VOGONS


Reply 20 of 54, by maddmaxstar

User metadata
Rank Member
Rank
Member
swaaye wrote:

The early Athlon motherboard hardware had a pre production feel to it. Little refinement, bugs, high sensitivity to power supply output, AGP instability, etc. There are a few boards with huge areas of unused, unnecessary PCB (very sloppy work). AMD 750 was AMD's contribution there because that chipset has broken AGP 2x. The CPUs themselves were perfectly fine AFAIK.

I'll have to agree with that. My Slot A Athlon 550 featured in my Avatar is one of the early Argon cores that's labeled "7th Generation Processor", and apparently that's not uncommon. Not to mention that because it was such an early chip, the CPUID string says "AMD K7" rather than "AMD Athlon", which threw Windows XP for a loop when it was first released (ie: no processor enhancements enabled, it ran slow without a patch).

Not to mention it's matching FIC SD11 motherboard was a humungous but almost bare ATX board that had few enhancements or modifications from the early pre-production versions of the board.

Other than that, the machine ran pretty stable. I was using a Voodoo 3 at the time, so I didn't have many AGP 2x issues with the SD11's Irongate Northbridge. The machine was an incredible performance beast as well compared to what my friends had, that 2 of my friends and my Brother got Slot A systems not long after I got mine. I still have my Athlon T-shirt somewhere that I got for registering with AMD.com back then, I used to only wear it to LAN Parties. 😜

= Phenom II X6 1090T(HD4850) =
= K7-550(V3-3000) =
= K6-2+ 500(V3-2000) =
= Pentium 75 Gold(Voodoo1) =
= Am486DX4-120(3DXpression+) =
= TI486DLC-40(T8900D) =
= i386sx-16+i387(T8900D) =

Reply 21 of 54, by jaqie

User metadata
Rank Member
Rank
Member

I actually had a grey duron T shirt, at the time my duron 600@952 was a very good gaming beast, I bought the v4 4500 as soon as it came out for it. My comp was one of the best at the LAN parties we used to have. Ah, those were the days. 😀

Reply 22 of 54, by swaaye

User metadata
Rank l33t++
Rank
l33t++
maddmaxstar wrote:

Other than that, the machine ran pretty stable. I was using a Voodoo 3 at the time, so I didn't have many AGP 2x issues with the SD11's Irongate Northbridge.

The Voodoo3 saved you. 😀 I have an ASUS K7M that I've experimented with a bit and cards other than 3dfx tend to not work reliably. The AMD AGP driver disables AGP 2x (because it literally is non functional in the nortbridge). But even with cards running AGP 1x, with Radeon and GeForce cards I see occasional image corruption and lock ups. It's like a Super 7 board.

jaqie wrote:

I actually had a grey duron T shirt, at the time my duron 600@952 was a very good gaming beast, I bought the v4 4500 as soon as it came out for it. My comp was one of the best at the LAN parties we used to have. Ah, those were the days. 😀

Duron was glorious. I knew people who upgraded from K6-2 to the cheap Durons and it blew their minds. They would often see 3-4x frame rates in 3D games.

Reply 23 of 54, by elianda

User metadata
Rank l33t
Rank
l33t
swaaye wrote:
maddmaxstar wrote:

Other than that, the machine ran pretty stable. I was using a Voodoo 3 at the time, so I didn't have many AGP 2x issues with the SD11's Irongate Northbridge.

The Voodoo3 saved you. 😀 I have an ASUS K7M that I've experimented with a bit and cards other than 3dfx tend to not work reliably. The AMD AGP driver disables AGP 2x (because it literally is non functional in the nortbridge). But even with cards running AGP 1x, with Radeon and GeForce cards I see occasional image corruption and lock ups. It's like a Super 7 board.

Well, you must own a very picky board. I usually found that AGP x1 fallback quite annoying and forced by default back to AGP 2x using Powerstrip. I used a GF2 GTS from Elsa at this time and used the machine to run Everquest most* of the time. (*literally)

Reply 24 of 54, by swaaye

User metadata
Rank l33t++
Rank
l33t++

Yup the ASUS K7M could be part of the problem. I looked up Usenet posts for details on the AGP problems with this board and found a lot of detailed posts with various cards. I just use my Voodoo5 and all is well.

