good selection of heatsinks and fans. You can add different fans to whatever HSF you buy.

For memory, I'd consider going with Corsair. [H]ardOCP, which is a pretty respected site, has named it best memory two years in a row. I push mine pretty hard and haven't had any problems with it.

<A HREF="http://www.hardocp.com/articles/hardestof21k/" target="_new">[H]ardOCP Awards</A>

Anyway, that's my $0.02 ...

<i>I made you look. But I can't make you see.</i>
February 15, 2002 5:49:57 PM

To my knowledge the via "patch" only addresses specific brands of IDE raid controllers and not the issue itself. What about the various SCSI controllers or sound cards or anything else that uses the pci bus? Pointing out that Intel has one chipset that has simular problems does not justify the ignorance VIA has shown in their design of the pci bus.....sorry it ain't going to cut it, not in my book anyway. Intel had the limnitations of thier bus documented for all to see before it went to market, which is a much different scenerio than the one with VIA. And its still a bug, one that needs to be addressed by a workaround. A "fix" would be the introduction of a southbridge that didn't require the patch.

It's not what they tell you, its what they don't tell you!
February 16, 2002 12:08:43 PM

Quote:
To my knowledge the via "patch" only addresses specific brands of IDE raid controllers and not the issue itself. What about the various SCSI controllers or sound cards or anything else that uses the pci bus? Pointing out that Intel has one chipset that has simular problems does not justify the ignorance VIA has shown in their design of the pci bus.....sorry it ain't going to cut it, not in my book anyway. Intel had the limnitations of thier bus documented for all to see before it went to market, which is a much different scenerio than the one with VIA. And its still a bug, one that needs to be addressed by a workaround. A "fix" would be the introduction of a southbridge that didn't require the patch.


Well, you are wrong lol.


The patch fixxes the pci bus quite a bit.


Quote:
Increase in burst transfer rate is now 20 to 30MB/s with all the cards it is tested.

That brings the Via PCI bus up to the speed of Intel chipsets, says the site.



Where do you get, specific cards, it affects all pci bus using parts.


Again, I stress, I am not being some via lover, but the current kt266a chipset HAS NO UNPATCHED/REPAIRED BUGS THAT I AM AWARE OF.


Complaining and saying via sucks for taking years to fix the issue does not change the fact they fixed it. Period.


"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 16, 2002 3:10:16 PM

Quote:
Complaining and saying via sucks for taking years to fix the issue does not change the fact they fixed it. Period.

Maybe maybe not. It sures makes one question the abilities of a chip maker that took this long to first discover and then address it.


It's not what they tell you, its what they don't tell you!
February 16, 2002 3:17:02 PM

Taken from Tec channel:
Quote:
According to VIA, the new version of their patch will increase performance with RAID cards from ACARD, Adaptec and Highpoint and not only with Promise-cards anymore. We did a few test with all these cards. In addition to that, we also checked the new patch with the Advance 29133, another Ultra-ATA/133-PCI-Controller, which is not on VIAs list yet.

As you can plainly see it only mentions Ide raid cards and does absolutely no testing with any other devices such as SCSI raid, Sound cards, etc etc. Maybe it helps maybe it doesn't. Until you see such results I would be wary of claiming that the patch is a fix all. I would love to see how this patch effects users of the 686b and sound blaster live coupled together.

Now I must add that I have used via chipsets in the past. This really is only an issue to some users with bandwith demanding pci cards. But this is one of the main reasons that you struggle to see VIA ( and in large part AMD) penetrate the workplace. What other "bugs" lurke in there hardware and how long will it take to address them? If I by a KT333 board and an issue is discovered will they take 2-5 years to fix it as well?

Is it the fact that this "bug" existed the largest cause for concern or it it the fact that they reportably did not even know about it that should be the main cause for concern? The latter is what really bothers me!

