QDI P5I430TX - is it the best Socket 7 mainboard?

In this video I take a look at QDI P5I430TX mainboard. The best Socket 7 mainboard in my opinion. Let me explain why.
Music by Model Povedeniya
modelp.bandcamp.com/
Patreon:
/ necroware

[@tellyjoossens4186]

On my first job I worked at the technical department of a wholesale computer parts supplier. Pc Chips and later Ecs were the crappy unstable brands. Qdi however had less than 5% defective rma returns. Asus was good, but QDI was the legend price/quality wise. The pentium 2 with bx chipset was their most famous and stable board ever. Even years later, I never heard of one with bad caps.

[@necro_ware]

This absolutely meets my experience. Even between Asus and QDI I'd always go with QDI.

[@CptJistuce]

PC Chips also made the LEGENDARY 486 boards with patented fake-cache technology!

[@user-zi1ip2tj8f]

The main issue of most 430TX mainboards was the Tag Ram, it supports only 64 MB Ram in cacheable range. If you put more Ram inside than 64 MB, the system gets much slower... ^^

[@charlesdorval394]

Oh I didn't know that! Thanks for the info!

[@looks-suspicious]

Since 64 MB is the perfect amount of RAM for a Socket 7 DOS/Win95 machine, this hardly matters.

[@nigelrhodes4330]

@@looks-suspicious People tend to try to max out their boards thinking it will run faster so it is good to point this out to people.

[@oscrthgrch7]

That is a concern with most CPUs, though it's a non issue if you use a K6-3/2+/3+ CPU. With one of those you can use the maximum of 256 megabytes with no problems.

[@nalinux]

Surprising. I upgraded to 96 Mo because Unreal swapped with 64 Mo, and don't remember the system getting slower. Of course I loved trying benchmarks. With a 233 MMX at 290 MHz and a Voodoo1 .

[@Stratotank3r]

I am a blessed owner of a QDI Titanium iB+. The + variant can handle K6-2 up to 400MHz (6*66) within all specs. Bit all QDI Boards were very very well made.

[@justjuliao5187]

Are you aware that Jan Steunebrink made a new bios for it in 2024?

[@necro_ware]

The IB board can also handle that. It's just a software limitation, which hopefully can be removed.

[@warrax111]

​@@necro_ware there is anandtech review for the board, one of the first reviews in early 1998. he mentioned there, that there is 83 mhz FSB in unofficial BIOS.

[@JamieBainbridge]

Yesterday Bits und Bolts, today Necroware. "Germans fixing motherboards" is one of my favourite KZbin genres 😎

[@SnipE_mS]

not sure if he's German but Epictronics is good for that too.

[@LeeMc007]

Half an hour to myself and Necroware drops a video, great timing, grabbed a brew and some dinner and off we go. 😁👍

[@flashgordon8554]

This was my first Socket 7 board that I had back in the day, when I was 18 years old, and also my first computer build. I remember like it was yesterday, when my father gifted this to me (at my recommendation, of course :) ), that this literally brought tears to my eyes. I was laying my hand on a revolutionary jumperless board and I couldn't believe it! I paired it with an Intel MMX 200Mhz processor and a Matrox Mystique 220 graphic card. I was coming from a 486 DX4 100MHz AMD processor and some VLB motherboard. It was a big upgrade. Thank you for the video and the nostalgia!

[@chu-icehugehard1820]

Great video! If you let me choose a Socket 7 mainboard, I'd prefer Super Socket 7 ones - AGP support, faster IDE controllers, wider CPU options as well as ATX form factor. QDI was a subsidiary of Lenovo (then called Legend), and they made mainboards for Legend-branded desktops/servers and the DIY market just like today's Asus. Sadly it seems that they no longer make mainboards after the Intel 945 era...

[@Adam-ni5lt]

Whether it's of any interest to anyone: I had an Intel 430TX motherboard back in the day with both DIMM and SIMM slots. The PC came with 1x 32MB EDO (yes, EDO) DIMM and, since it was cheaper, I ordered 2x 16MB EDO SIMMs as an upgrade. The 32MB DIMM and the 16MB SIMMs worked happily together and I had a massive 64MB RAM to play games with. 😊 (It probably would have burnt out the DIMM stick if it was an SDRAM module instead of EDO module). Really enjoy your videos on Socket 7. Brings back memories. 👍

[@jbaroli]

Yes,it would have burn out the 3.3 v SDRAM. EDO DIMMs work at 5 v.

