Look at the sh** on hardocp.....

[Scout255]

Some idiot sent in an e-mail rant about matrox hardware not being able to handle AGP 4X....
here is the direct cut+paste
Richard Sanders gives us a heads up on Matrox. Seems the G400 that was touted as being 4X AGP compliant is not. I will just post the whole mail for you to check out. Not to be purveyor of bad stuffs, I don't think Matrox will be the last to have to do this. OUCH.

Seems the first batches of the G400 MAX's wont cut it on AGP 4x transfer mode, and Matrox has released new spin on the boards that can handle it. I know that's old as hell but what I was interested in was if you knew if Matrox was gonna bite the bullet and let all of us that waited 8 damn weeks for our cards trade them in for one that actually lives up to the specs they published. They have a nice "Specifications subject to change" on the tech info now, so we are most likely just gonna have to bend over and take it.

More of a rant than news, but yet again a graphics card company has screwed their customers.

Dunno bout you, but i have never see an official matrox document that specifically shows "AGP 4X Transfer mode compatable"
and it definatly isn't on my box. I'd also like to see a company that would pay millions to replace a card so that it can run 4X.... I still doubt that there would be much of a performance increase. It is simply trivial. As for matrox "Screwing" there customers, how the hell did they do that?
I am very happy with my card, and don't give a crap wether it will fit in an AGP 4X Mobo. Like comeon here, who can honestly say that they will be useing this grafics card in a year when this technoligy becomes widely knowen.

This site gets over a million hits/day. I still can't beleve he posted crap like this. I really doubt that he would do a something like this to Nvidia or 3DFX Because he KNOWS they have never stated 4X Transfer rate. However, he knows nothing about this subject!
ARRRRRRRRRRRRRRG!

[Kruzin]

Go have a look ar hardocp now...seems they are in the "equal time" mood today

[agallag]

let's all go out and get a blasted on Scotch and throw rocks at an Electronics Superstore all night

Sounds like the best bit of advice on this topic in a long time...

[Scytale]

Still, Full AGP2X/4X device as stated in http://www.matrox.com/mga/press_room...f/chipspec.pdf is a tad misleading, to say the least.

[SteveC]

Yeah, the chip is, the card is not....

------------------
Cheers,
Steve

"The chances of anything coming from Mars, are a million-to-one", he said.

[Buffy]

This is my theory

What should really be the target here is the delay of the motherboards using 4x. Matrox didn't want to release their cards being 4x to find out that this would cause trouble/conflicts once the boards were released. Remember intel delayed these motherboards a long time. When the announcements were made I am sure matrox *thought* they would have ample time to test them on a finished intel motherboard. So rather than release them not being tested they went the only road they found posible.

So to sum it up in order to do what a manufacturer wants to do they depend on other manufacturers to do what they do. When intel announces a release date then pushes it back it affects all hardware being built to be compatable with it.

Remember this is MY theory

[The PIT]

The Intel motherboards were delayed due the fact that third dimm slot would cause problems if used. So Intel eventually did a quick fix (After thinking about it for ages) and just removed the third slot.
Now who tested those boards I wonder if there still working for Intel. Probably but a tad embarressing for them.

[Mark F]

Ahhhh the third slot...
Well that was not a RDIMM or chipset problem,
the problem was with the construction of the INTEL boards, so they redid their boards with two, and forced others to comply.
Other manufacter's boards with three RDIMM slots worked fine, before INTEL's antics.
Personally, it's a doomed chipset.

Mark F.

------------------
OH NO, my retractable cup holder swallowed a CD

[The PIT]

Last time I heard they were still working on a redesign that would work with 3 sockets. Suppliers can use three but there's no guarntee that they work reliable. A bit like few early Bx boards with 4 Dimms.
As for a doomed chipset that depends wether Rambus drops in price by a lot.

[CrazyBee]

Heh...

more like if RAMBUS becomes available in its faster speeds and also drops in price. ALOT.

the only RAMBUS available is the first speed which is also insanely priced and the two other faster speeds are either non-existant and/or insanely insanely insanely priced.

SDRAM will be here until RAMBUS becomes more avalable. Just like EDO was still around for a good time after SDRAM was being widely used and became dirt cheap.

CB

[Wombat]

SDRAM, or DDR RAM, will be here longer than Rambus. SDRAM shows a marked improvement with increases to the FSB, and right now SDRAM @ 133MHz meets or exceeds Rambus performance.

[Scout255]

But still, who knows. What happens if intel launches a chip that runs at 800 Mhz FSB and a 1 multiplier. Only rambus would be able to handle it....
man, that would just kick ass.... *drool*

[Mark F]

FYI - PC800 RDIMMs run at 400MHZ. The next chipset with a jump in FSB speeds will have a 200MHz bus.

-----------
Why do I beleave the i820 is doomed? Well, yes RDIMM are too expencive now, but as you said they will fall in price. However that's still a ways off. Those who are not put off (or need it's advantages enough to justify) by it's high price are better severed right now by the i840. Those looking to upgrade from the BX chipset are most likely not going to be willing to make the large inverstment in a RDIMM system with only minimal performance gain, and the SDRAM i820 cludge is actually slower than BX. In the next few months INTEL will release the i815 all the advanteges of i820 but supporting PC133, a much better at a lower cost solution.
Sure one day RDRAM will not be as rediculusly high, but by then there will be other competative memory solutions, like DDR, and even INTEL will have released newer chipsets. So maybe RDIMM will proliferate but the i820 is not the vehicle that will propell it to a lare audence. Hence my asertation that it is a doomed chipset. Just my beleave.

Mark F.

------------------
OH NO, my retractable cup holder swallowed a CD

[Bohrn]

Hey Scout255, it's good to hear you saying nice,(even patriotic),things about Matrox now. I remember when you were really mad and wanted a TNT2U or some other such Nvidia card. And we(well others here, not me) got ya thru it. Nice to see you are still here and happy with the G400.

------------------
AsusP2B ,iCeleron525(7x75), 128megs PC100, G40032megSH, Yamaha PCI sound, 2 small HD's, 42X Sony CDrom and 98SE w/shutdown patch, PD 5.41 w/Turbo MCD and beta ICD

[Agent31]

I think there should be a law written that restricts all companies from claiming anything other than the facts about their hardware.

Future press releases would read:

"Matrox video card - can be plugged into your computers motherboard. Can also connected to your computer monitor."

This way, you'll get a pleasant surprise when you find your computer will start up, and you can see something on the display, instead of being disappointed when you find that it doesn't support AGP 4x transfer mode and fast writes that haven't, and probably never will be, used for anything other than bullet points on packaging.

------------------
PIII 500@560 (2.1V) - ABit BH6 (NV) - 128MB Corsair CAS2 - 3D Blaster GeForce (135/183) - 12 MB Voodoo2 - SBLive! Value - ViewSonic PS775 - 8.4 + 4.3Gb Quantum Fireball CRs & so on.