It's not what they tell you, its what they don't tell you!
February 17, 2002 1:11:29 AM

Quote:
As you can plainly see it only mentions Ide raid cards and does absolutely no testing with any other devices such as SCSI raid, Sound cards, etc etc. Maybe it helps maybe it doesn't. Until you see such results I would be wary of claiming that the patch is a fix all. I would love to see how this patch effects users of the 686b and sound blaster live coupled together.



LoL, they fixxed the pci bus bandwidth problem, it will affect ALL cards, EVERY card they tested WORKED fine.

Stop assuming scsi wont work, or anything for that matter.


There is NO current bug in the kt266a chipset.


Quote:
Is it the fact that this "bug" existed the largest cause for concern or it it the fact that they reportably did not even know about it that should be the main cause for concern? The latter is what really bothers me!



THe reason they didnt know about it is because it never affected performance untill now, the pci bandwidth was plenty high for 99% of pci cards, but high end ide raid would be some of the first things that did not have enough bandwith.


Quote:
What other "bugs" lurke in there hardware and how long will it take to address them? If I by a KT333 board and an issue is discovered will they take 2-5 years to fix it as well?



And that sentament is why I provided the link to intels bug, intel is regarded as the best chipset maker, and if they have the SAME bug as the via chipsets, then either intel is as bad as via, or via is not as bad as everyone says.


Thats my sentament in a nutshell.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 17, 2002 1:13:45 AM

Quote:
As you can plainly see it only mentions Ide raid cards and does absolutely no testing with any other devices such as SCSI raid,



The only reason they mention ide raid cards, is because those were what they had problems with before the patch, if a chipset had a bug which causes photoshop 6 to crash, and the chipset manufacturer released a patch, would you not expect them to test it with photoshop? Would you then say that because they didnt test other photo applications, that the patch didnt work?

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 17, 2002 2:30:59 AM

Quote:
LoL, they fixxed the pci bus bandwidth problem, it will affect ALL cards, EVERY card they tested WORKED fine.

Stop assuming scsi wont work, or anything for that matter.

Actually it is you that made the assumption that the board you got had no issues when in fact it did. You should be carefull to make further assumptions as to it now being bug free. Can you say sound blaster and 686b? If it has fixed the entire pci bandwith problem then by all means show me the proof? If not stop making assumptions!

Quote:
THe reason they didnt know about it is because it never affected performance untill now, the pci bandwidth was plenty high for 99% of pci cards, but high end ide raid would be some of the first things that did not have enough bandwith.

Thats a load of Bullcrap! High end raid cards have been around for some time, maybe not IDE raid but none the less they have been around. Your explanation of it just being discovered in unexceptable....PERIOD!!! After all the PCI bus is a standard dammit!

Quote:
And that sentament is why I provided the link to intels bug

Stop bringing this up as it is a none issue. Intels "bug" was well documemted before it was released. Plain and simple.

It's not what they tell you, its what they don't tell you!
February 17, 2002 5:09:05 AM

Quote:
Actually it is you that made the assumption that the board you got had no issues when in fact it did. You should be carefull to make further assumptions as to it now being bug free. Can you say sound blaster and 686b? If it has fixed the entire pci bandwith problem then by all means show me the proof? If not stop making assumptions!


A: I dont own a kt266a board.
B: I dare you to find a single card to which this patch did in improve bandwith.
C: the patch inproved pci bandwith, that is NOT variant on the card being used, it is pure bandwith. If pre patch the pci bus could only transmit 70MB/sec, and the patch raised that to 100MB/sec, that gain is for EVERY pci card, not just ide, and not just specific brands of it.

PROOF

<A HREF="http://www.networking.tzo.com/net/software/readme/vlate..." target="_new">http://www.networking.tzo.com/net/software/readme/vlate...;/A>


Quote:
Copy large files to and from multiple hard disks while playing sound files. The files should be at least 100MB in size. If the computer freezes during such activity when this driver is absent, but works when the driver is present, then the driver is working.