[@dallesamllhals9161]

@@jbaroli Huh! I've got 128MB* SDRAM in my Iwill P55XB2 right now? *YES-yes I know only 512KB onboard cache, but that's what i had around after needing my 2x16MB SIMM in my Socket 3. EDIT: Oh, you mean SIMMs and DIMMs at the same time?

[@christopheoberrauch784]

Some time ago I did some benchmarks regarding the RAM. EDO 60 ns RAM is significantly faster than SD RAM.

[@nigelrhodes4330]

Glad to hear you say that, I found the same thing on my GA-586TX3.

[@warrax111]

you probably screwed something with timings on SDRAM.

[@registrazioniduemillaotton6030]

Oh I've had this board!! mine was even more special because the pcb color was not the usual green but a more teal or blue-ish color... Excellent board!

[@Constantin314]

so many motherboard manufactures back in the days compared to the 4 or 5 we have today. QDI, ABit and DFI were my favorites. i liked SOYO but i think the reason is that i liked how it sounds :)

[@bengelman2600]

Nice! Retro computer with jumperless settings will be super useful.

[@Gectms]

Had this exact same board. One of my favourites of all time.

[@SrLobo90]

I can confirm it, I have the same one and it is a great motherboard, it was one of the first boards to have automatic FSB and vcore adjustment based on the microcodes, dc-dc converter, on top of that it is very stable. I have had it working on that motherboard with 256MB of ram and Windows XP, and it was working really well.

[@johnbee1574]

i`m sure that tool specifically is a schrader valve removal tool. amasing video by the way :)

[@rs-qq3os]

I have this model of board, I bought it new in the box in 1996. It still works without problems and is my base for a Socket 7 retro machine.

[@warrax111]

You had probably its predecessor, QDI P5I430VX-250DM V2.0 (S2.1) (Explorer II/SpeedEasy) Looks very similiar, if not the same. Along with Abit, it was first board to feature jumperless design (tweaking multiplier, CPU voltage and FSB, in BIOS). TX board came out in middle 1997. You can find Anadtech review in early 1998. first BIOS is around the june 1997.

[@rs-qq3os]

@@warrax111 No IT was Titanium I on TX chipset, probably I messed up with the datt.

[@mathiasheyer3079]

This was my first own PC mainboard. Still own it. Still works.

[@TheWizardHW]

Amazing video Necroware. We need a tutorial to make multiple DOS configurations boot menu!!!

[@danielpetrov9179]

This is like deja vu for me, yesterday I fixed just the same board with missing or broken caps, bent pins and even broken coil.

[@PKmuffdiver]

I also had this board. It was fantastic. I still have the P1 233 MMX cpu, but sadly, time has made off with the mobo. Great Video.

[@RacerX-]

Nice video and that is indeed a versatile board. I have a Chaintech Socket 7 board that calls the jumperless SeePU technology. Haha they must have called them all sorts of things. It is one reason I like the Slot 1 era so much. Usually jumperless and auto detecting. It made things so much more easier at the time when building systems.

[@warrax111]

I have PCChips motherboard socket 7. Not only it is fully jumperless, but it also supports voltages from 2.2 to 3.5v in 0.1 v steps. Didn't expect it from PCchips, it supports even K6-2. They were known to not have overclocking. But this board has even FSB , 75 and 83 Mhz in BIOS. What is more strange, board is from late 1997. So PCChips were this good in late 1997, they adapted very quickly. Unfortunately, they didnt hold this trend on Slot 1 boards.

[@sonic2000gr]

I had this specific one. Amazing board.

[@zorrzdxqwetti6482]

Отличное видео, благодарю.

[@David_Ladd]

Thank you for sharing @necro_ware Great video as always :) I love seeing you save these boards and other devices from ending up in landfills or destroyed for recycling.

[@effectpower]

I have one of those boards whith K6 modified bios running an K6-3 @ 400Mhz

[@SianaGearz]

The 10 Ohm resistor is a cheap way to control back EMF spikes of the potentially electrically noisy fan motor of that era, to make sure not much of that leaks back to the source and mars the rest of the board's electrical performance. There would also be a capacitor or two in parallel with the fan.

[@nigelrhodes4330]

That was my thought too, it should only be a small value such as that.