But I have read that the 750 chipset has defective AGP functionality
http://web.archive.org/web/20000818052927/htt … sets/index.html

"Even if one chooses to go with an Athlon solution over the P3/Coppermine, the only chipset readily available now, the AMD Irong […]
Show full quote

"Even if one chooses to go with an Athlon solution over the P3/Coppermine,
the only chipset readily available now, the AMD Irongate chipset cannot
handle AGP properly. This causes cards to resort to AGP 1x "command mode"
which somewhat slows down the card. Realize that AGP 1x "command mode" is
NOT AGP 1x, textures and other data is still retrieved at AGP 2x speed.
However, even AGP 1x "command mode" still has a problem, and that problem
becomes evident when we look at what's wrong with the Irongate chipset.

Essentially, the Irongate chipset has a bug in it which corrupts data/etc.
sent through the AGP bus. In the case of things such as texture data, the
number of corrupt bytes is very small; in actual gameplay, one or two pixels
which are not what they are supposed to be will never be noticed. However,
if a graphic command is passed through wrong, the effects can cause even a
system crash. For this reason, BIOS fixes for Irongate chipsets play it safe
and resort to AGP 1x when sending commands. It is interesting to note that
once applications begin to make extensive use of T&L, the sparse data errors
can cause hugely noticeable effects, for example, entire polygons may be
severely skewed or completely missing if simply one bit is flipped.
Hopefully VIA's recently released Athlon chipset will bring about good
things for Athlon."

The part about T&L corruption reminds me of the distorted models I saw in 3DMark2001 Lobby Test with a Radeon 7500 on that board.

Reply 25 of 54, by elianda

User metadata
Rank l33t
Rank
l33t

AFAIR the 750 has some higher noise levels on the AGP port that could lead to the wrong data transmitted at AGP 2x. Still noise is a somewhat analog problem and is highly dependent on the mainboard layout, graphics card used, probably even the used power supply. Still the hardware manufacturers play safe and fall back to AGP 1x if an irongate is detected. This does btw. also the NVidia driver.
Still I strongly recommend everyone with such a setup to test if his setup is prone to the data corruption and thus lockups. Usually running 3DMark2001 in a loop for several hours will reveal instabilities.

Retronn.de - Vintage Hardware Gallery, Drivers, Guides, Videos. Now with file search
Youtube Channel
FTP Server - Driver Archive and more
DVI2PCIe alignment and 2D image quality measurement tool

Reply 27 of 54, by maddmaxstar

User metadata
Rank Member
Rank
Member

SD11 is definitely a sparse board:

6699055635_8540b4f33c_z.jpg

As for the Asus K7M, I recall hearing numerous bad things about the board back then. At the time I was doing research for building my Slot A system in early 2000, numerous sites were recommending staying away from the K7M or waiting for VIA. My old VA-503+ worked well enough, so I went FIC instead.

I actually didn't know about the Irongate AGP2x issues until later, I don't think I've ever used anything other than the Voodoo3 in the SD11. By the time I got my GF2Ti I'd moved on to my Duron 1GHz. (btw: ECS K7S5A = Worst board ever!)

= Phenom II X6 1090T(HD4850) =
= K7-550(V3-3000) =
= K6-2+ 500(V3-2000) =
= Pentium 75 Gold(Voodoo1) =
= Am486DX4-120(3DXpression+) =
= TI486DLC-40(T8900D) =
= i386sx-16+i387(T8900D) =

Reply 28 of 54, by swaaye

User metadata
Rank l33t++
Rank
l33t++

Inexplicably, the ECS K7S5A and SiS 735 have a fanboy crowd. Watch out. 😉

Here's the other FIC board with wasted space. The AZ11. It looks like you could make your own micro ATX board with a bit of sawing.

Attachments

  • az11.jpg
    Filename
    az11.jpg
    File size
    41.89 KiB
    Views
    2636 views
    File license
    Fair use/fair dealing exception

Reply 29 of 54, by F2bnp

User metadata
Rank l33t
Rank
l33t

Wtf! I've seen unusable space in boards before, but this is just ridiculous...
Look at that SD11...

Athlon boards have been nothing but pain in my case. I've had 3-4 of them and not even one worked properly, with a lot of different CPUs. It's a shame really, the CPUs apparently kick ASS, but what are you going to do with them when you've got a VIA chipset :@ ?
nForce2 seems to really rock it out though, very feature rich and apparently stable and all, but the only board I ever got was DOA 🙁. Luck is not on my side it seems 😜

Reply 30 of 54, by elianda

User metadata
Rank l33t
Rank
l33t

Well, I used this: http://mail.lipsia.de/~enigma/neu/pics/ga7ixe.jpg
which is a changed design of this predecessor http://mail.lipsia.de/~enigma/neu/pics/ga7ix.jpg

Maybe this also helped AGP 2x to work more reliable.