Fixxes the sblive/datatransfer issue.

<A HREF="http://www.tecchannel.de/hardware/817/7.html" target="_new">http://www.tecchannel.de/hardware/817/7.html&lt;/A>
Quote:
After we installed the unofficial latency patch the effect described completely disappeared. Whether we played one single note with a number of instruments or went through complex songs with up to 64 note polyphony and dozens of instrumenst - not a single note was wrong

Cleared up pci transfer erros to a professional soundcard.


So please tell me, how a patch specifically aimed at IDE raid cards helps sound cards, I would appreciate it.
Quote:
Actually it is you that made the assumption that the board you got had no issues when in fact it did.

I said HAS not HAD, there is a HUGE difference between what you think I said, and what I did say.
Quote:
Thats a load of Bullcrap! High end raid cards have been around for some time, maybe not IDE raid but none the less they have been around.Your explanation of it just being discovered in unexceptable....PERIOD!!!


HIgh end raid for ide, is new.
HIgh end raid for scsi wouldnt go into a via board, scsi raid is a SERVER application, and most servers are NOT via.
I can easily fathom how an issue like this could go undected for several years, and it is no coincidence that several months after ide raid is introduced, and it becomes mainstream the problem is detected.

Quote:
After all the PCI bus is a standard dammit!

Intel screwed up their pci bus too, its a standard damnit! Now, what the hell does pci being a standard have to do with anything? Again, like I said, the bug is patched, the via PCI bus now performs on par with other chipsets. The time it took to patch the bug does NOTHING, to discredit my statement that AT THIS TIME, THERE ARE NO BUGS WITH THE KT266A THAT I KNOW OF.

Quote:
Intels "bug" was well documemted before it was released. Plain and simple.

If this bug was doccumented before it was released, umm, why didnt they fix it? Would you not consider that a far worse offence than not knowing about a bug at all?




"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 17, 2002 2:52:21 PM

Why are you linking to George Breese's unofficial patch? Were we not discussing the official VIA patch??

Taken from your link:
Quote:
This product could, in a rare case, damage a computer. Use it carefully and at your own risk.

Yes I agree George Breese's patch does indeed seem to affect more than just IDE Raid cards and there is documentation as such. However, this patch is unsupported by VIA and was not even developed by them ( another embarrasment for VIA ).

Now as for the official patch, if you remember correctly VIA first released a patch aimed at promise cards singularly. There next version of the patch then added support for other IDE Raid cards, and in thier readme this is all they claim it is a fix for. Does it cure other pci bandwith problems as well? Perhaps, but until you link to documention stating such, I would be carefull to make claims otherwise.

Your reference as to VIA never being deployed in a server enviorment is laughable, but hey, I agree. So in effect what you are saying is that VIA is in effect sub-standard. I wonder if they place such a disclaimor on all their motherboard chipsets " Warning, this product is not up to the standards required in any server enviorment".

You keep bringing up the Intel I850 in defense of VIA. Sorry, thats a no go. I can just imagine you as a trial lawyer, your defence would go like this:

"Your honor, it is true my client did in fact kill the defendent, but look, Mrx. also has committed murder so then it must be ok" Just doesn't work does it?


It's not what they tell you, its what they don't tell you!<P ID="edit"><FONT SIZE=-1><EM>Edited by ncogneto on 02/17/02 12:09 PM.</EM></FONT></P>
February 17, 2002 4:05:01 PM

Quote:
Tests / VIA Chipsets slow down PCI cards

Burst transfer rates
PCI card No Patch [MBytes/s] Patch 1.04 [MByte/s]
ACARD AEC-6280 89,6 106,5
Adaptec 29160 78,2 104,5
Advance 29133 82,9 105,8
HighPoint RocketRAID 133 81,2 99,9
Promise Ultra133 TX2 90,1 111,3

well,

& what's about the other pci cards ?