[@teknologyguy5638]

Great video and find

[@Trailtraveller]

Sold these boards, QDI excellent brand.

[@carlosfernandes8919]

Muito bom , Classe A , Voce fazer um HARDWARE, antigo voltar à Vida...Parabéns... Exercicio pra mente pro Cerebro...

[@mitch075fr]

I had a similar mobo : Shuttle HOT! 569 iTX. While it didn't have CPU settings in BIOS and instead relied on a series of switches (not jumpers), it added support for 83 MHz FSB speed and a third SDRAM slot. Let me tell you, running a Pentium 133 at 166 MHz on it (83x2) was a BLAST - benchmarked at Pentium 200 level. I had started looking for an AMD K6 to plug on it when I chanced upon a nice deal on a Celeron 300A and an Abit BH6 - people from that time will realize that there wasn't much on Socket 7 that could rival that setup, and even Super S7 was hard pressed.

[@pavelfara9333]

Nice! I have saved the cheaper 430VX QDI which has the SpeedEasy feature as well! It is kind of a hybrid of old and new stuff. Btw - it is quite common to have a VRM failure on those boards that is manifestig in the way that only single voltage cpus work! In some cases one of the power transistors / mosfets is bad. In my case it was one voltage reference (small 3 pin part) that failed. Used parts from a dead broken board and saved the better one. 😂 And the funny part . I had to fix the same broken Simm slot and IDE connector as you 😀. Used 100% the same method and all work. In my case the board beeing VX I need the simms more.

[@fattymcboomboom9254]

6:50 wow you bombed that deoxit there. I only apply very little of it since I don't like it when the board becomes greasy. Or do you have some alternative without lubricant?

[@necro_ware]

Yea, unfortunately the spray which I have is quite strong. However, I knew, that I'll have to wash the board anyway. It was so sticky, that a bit of dioxide couldn't make it any worse.

[@nigelrhodes4330]

I will sometimes go over boards with a hydrocarbon spray to clean them if they look to have a lot of oil residue.

[@ukaszkowalczyk3011]

There's also a PC Chips m570 Socket 7 motherboard with such jumperless CPU configuration. It has also an AGP slot, integrated Sound Blaster Pro compatible sound chip and USB support. I have it in my collection and I use it for quick testing of components from this era.

[@warrax111]

There is even PCChips M572, which is fully Intel TX , no AGP socket 7 board, so direct competitor. I have it, and it shocked me. Fully jumperless, voltage and FSB in BIOS, and voltages from 2.2v to 3.5v... in 0.1v STEPS! Couldn't believe it. Yes, it owned even this QDI Titanium IB, i stood shocked when I first tested it.

[@alaricjeard269]

I have an Abit TX5 that have exact same feature, jumperless! What a great thing on a socket 7 motherboard!

[@Wormetti]

BIOS CPU settings, what will they think of next! 😮

[@warrax111]

unfortunatelly, this was last good thing QDI made. From this point, they only fell deeper and deeper. Their boards started to be pretty average, and their overclocking in BIOS sucked. Even on Slot 1, they lost it with no voltage setting in BIOS, and FSB only "66 to 75 or 83, and 100 to 112" overclocking. Nothing something like 66 to 100, or free 66-133. Or 100 FSB underclock to 66. Anyway, Asus, Abit, MSI and Soyo owned it instead. After socket A, they sucked hard, on socket A they were average, or below average. Their boards had weak BIOS, compatibility problems, random freezes. They were close to ECS / PCChips, just slightly better. They never made expetional board after this Speedeasy Explorers and Titanium. Like this board was praised in review, they were good on positions charts, but after this, they almost disappeared. They were pretty average after.

[@tigheklory]

I still have my DFI 586IPVG (Diamond Flower) Socket 7 board from back in the day, It has VX chipset, IR and USB built in and is an awesome board. It's not jumperless, but check one out if you get the opportunity.

[@p_mouse8676]

I don't know the power rating of that little resistor, but it could potentially draw P=(0.9/12)^2*68=0.38W, which is a bit much for such a small resistor.

[@FrancisFjordCupola]

100Mhz Front-Side Bus or go home, as far as I'm concerned. The TX, if I recall correctly could only cache the first 64Mb. Put in some decent memory and that ditches the performance. The HX or VX or something like that could cache up to 512Mb. The board also doesn't seem to be an ATX board. Also seems to lack AGP. Setting a few jumpers never bothered me. Although if you can do it with software, it's easier indeed. Still good to see the board receive some love. It is a nice oddity.

