Difference between revisions of "Talk:Socket 7 Builds"

Latest revision as of 01:18, 6 April 2013

The presented VX Pro and TX Pro facts should be removed since they do not contain valuable information. Even the linked Redhill Guide article mentiones that the VX Pro is slower and that the TX Pro ran quite well even at 75 MHz. Quite a contradiction to whats written here.

TX Pro is btw the SIS5591 chipset and VX Pro+ the Apollo VPX chipset.

I have boards with both chipsets and I don't have anything to complain about these boards. Lower performance compared to i430VX might be the case for the VX Pro+, i did no direct comparison. The comparison should be done against another Apollo VPX board.Enigma (talk) 10:53, 4 April 2013 (EST)

Is that better? :) --RacoonRider (talk) 16:54, 4 April 2013 (EST)

Yes. Still the fact with the high defect rates has no source. Redhill writes that PCChips builds mainboards with lowest quality, but defect rates for VX Pro+ and TX Pro are not given. So it is open, if these boards had high defect rates. At least the boards I have still work, but that doesn't tell statistics. Also defect rate tend to follow the bath tub curve. So if anyone gets such a mainboard and it is still working it is more unlikely to fail soon.

So I would change the argumentation this way, that the boards were manufactured by PC Chips, a company that is known to use lowest quality components. I would not mention defect rates for these boards, as I do not know them and there is no source for defect rates. Enigma (talk) 02:12, 5 April 2013 (EST)

As a matter of fact, there is. http://redhill.net.au/b/b-bad.html

Second, they are made by PC Chips, renowned as the industry's cheapest, lowest-quality manufacturer. Their defective rates are legendary, between 10-15% in an industry where 5% or less is considered good. PC Chips was also the originator of putting "fake" plastic cache chips on motherboards. With these VX Pro boards you really do get what you pay for. ... The price is unbeatable, but price has always been PC Chip's sole selling point. ... Why would you buy a motherboard made by the industry's absolute lowest bidder? --RacoonRider (talk) 04:06, 5 April 2013 (EST)

I read this. My interpretation of this text is that they extend the prior defective rates to the at this time new VX Pro boards. For me this is rather speculative. It could has been as well that while the boards were cheap the defective rates were ok, especially since the TX Pro worked well. But maybe some native english speaker reads there something between the lines.

However, today the focus whats important is different. If you get such a board and it still works then it is still a Socket 7 system. If performance matters just get another board. No manufacturer knows how the defective rates change after 20 years. Enigma (talk) 04:43, 5 April 2013 (EST)

Ok, I think, I got your point:) I really appreciate your attitude, it's pretty much scientifical and seems to oppose any kind of "holy war". I changed a few lines in hope this would be the most adequate angle, at which the user would know what he's dealing with, but would not worry much in case it works OK. Still, I think that the owner of the actual hardware should have the final word here. By the way, can you provide some benchmark results for the special section? It really lacks data collected from non-intel chipsets.

And what do you think of RAM paragraph? --RacoonRider (talk) 05:38, 5 April 2013 (EST)

Well, most importantly that PS/2 SIMMs have to be plugged as pairs in S7 systems is missing. The 3.3V vs. 5V issue is not too critical as most memory chips support the whole voltage range. The heat dissipation as written should be considered. The golden rule for SDRAM is to choose DIMMs with a lot of ICs for high capacities. On the other hand it may not be a problem if only a partial capacity is detected. I would always recommend to run memtest86 at least 1-pass for a large memory setup to find out if chipset and DIMMs work in combination. PS/2 SIMMs are usually not problematic regarding memory layouts exceeding the memory controllers specs.Enigma (talk) 06:08, 5 April 2013 (EST)

Forgot to mention that SIMMS work in pairs. However, that's only partially true for EDO, on 2 of 3 boards I posess EDO RAM works well without the same stick in the adjacent slot.--RacoonRider (talk) 17:06, 5 April 2013 (EST)

A pair means same memory organization. It doesn't mean that it has to be the same manufacturer. Of course the risk is higher with different manufacturers to choose mismatched sticks. But this can be seen by the amount of detected memory. Maybe it is also worth to mention that 430VX started to support EDO, while 430FX f.e. did not. And what about S7 boards supporting BEDO?Enigma (talk) 22:39, 5 April 2013 (EST)

I didn't think about BEDO, I own neither mobos that support it nor modules of it. I'll search the net on the matter. Concerning 430FX, it did support EDO. 430VX was the first intel chip set to support SDRAM. Links:

• http://www.pcguide.com/ref/mbsys/chip/pop/g5iI430FX-c.html
• http://www.pcguide.com/ref/mbsys/chip/pop/g5iI430VX-c.html

--RacoonRider (talk) 01:18, 6 April 2013 (EST)