(gc,tv,capture,...)

if you know you don't know, the way could be more easy.
February 17, 2002 4:16:02 PM

Quote:
that gain is for EVERY pci card, not just ide

lol.


if you know you don't know, the way could be more easy.
February 18, 2002 5:37:41 AM

Quote:
Your reference as to VIA never being deployed in a server enviorment is laughable, but hey, I agree. So in effect what you are saying is that VIA is in effect sub-standard. I wonder if they place such a disclaimor on all their motherboard chipsets " Warning, this product is not up to the standards required in any server enviorment".


I am not saying via is good, I am saying that the kt266a chipset does not have any unpatched bugs that I am aware of.

I was unaware that the patch I linked to was different from the via official patch however(thought they were the same). Still, a patch is a patch, and I have no qualms in reccomending a kt266a motherboard to anyone.



"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 18, 2002 5:39:36 AM

Quote:
well,

& what's about the other pci cards ?

(gc,tv,capture,...)



I dont follow, the link you provided showed the patch increased performance, on everything.

Why should you assume that a mpegcapture card would be any different than a raid controler card?

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 18, 2002 5:41:02 AM

Quote:
You keep bringing up the Intel I850 in defense of VIA. Sorry, thats a no go. I can just imagine you as a trial lawyer, your defence would go like this:

"Your honor, it is true my client did in fact kill the defendent, but look, Mrx. also has committed murder so then it must be ok" Just doesn't work does it?



NO, you say via is below industry standards, Intel sets the industry standards, they are regarded as the best chipset maker of them all, DO YOU DENY IT?


My defence is that if via is below industry standards, which intel sets, and intel has the SAME ISSUE AS VIA, how can via be below industry standards.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 18, 2002 5:41:50 AM

Quote:
Your reference as to VIA never being deployed in a server enviorment is laughable, but hey, I agree. So in effect what you are saying is that VIA is in effect sub-standard. I wonder if they place such a disclaimor on all their motherboard chipsets " Warning, this product is not up to the standards required in any server enviorment".



BTW, I would appreciate it if you stopped assuming I am defending via, when I am merely defending my comment about a specific chipset.



"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 18, 2002 7:41:50 AM

Quote:
NO, you say via is below industry standards, Intel sets the industry standards, they are regarded as the best chipset maker of them all, DO YOU DENY IT?


My defence is that if via is below industry standards, which intel sets, and intel has the SAME ISSUE AS VIA, how can via be below industry standards.

First of all yes, maybe. I815 and BX, yes, great chipsets. Lately though I don't think Intel chipsets are all that special. The I850 is a bit of a puzzle to me as I have not looked into the issues it has with the pci bus other than to read the same articles as you. It really hasn't been an issue with me as I do not see myself using it anywhere in the near future. But let us clarify one thing here, intels issue deals with one and only one chipset the I850, while the via bug effects all via chipsets from the mvp3 up to current chipsets so that would be in the following order, mvp3,kx133,kt133,kt133a,kt266,kt266a. Theres a bit of difference hear don't you think? And why don't SIS and ALI chipsets suffer from this bug? It is not just a IDE raid issue it is a busmastering issue.

It's not what they tell you, its what they don't tell you!
February 18, 2002 7:56:22 AM

Quote:
Still, a patch is a patch, and I have no qualms in reccomending a kt266a motherboard to anyone.

I don't agree. Using unofficial and unsupported "patches" Is no way to gain consumer trust for VIA. The possibility of damaging your board with no recourse is not a way I would find acceptable to make a statement such as the one you make.

For the record I am not saying that it is complete crap ( the kt266 ) only that it does have issues buyers need to consider beofre blindly making a purchase. For that reason I take issue with your recomendation of it having no "bugs" as you put it. Until enough of an uproar is heard from us to BOTH Intel and VIA about such matters, why would they be motivated to make better products? Bottom line is money talks, and as long as they can keep getting us to spend our hard earned money on buggy chipsets that need third party patches they will see no need to shape up thier act.