[@necro_ware]

Valid points, but 100MHz + AGP is a Super Socket 7 and not Socket 7. That is of course another story.

[@warrax111]

@@necro_ware supersocket 7 is , not only 100 Mhz FSB and AGP, but also 2.4v or lower CPU voltage.

[@emmettturner9452]

If 68Ω was the lowest value SMD resistors you had and you had more than one then you could’ve squeezed two of them side-by-side for half the resistance (34Ω). You could stack two more on top for 17Ω. That’s still 70% more than the original 10Ω resistor we’re replacing but drastically closer.

[@glitchwrks]

Very nice board! If you don't mind jumpers, the Shuttle HOT-569 is very similar.

[@krzbrew]

Nice Baby-AT board

[@DominatorHDX]

I have the exact same motherboard fully recapped and working ready to start my Pentium build one day when I get around to it. Together with a P233 MMX CPU, 64MB RAM, S3 Virge GX (STB Nitro) and Diamond Monster 3D 3DFX Voodoo 1 😇

[@emmettturner9452]

What tool are you using to remove the solder mask and do you think it would work well for removing the nickel strip spot welds on 18660 cells? I’m looking for a clean way to drill out just the spot welds so that the nickel strips can be reused when replacing bad cells in a battery pack. Most people using a rotary tool grind stone are just trying to smooth the ends out after ripping the nickel strips off.

[@Bainderosa_Technologies]

Thanks!

[@necro_ware]

Thank you too ;)

[@PROSTO4Tabal]

This QDI board is good for 3Dfx Voodoo1 upgrade as well as awe64 for 3D games such as Quake 1 or Turok. Looking forward to video #2 of this board

[@tegarz]

I see it has atx power connector. Have you tried directly use atx power connector instead of using atx to at adapter on it?

[@necro_ware]

Yes, works fine.

[@LG-HH]

Every child knows that the Asus P55T2P4 was the best 😉😬😁

[@SobieRobie]

😆

[@LeeMc007]

I agree, it's a great board, I run a K6/2 500(6x83) in it and 128mb of ram, once I added the Tag ram it was all cached, I'd say it's only weakness is no no Sdram slots.

[@nalinux]

Looks a lot like the MB I had. With undocumented 83 MHz FSB. 233MMX at 290 MHz very stable.

[@warrax111]

No, FIC PA-2007 is best. VP2 chipset. owned even Asus TX board back in the day, reviews were full of it.

[@jonatas464]

Awesome vídeos, as always. Do you haver some materiais about P3 Tuallatin? Something about Via C3, that's only runs at Tuallatin bords. And... Do you know something about CPU Galaxy Channel? He Just vanish.

[@danthompsett2894]

yeah i briefly owned a QDI board think it was Athlon Thunderbird or Athlon Xp based but i cant remmember but it was a quaility high performing board.

[@GizmoTheGreen]

maybe not so related question. I had a pc die on me when I removed the cooler to change thermal paste, after new paste it wouldn' boot. I bought a PCI post debug card and it doesn't show any numbers, no 0, only the default "----", does this mean dead motherboard? or cpu? the motherboard is a GA-7ZX rev 1.01 (actually seems to land between 1.01 and 5.0 because it exhibits features/lacking of both...)

[@PROSTO4Tabal]

Best 430TX in my opinion is Shuttle HOT 569, I use this board on many pc builds.

[@nigelrhodes4330]

I am currently running a GA-586TX3 mobo with a K6-2 450MHz running @400MHz as that is the highest it supports, quite a nice bump over a P200MMX :).

[@johnk7134]

If you want to use glue for plastics do not use the one in the video as it for styrene plastic. i would suggest you use Deluxe materials Plastic Magic which is for various kinds of plastic such as ABS.I have this board myself. Indeed i was amazed by the options, but there are also S7 boards with AGP slot as well.

[@hypercube33]

If you look at the traces that look corroded like that - they probably used soldermask and squeegeed it on and it wasn't the right viscosity to stay or get on those spots.

[@GizmoTheGreen]

How does the ram click in? the retention clips don't seem to reach the retention notches??

[@necro_ware]

On this model the white clips are just to push the modules out of the slot. Here they are not designed to reach retention notches. At 6:55 you can see it from the side once again.