I currently have a setup with the first board as transfer system for old home computers that I use rarely:
CPU: Athlon 650 MHz (Slot-A)
Mainboard: Gigabyte GA-7IXE
RAM: 512 MB SDRAM (due to Win98SE)
Graphics: Voodoo 3 3500TV with Video In/Out
Video: Hauppauge WinTV PCI
Sound: Terratec EWS64XL
HDD: 2 GB Sandisk Compactflash on CF to IDE adapter

Reply 31 of 54, by sliderider

User metadata
Rank l33t++
Rank
l33t++
elianda wrote:

AFAIR the 750 has some higher noise levels on the AGP port that could lead to the wrong data transmitted at AGP 2x. Still noise is a somewhat analog problem and is highly dependent on the mainboard layout, graphics card used, probably even the used power supply. Still the hardware manufacturers play safe and fall back to AGP 1x if an irongate is detected. This does btw. also the NVidia driver.
Still I strongly recommend everyone with such a setup to test if his setup is prone to the data corruption and thus lockups. Usually running 3DMark2001 in a loop for several hours will reveal instabilities.

This may explain why the SD11 is such a huge board. Longer traces might absorb some of the noise, abating the problem somewhat. I bought an SD11 board recently because it was cheap, but unfortunately it arrived with a cap broken off so I haven't been able to use it. It's on the "to do" stack to replace that cap so until it gets fixed I have to make do with proprietary OEM boards for my K7's.

And you never had a problem using a GeForce with the K7? K7 boards with the AMD chipset were notorious for having issues with GeForce cards more than any other.

Reply 32 of 54, by swaaye

User metadata
Rank l33t++
Rank
l33t++

I've tried TNT2, GF256, GF2MX and GF3 on my ASUS K7M and had problems with all of them in some way. Same with ATI cards. It seems that only 3DFX's cards which don't use AGP features work OK.

Last edited by swaaye on 2012-02-23, 19:12. Edited 1 time in total.

Reply 33 of 54, by elianda

User metadata
Rank l33t
Rank
l33t

I didn't say never. In retrospective I would say on average it crashed once a week. But this was ok at this time. Other boards had also their glitches 😉.

Retronn.de - Vintage Hardware Gallery, Drivers, Guides, Videos. Now with file search
Youtube Channel
FTP Server - Driver Archive and more
DVI2PCIe alignment and 2D image quality measurement tool

Reply 34 of 54, by maddmaxstar

User metadata
Rank Member
Rank
Member
swaaye wrote:

Inexplicably, the ECS K7S5A and SiS 735 have a fanboy crowd. Watch out. 😉

Here's the other FIC board with wasted space. The AZ11. It looks like you could make your own micro ATX board with a bit of sawing.

Yeah, it's hard to imagine why there wouldn't be K7S5A fans out there, the SiS 735 was a good chipset and it's support for DDR and PC133 made for a killer upgrade - which is how I got sucked in. However the K7S5A I had in 2002 with my Duron 1.0 was a poor but not poor enough board that lingered around like a bad cold for years.

The board I had had CMOS battery issues (brand new CMOS battery? won't hold memory if the system's unplugged), and stability at times was poor. I had it for 4 months before I sold it to a friend of mine and bought an Asus A7S333(GREAT board). A year later I do a sale/trade with him and ended up with the K7S5A back. Worked for a while as a backup until I replaced it in 2005. I took it out a couple months later to get it working, the board kills 2x 1GHz Durons(including my original Duron, the fan was secure and using paste, bad VR maybe?), and before I went to throw it out, I test an AXP 1600+ in it, and the damn thing works fine. I in no way wanted to trust the chip killer, so I sold it to another friend of mine with the warning that it may not work forever, and a year or so later he gets a new machine and wants to give the K7S5A back to me in 2007. I told him to turf it.

That plus the number of K7S5As I've seen with bad caps or other problems, and bad ECS/PCChips boards I've seen over the years (even on newer machines today, like Acers) gave me an extreme distaste for the brand.

I've got nothing against SiS chipsets though, the SiS 745 in my A7S333 was so stable, I once ran it for 137 days straight without a reboot in 2003. I miss that machine.