And as a side note after looking at two benchmarks comparing the new kt333 boards ( asus & msi) it would appear that now they have a different problem as when running at a memory bus of 333 Hard drive performance is worse then when running at 266. It is a bit early to tell if this is just a pre-release issue they need to iron out before its official release but none-the-less interesting when looked at in the context of this conversation.

It's not what they tell you, its what they don't tell you!
February 18, 2002 9:39:59 AM

Quote:
For the record I am not saying that it is complete crap ( the kt266 ) only that it does have issues buyers need to consider beofre blindly making a purchase. For that reason I take issue with your recomendation of it having no "bugs" as you put it. Until enough of an uproar is heard from us to BOTH Intel and VIA about such matters, why would they be motivated to make better products? Bottom line is money talks, and as long as they can keep getting us to spend our hard earned money on buggy chipsets that need third party patches they will see no need to shape up thier act.



Even unpatched the bandwith issue does not post a threat, and does not cause unexpected operation, it only limits data transfer at extremely high rates, only affecting raid performance and very pci intensive things.


Its a bug, there is a via official patch which solves the problem for everry thing they tested, but I cannot find any links to see if it solved any other pci cards(due to the newness of the patch).

There is also an unofficial patch which DOES resolve issues with other than ide cards(not saying vias official patch dosent, just saying that I have no links to show if it does due to the time length the patch has been out).


All in all, the issue with the kt266a chipset is minor, and has a working patch from via. I still have no qualms about reccomending it to anyone looking for a motherboard.


Furthermore, The kt266a performs very well even WITH the bug, and except in the event of a ultra high bandwith device, the bug is NON DETECTABLE, as evidenced by the fact that it was only recently discovered but affects all via chipsets with that particular southbridge or its variants.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 18, 2002 9:41:22 AM

Quote:
First of all yes, maybe. I815 and BX, yes, great chipsets. Lately though I don't think Intel chipsets are all that special. The I850 is a bit of a puzzle to me as I have not looked into the issues it has with the pci bus other than to read the same articles as you.


If intel is not the industry standard in chipsets, then what is, how can via be below something we cannot define, ALL chipsets have errata, and many have unpatched erratta, I fail to see why the kt266a chipset is any worse than any other chipset.

"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 18, 2002 9:47:57 AM

Quote:
But let us clarify one thing here, intels issue deals with one and only one chipset the I850, while the via bug effects all via chipsets from the mvp3 up to current chipsets so that would be in the following order, mvp3,kx133,kt133,kt133a,kt266,kt266a. Theres a bit of difference hear don't you think? And why don't SIS and ALI chipsets suffer from this bug? It is not just a IDE raid issue it is a busmastering issue.


All those chipsets used the same or similar southbridges, where the issue lies, FURTHERMORE, the length of time the bug existed unrepaired is moot, how can they fix an issue they didnt even KNOW ABOUT, untill very recently.


In fact, within a few weeks after the busmastering issue was discovered via had a patch for it. This issue confirms for me that via is on the ball, and are genuinely trying to fix whatever issues arrise in their chipsets.

The unofficial patch for sure fixxes the sblive issue(according to those links anyways) The via official patch for sure fixxes ide transfer rates for raid cards(and IMO will repair many things when further testing is conducted). This IMO shows that the bug, if not totally fixxed, is patched to working standards, if you say that due to this bug, one should NOT buy a kt266a chipset board, then you MUST say, that one must NOT buy an i850 motherboard as well.

If you say that, then I will agree to disagree and believe you have high standards, if you do not, and excuse the same bug in another chipset, then you will be judging a chipset based on the fact its a via, and without ground IMO.




"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 18, 2002 3:52:20 PM

Quote:
All those chipsets used the same or similar southbridgesp

Not even close other than the fact they were all bandwith limited and all made by VIA.

Quote:
If you say that, then I will agree to disagree and believe you have high standards, if you do not, and excuse the same bug in another chipset, then you will be judging a chipset based on the fact its a via, and without ground IMO.