[@GizmoTheGreen]

@@necro_ware what holds the ram in? I have a motherboard with such clips (both 72 pin ram installed by default) and I couldn't get the ram to seat, one side keep popping out.

[@Very_Dark_Engineer]

Partialy damaged smd resistor seems to be a fuse, such small sized resistor just simply cant disipate any energy so it is smd fuse with about 1 Ohm resistance

[@nalinux]

My 233MMX mainboard used jumpers, like most of them. Also with a TX chipset, I think it was an Asus. There was an undocumented jumper setting that allowed it to run at 83 MHz. So the 233MMX at 290 MHz very stable. Not sure it's possible with jumperless and settings in bios :)

[@warrax111]

this board can also run at 83 Mhz FSB. Just read Anandtech review. ;-)

[@dallesamllhals9161]

Iwill P55XB2 anyday I need those 5 PCIs 😛

[@SobieRobie]

Isn't this overheated trace powering the fan?

[@perhansson6718]

Looks like it goes to jumper JP6 - Clear CMOS, possibly if you short that and turn the board on it destroys the trace if the board is not designed with that in mind...

[@fft2020]

I have a QDI Explorer II motherboard with speedeasy that displays similar symptom like this one and always defaults to the pentium 75mhz after 3 boots it displays the message "you have failed to run 3 times this motherbaord and now it is in safe mode" and I can never get it to recognize the CPU it works but just at 75mhz... I have no idea what is wrong I have flashed a new bios and the problem persists

[@warrax111]

just use battery. also ESCD reseting can be out of power. Not sure how to solve it. Try to desolder BIOS, and use new.

[@fft2020]

@@warrax111 thanx even with battery it doesn't change :( maybe replacing capacitors around the CMOS will help

[@warrax111]

@@fft2020 try to update BIOS to the latest. Explorer II was buggy , it is early motherboard with BIOS tweaking, first QDI made. Also motherboard can have used BIOS chip. It can update ESCD only several hunded times. Motherboards use special RAM for that, but it has only 256 bytes of RAM. So later motherboard are writing part of configuration into BIOS NVRAM, because they store more data. I'm sure, Speedeasy probably doing it, because it has BIOS CPU tweaking options. Just my thoughts. :) Maybe someone played too much with it, and NVRAM of BIOS is used up and cannot remember new data written.

[@fft2020]

@@warrax111 Wow! thank you my friend! I never heard of that memory problem. I will investigate! So maybe the solution is to replace with a new bios chip no ?

[@warrax111]

@@fft2020 If NVRAM is used, then yes, you need to desolder old chip, and flash BIOS to new chip. If it is socketed, it will be easier, without soldering. I don't say, it is that, but it's another case, which can causing it. Maybe someone was too active, it was first motherboard with CPU tweaking in BIOS (along with Abit motherboards , they had Softmenu, QDI had Speedeasy) Good luck. If it will be not cause and BIOS is updated, then I am out of ideas, what can cause it. PS: If you want investigate, start with short description with ESCD in wikipedia. It is Extended System Configuration Data. It is explained there well and compact. Also check BIOS NVRAM (Nonvolatile BIOS memory) on wikipedia.

[@charonunderground8596]

I wonder how much of a speed difference there is between EDO and SDRAM on this board.

[@warrax111]

not much. But SDRAM is faster. I have it in my primary build. used SDRAM and EDO RAM too, even FPM.

[@adenunzio2000]

Just go with Super Socket 7. I would not waist my time on this, but still fun to play. Go with a board that can have AGP for one, Multiplier x6 or 6.5 , volts need to go down to 1.8., FSB at 100 or 112. Run a AMD k6-iii at 650 mhz

[@warrax111]

No. then you can just use Tualatin.

[@glitchwrks]

A lot of those boards that were early to add CPU fan headers don't provide much power at all for the fan. On my Shuttle HOT-569, it's current limited with a resistor on the motherboard, and is insufficient for most 12V fans you'd find on a Super 7 heatsink.

[@DO5MZ]

i think for me my ASUS P/I-P55T2P4S i use with a AMD K6III+ @400MHz all VID jumpers closed for 2Vcore 384MB Ram

[@marcsvenschulte]

Be careful when washing boards that have BGA chips (ball grid array) soldered to them. Salt residue from the water can remain between the balls and is difficult to remove.