As for the AZ11 with lots of wasted space, that's crazy, looks like they could have cut some sections off to save on materials. There were alot of boards like that back then, especially with early ATX boards. You'd think they'd could have redistributed the components a little and placed everything on the board within the middle-to-back mounting holes and make the whole board smaller. I wonder if someone could actually take a dremel and saw the last slot and the empty space on the AZ11 off and make a MicroATX board with it. 😜

= Phenom II X6 1090T(HD4850) =
= K7-550(V3-3000) =
= K6-2+ 500(V3-2000) =
= Pentium 75 Gold(Voodoo1) =
= Am486DX4-120(3DXpression+) =
= TI486DLC-40(T8900D) =
= i386sx-16+i387(T8900D) =

Reply 35 of 54, by DonutKing

User metadata
Rank Oldbie
Rank
Oldbie

Yep, I'm another one that had a K7S5A with nothing but problems.
Mine would randomly fail to cold boot, and after less than 6 months ended up stone dead. Visually all the caps were fine but I suspect the failure was a combo of bad caps, cold solder joints and cheap/poor board design.

Rumour was there was a few different factories making these which accounts for the seemingly huge variation in quality (and the huge split between lovers and haters of the board 😀- I'm firmly in the latter camp)

I originally bought the board because I had 128MB SDRAM from my old system and couldn't afford new RAM at the time. It was paired with an AXP 1800+ until I managed to buy 512MB DDR. I upgraded from the K7S5A to an A7V333 which ran pretty well until the PS/2 ports died on it somehow. It was still a much better board than the K7S5A was...

If you are squeamish, don't prod the beach rubble.

Reply 36 of 54, by therelaxer

User metadata
Rank Newbie
Rank
Newbie

I have been using a Athlon Classic 800 on a Asus k7V with 512MB of pc133 cl2 with a 32MB Geforce 2 GTS and have been having a blast. My objective was to build a machine from about June/July of 2000. It even Handles RTCW at high settings at 1024x768 very nicely. I have played around with all sorts of machines from this era, and like the slot Athlon the best, as I was too poor to afford one back when they were popular. I went from a K6-2 500 to a Thunderbird 1 gig. I also never had a geforce 2 that wasnt an mx... Skipped right over to a GF3 when I went to an Athlon XP. I actually RMA'd the Crucial stick of 512 to get this rig going. It runs flawless now...

Reply 37 of 54, by Anonymous Coward

User metadata
Rank l33t
Rank
l33t

I really don't miss the Asslon/Pee4 days. I worked with a lot of K6 and early K7 systems, and they were nothing but poorly designed nightmares (mostly due to poor motherboards and chipsets, but many of the CPUs also had heat problems.) To this day I refuse to build AMD systems from years of being burned.

The problem back then though was that if you wanted the reliability of Intel, you had to settle with RAMbust and Pee4, so basically you were stuck between a rock and a hard place. In my dormitories, I was basically the only one still running P3 and SDRAM. I skipped right over the entire P4 generation and went right to Core. Sadly, the Coreduo and Core2duo chips and their related platforms still weren't nearly as reliable as what we had in the socket7 days. I now run a Core i3 based Celeron, and this is probably the best system I've had quality-wise since the mid 90s.

It's really too bad that AMD bought ATi and soured their relationship with nvidia. It seems that nvidia makes some fairly decent chipsets. I wouldn't mind trying an AMD again as long as I didn't have to run them on one of their own chipsets.

I don't know why you would want to relive the early 2000s. You're really not missing anything.

"Will the highways on the internets become more few?" -Gee Dubya
V'Ger XT|Upgraded AT|Ultimate 386|Super VL/EISA 486|SMP VL/EISA Pentium

Reply 38 of 54, by luckybob

User metadata
Rank l33t
Rank
l33t

I'm sitting outside in my workstation now that its FINALLY under 75F outside. (its 12:30 am) I have a complete setup ready for windows for slot A. I also have about 5 processors I can test. from what I see off hand... I have a 5, 6, 7, 950 and 1ghz slot A. I also have 5,6,667,850,1g pentium 3's...

speedsys scores would probably be "easiest" as I wont have to dick around with windows, however I dont think its a very good comparison. any suggestions for one or 2 benchmarks? 2 cpu marks and a video? to see if the cpu makes a difference in game fps?

It is a mistake to think you can solve any major problems just with potatoes.

Reply 39 of 54, by [GPUT]Carsten

User metadata
Rank Newbie
Rank
Newbie

Just wanting to add a thing to the Coppermine topic.

It was the first processor line from Intel that featured what was marketed as "Advanced Transfer Cache", ATC for short. Technically, it had a 256 bit interface to the processor core and that was what saved Intels performance for everything that went throught L2 cache.

Another thing, I remember fondly was my Athlon 600 back then - which also had the 7th Generation AMD Processor marking on it's case. A shame that I removed it for the chance of using a goldfinger device to OC it. Results where mediocre though, only got it stable to abou 680 MHz.

edit: I still have the CPU though, might check one day, if it's also got the K7 id-string.😀