Like I said, I do not know enough about the Intel Issue as that motherboard was never a serious concern for me. I am not defending Intel but nor am I going to condemn them for something I know little about. If there chipset suffers the same problems as VIA does I could easily say that. You keep bringing up the IDE RAID patch and resultant improved transfer rate on VIA boards. Remember that is a result of the problem ( decreased transfer rates) and not the problem itself which is the ability on a bus mastering pci card to accesss the via pci bus for a continous uninterupted duration. The VIA bus interupts this data flow several times which as a result decreases sustained thorouput. It is the same issue that effected the 686b and the sound blaster live. It is the same issue why many users had problems using an aureal sound card in the past on a via chipset board. This being the case this issue did not just suddenly surface as you suggest. I am not sure how VIA could just patch such an issue and I suspect the atch is just a workaround and not a fix as you suggest.

There are other reasons why internal transfer rates could be limited on the intel board which might not have as broad sweeping impact on all pci cards as does VIA's.

It's not what they tell you, its what they don't tell you!
February 19, 2002 6:23:27 AM

Quote:
Not even close other than the fact they were all bandwith limited and all made by VIA.



The kt133 uses the 686 a
the kt133a uses the 686b
the kt266 uses the 686b
the amd761 uses the 686b SB
the kt266a uses the VIA VT8233 Southbridge, which explains why it is NOT affected by the sblive bug.

However the transfer was NOT discovered untill AFTER the release of the new chipset, and I am confident that they will repair the issue in the next silicon.


YOU claimed I was wrong about the southbridge's being the same or similar, YOU were wrong, now how do you propose to debate when you are using such incorrect facts.

The bug is in the southbridge, the southbridge had no reason to change, the bugs were not discovered. When they discovered the sb bug, they repaired it. Now they have found a new bug, the official patch repairs every card brought to their attention by the tech site which broke the news, and an unofficial patch repairs other issues(altho I bet those other issues were for older chipsets and not the kt266a). The current state of the kt266a chipset is stable, and the one bug present has been patched. I reccomend it wholeheartedly.


Quote:
I am not sure how VIA could just patch such an issue and I suspect the atch is just a workaround and not a fix as you suggest.



Tell me, what is the difference between a workaround and a fix, obviously this issue is in the silicon, and a fix (per se) will not happen untill a new sb revision is released. If the "bug" with the kt266a is not visible unless you use an ultra high end raid card, and all the high end raid cards they tested were fixxed by vias official "workaround" then what is the problem?


"The Cash Left In My Pocket,The BEST Benchmark"
No Overclock+stock hsf=GOOD!
February 19, 2002 12:07:57 PM

Quote:
YOU claimed I was wrong about the southbridge's being the same or similar, YOU were wrong, now how do you propose to debate when you are using such incorrect facts.

HUH? any reason why you left out the mvp3??????? Just like to help make your point by omitting a little information don't you? three differnt chipsets spanning several years.

mvp3
686 a/b
VT8233

A debateable point can be made about the fact that the 686a and 686b being significantly different enough to designate them as seperate chipsets in themselves ( the 686a after all did not have the sblive issue ).

It's not what they tell you, its what they don't tell you!
February 19, 2002 12:27:23 PM

You keep coming back to with that VIA only recently discovered this bug. That is the one troubling issue I have with the whole thing. I guess VIA has no concept then of quality control if this could go undetected for several years. You are also making an assumtion that only with high end Raid cards will a user suffer a degradation in performance, unfortunatly this is not the case. Even before reaching the 80mb/s threshold ( which should be 133mb/s minus some for overhead) These boards suffer from slower than normal transfer rates. This can be shown by users who are not currently using a raid setup still gaining transfer speed after installing the patch. This makes you statement about how it was just discovered a little questionable doesn't it?

It's not what they tell you, its what they don't tell you!
!