[@necro_ware]

Yeah, salt between the balls is absolutely terrible indeed ;)

[@MrCof1ee]

I'm currently listing this motherboard on ebay :)

[@ruxandy]

So you are one of the crazy bastards who wants the cost of a kidney for it? 😁

[@wettuga2762]

I think I have a Titanium IB+ stored somewhere, according to the RetroWeb the only difference to that one is that it includes IrDA, so mine is better than yours 👅

[@YarisTex]

Since it’s a topic for another day I guess you already came across Jan Steunebrink’s BIOS for this board

[@necro_ware]

That BIOS is for IB+ and doesn't work on this IB board, but I'm already in contact with Jan about it.

[@drewduncan5774]

14:55 #R420replaceit

[@Evhen_Velikiy]

17:00 - Its not. With lower voltage fan will consume less power.

[@SidneyCritic]

Should stack 2 off 68Rs, ie, 34R - lol -.

[@oscrthgrch7]

I guess my comment was deleted because I linked to the Retro Web? Anyway, Jan Steunebrink has a patched BIOS for this motherboard that supports K6-2+/3+ CPUs and it is both on his site, and the Retro Web and I won't link to it because KZbin is dumb. This seems like a very nice board, but the TXP4 has a setting for 83 MHZ FSB.

[@necro_ware]

I'm in contact with Jan. He patched the BIOS for IB+ and this is IB, a slightly different board.

[@oscrthgrch7]

@@necro_ware Sorry, I didn't notice the +. I should have known that you would check with Jan.

[@dominikschutz6300]

Loads of SIMM sockets on the interwebs. Even mouser has them :)

[@ruthlessadmin]

I'm pretty sure the HX was the fastest...Wasn't the TX between the HX and VX? [SOLVED] Also, I'm sure this comes up a lot...where can I get a copy of that disk with all the benchmarks? I thought it came from Phil's Computer Lab, but I do not see it anywhere on his site. After searching for DOS Benchmark Pack, I found it...thanks to the person who named it but deleted their comment...

[@totallyuselessvideoz]

Not even close. TX supports 64-Mbit DRAM/SDRAM HX and VX don't

[@Choralone422]

The 430TX was released 1 year (Feb 97) after the 430HX and 430VX chipsets (Feb 96.) The TX, like the VX before it, were mainstream chipsets that supported SDRAM as well as 72 pin EDO/FPM memory but they both could only cache a maximum amount of 64 MB of RAM and only supported single CPU configurations. The maximum amount of RAM the VX supported was 128 MB and the TX supported 256 MB, again with only 64 MB of that being cacheable. The HX chipset saw use in mainstream mainboards but was also used in workstation and server boards as it supported a maximum of 512 MB of RAM with all of that cacheable with the proper cache tag RAM installed. It also had support for ECC RAM and dual CPUs. The Tyan S1562D Tomcat II is one board that is a good example of this. In general the HX chipset was thought of as being the fastest as there were a number of boards using it that featured 75 and 83 MHz FSB settings. I don't remember any VX or TX chipset boards that supported FSB speeds above 75 MHz. The Abit IT5H and Megatrends HX83 were 2 HX based examples of boards with 83 MHz FSB support. The Abit board also featured jumperless CPU setup which Abit called SoftMenu. Back then I owned a Megatrends HX83 board and ran it with a K6 200 at 83 x 2.5 or 208 MHz instead of 66 x 3 for 200 MHz. Often times the extra FSB speed gave a large boost in performance even though the CPU was only running a few Mhz faster than stock. That K6 wasn't entirely stable at 75 x 3 for 225 MHz and would not boot at anything higher than that so 208 MHz is where it stayed for the life of that machine.

[@ruthlessadmin]

@@totallyuselessvideoz I don't know...I found a thread on AnandTech forums from 2002 that seems to conclude HX is fastest. I wish YT allowed links but you'll find it if you search "i430 hx vs tx vs vx"...

[@oscrthgrch7]

Phil has it on his site under "DOS Benchmark Pack."

[@ruxandy]

@@Choralone422 "I don't remember any VX or TX chipset boards that supported FSB speeds above 75 MHz" - there are many TX boards out there that support 83 MHz (I have at least five).

[@ruben_balea]

66+ means 66.66 MHz instead of 66.0 MHz

[@necro_ware]

Really? Ok, will have to check it out. Thank you.

[@ruben_balea]

@@necro_ware Well, one user said 66.66 and other user said it was 68.5... I guess you'll have to use the oscilloscope to find out the real speed. The "Chipset Guide" article on AnandTech says that the Intel 430TX officially supported bus speeds are 50, 55, 60 & 66 MHz and the unofficially achieved bus speeds are 68, 75 & 83.3 MHz In their article "100MHz Bus Speed & Socket-7 Motherboards" they say "the 68MHz bus speed is referred to as the Turbo Frequency of the 66MHz bus speed, used Internally by manufacturers for testing purposes...however it makes a sweet little overclocking tool if you don't want to risk much" Wikipedia Socket 7 page says FSB speeds were 66.7, 68.5, 75, 83.3, 95, 97 & 100 MHz Some boards with jumpers already had a bus speed of 66+ but their manuals don't explain what that means. Maybe 66+ wasn't an exact frequency and could mean 66.7 on some systems and 68(.5) on others depending on what frequencies the chipset could generate between the official 66 and 75

[@warrax111]

incorrect. First, he had 60+ , not 66+. 66 means always 66,6. 66+ is QDI name for Turbo mode, and same 60+. Turbo mode means, you raise FSB a little bit, it was used for testing purpouses, for stability. that there is a little reserve. Because chips (including NB and SB) degrade over time, so they usually tested it on 60+ and 66+ ... it was for engineers. Lots of old boards have this settings. They are usually undocumented in manual, if set by jumpers. In BIOS, they can be eighter visible, or unlocked somehow. 66 with Turbo mode meaned usually around 68 Mhz... and in old reviews, they used this little trick to get overclock on boards, that didn't support it. Usually added like 3% performance. ( but it was for free)

[@eeeneeen381]

is it the best socket 7 mainboard? definetily not... its a very basic dummy board with an interesting but limited feature... nice try, but there are many much better board on the market... -only 256/512KB ext cache -very limited voltage range: the onboard Raytheon RC5041 VRM controller would be able to supply down to 2.1V, but i think it's impossible to implement it without hardvare modification... just check the VID configuration table in it's datasheet... -very limited multiplier settings: the question is, is the missing BF2 pin being connected to ground on the board, or is it left as floating? in first case, You will be limited to 4, 4.5, 5 and 5.5x multiplier options. in the latter case, limitation will be 2.5, 3, 3.5 and 6x with a K6-2 or K6-III CPU. if it's floating, then You can solder a jumper on BF2 line to achieve all possible multipliers on these CPUs -the W49C65-04 PLL clock generator is very basic without async PCI clock option - in 75MHz setup, PCI clock goes to 37.5MHz which can cause stability problems. the PLL would support 83.3MHz as well, but with 41.66MHz PCI clock - that's why they hid it from the options. there is no 75/33 and 83/33MHz async option. and in addition, it has no I2C but only hardware configuration pins... on many later S7/SS7 mainboards, You can reach the PLL IC through I2C bus and modify the setup on the fly from windows with CPUCool, CPUFSB, SpeedFan, etc according to PLL datasheet, 60+ means 60.6MHz and 66+ means 66.82MHz another problem is the missing microcode from BIOS, and the fact that the flash IC is being soldered on the board instead of socketed - it's a bit risky to hack the BIOS on a board like this... You probably met with SoftMenu II and SoftMenu III option on many P2/P3/S462 ABIT mainboard, but as i know, the first SoftMenu appeared "much earlier" on 430HX boards like ABIT IT5H

[@necro_ware]

I agree on some points, but you seem to compare mostly SS7 features with S7 ones. Async clock above 66 MHz was an SS7 feature, as well as voltage below 2.8V. More than 512K L2 cache is also not common on S7 boards. Nonetheless this board supports more, than you think. Some of those drawbacks are purely software limitations and there is a work ongoing to modify the BIOS accordingly. Of course fsb over 66MHz without PCI overclock is not possible though, but it's alright. Once again, we are speaking about S7, not SS7.

[@warrax111]

dude... by socket 7 , he means pure socket 7 ... not supersocket 7. supersocket 7 use socket 7, but they are considered by their own kind . supersocket 7. socket 7 boards used to have 512 Kb cache as maximum, in that era. So yes, this is expeptional good board... though 2.2-2.4v is definitely missing... even 2.7v, sometimes it is enough, and you can undervolt the processor.

[@alexsandroslin4113]

Очень трогательное видео