10-08-2015, 06:33 AM
Bofox,
It is easy to say stuff like that. "rampage was so great it would have blown everything out of the water" and "it was complete and ready but we had to lay everyone off before it could be mass produced."
talk is cheap. We always here stuff like that. Remember AMD and how they were building a true quad core CPU. A native quad and how much better it was supposed would be than intel's method of stacking two dual cores together. Remember all the hype before Bulldozer and how it was supposed to not onlyt put AMD back in the game but people literally expected it to bulldoze intel. See had AMD went belly up just before any of those architectures, those hyped up rumors would have survived and that is what we would remember them as....not what they actually turned out to be. See, no one expected the pathetic outcome of the original phenom. It was an absolute surprise but not in a good way. The bulldozer was epic as well, an epic failure to such a degree it too was hard to imagine....although by then, there were people who didnt expect much from AMD but even then, very few people thought it would turn out as bad as it did.
It is easy to imagine that 3DFX was just about to release a come back GPU that would surpass everyone else. Actually, as a chip maker in those days, it was an expectation that your fresh new launching chip surpasses other chips that were already out on the market.
It is easy to things and talk is cheap.
the link you have tells us that the Rampage was a long way from launching. Minimum 6 months, realistically up to a yr.
This would have put it launching in 2001. Since we dont know anything but some paper specs, no one could possibly know if it actually would have blown anything out of the water. There is more to it than bandwidth, a lot more to it. For starters, drivers. The "Specter" would have to run DX games, how well they done this is anyones guess.
Paper specs mean little, we all know that. Especially though, this couldnt be more true when we are talking about GPUs. But going by paper specs alone, if we look at the specter specs from wiki, those dont support the guy claiming that it was so far ahead of nvidia.
https://en.wikipedia.org/wiki/3dfx_Inter...bankruptcy
scroll to the spectre "Rampage" on the chart.
now look at nvidias
https://en.wikipedia.org/wiki/List_of_Nv...sing_units
The spectre doesnt look very impressive paper spec wise compared to the Geforce2 gts/pro, which would already have been out. Paper specs alone, its not impressive unless you consider their imaginary x2 card. There existed not a single one, not even an engineering sample. They literally just added the specs together, to represent a card that is made up in imaginations only. It was imagined to be two chips on one PCB. Whether or not that was feasible or how well it scaled is anyone's guess.
Just for the record, the voodoo5 6000 (that unreleased GPU that had a whopping 4 VSA-100 chips on it) had jaw dropping paper specs but in actual performance it would be demolished by the Geforce2.
http://www.x86-secret.com/articles/diver...6kgb-7.htm
The voodoo5 6000 wasnt even feasible in its day. Most people would have literally had to buy an extra power supply just to run that card. But 3DFX was such a mess at this time that they were not only dreaming up these things, they actually were telling the press it was on its way.
But lets forget about the terrible idea of 4 chips on a single PCB. If we look at the Voodoo5 5500, which was a dual VSA-100 power suckle, with its awesome paper specs it couldnt keep up with a geforce1 the geforce 256DDR.
So, the specs of their dual rampage are much more meaningless than the specs of the single chip versions. The scaling of twice a single chip is a joke. The paper specs of the 3DFX Rampage arent impressive, not the Spectre 1000 or the 2000. These specs were not impressive for the 2000-2001 era.
We just have a case where people attached to the company really wanted to see them do something great. And just like the phenom and bulldozer, this thing got hyped up. Sure, i would have loved to see what a rampage could have done in the real world. But that is the thing, they were a long way from launching. Drivers and games are very different than theoretical performance. We dont know how well it could have really done.
Just like today, we hear all this hype about AMDs Zen. If AMD goes under, this hype is all people will know about. That is how it will be remembered. See, 3DFX was a long way from launching their rampage. They just got the Voodoo5 x2 5500 chip out and went under. If you ask me, the rampage was at least a yr away from feasibility when the company went under, it wasnt ready for manufacture. It is hard for me to see how it could have saved them at all.
It is easy to say stuff like that. "rampage was so great it would have blown everything out of the water" and "it was complete and ready but we had to lay everyone off before it could be mass produced."
talk is cheap. We always here stuff like that. Remember AMD and how they were building a true quad core CPU. A native quad and how much better it was supposed would be than intel's method of stacking two dual cores together. Remember all the hype before Bulldozer and how it was supposed to not onlyt put AMD back in the game but people literally expected it to bulldoze intel. See had AMD went belly up just before any of those architectures, those hyped up rumors would have survived and that is what we would remember them as....not what they actually turned out to be. See, no one expected the pathetic outcome of the original phenom. It was an absolute surprise but not in a good way. The bulldozer was epic as well, an epic failure to such a degree it too was hard to imagine....although by then, there were people who didnt expect much from AMD but even then, very few people thought it would turn out as bad as it did.
It is easy to imagine that 3DFX was just about to release a come back GPU that would surpass everyone else. Actually, as a chip maker in those days, it was an expectation that your fresh new launching chip surpasses other chips that were already out on the market.
It is easy to things and talk is cheap.
the link you have tells us that the Rampage was a long way from launching. Minimum 6 months, realistically up to a yr.
This would have put it launching in 2001. Since we dont know anything but some paper specs, no one could possibly know if it actually would have blown anything out of the water. There is more to it than bandwidth, a lot more to it. For starters, drivers. The "Specter" would have to run DX games, how well they done this is anyones guess.
Paper specs mean little, we all know that. Especially though, this couldnt be more true when we are talking about GPUs. But going by paper specs alone, if we look at the specter specs from wiki, those dont support the guy claiming that it was so far ahead of nvidia.
https://en.wikipedia.org/wiki/3dfx_Inter...bankruptcy
scroll to the spectre "Rampage" on the chart.
now look at nvidias
https://en.wikipedia.org/wiki/List_of_Nv...sing_units
The spectre doesnt look very impressive paper spec wise compared to the Geforce2 gts/pro, which would already have been out. Paper specs alone, its not impressive unless you consider their imaginary x2 card. There existed not a single one, not even an engineering sample. They literally just added the specs together, to represent a card that is made up in imaginations only. It was imagined to be two chips on one PCB. Whether or not that was feasible or how well it scaled is anyone's guess.
Just for the record, the voodoo5 6000 (that unreleased GPU that had a whopping 4 VSA-100 chips on it) had jaw dropping paper specs but in actual performance it would be demolished by the Geforce2.
http://www.x86-secret.com/articles/diver...6kgb-7.htm
The voodoo5 6000 wasnt even feasible in its day. Most people would have literally had to buy an extra power supply just to run that card. But 3DFX was such a mess at this time that they were not only dreaming up these things, they actually were telling the press it was on its way.
But lets forget about the terrible idea of 4 chips on a single PCB. If we look at the Voodoo5 5500, which was a dual VSA-100 power suckle, with its awesome paper specs it couldnt keep up with a geforce1 the geforce 256DDR.
So, the specs of their dual rampage are much more meaningless than the specs of the single chip versions. The scaling of twice a single chip is a joke. The paper specs of the 3DFX Rampage arent impressive, not the Spectre 1000 or the 2000. These specs were not impressive for the 2000-2001 era.
We just have a case where people attached to the company really wanted to see them do something great. And just like the phenom and bulldozer, this thing got hyped up. Sure, i would have loved to see what a rampage could have done in the real world. But that is the thing, they were a long way from launching. Drivers and games are very different than theoretical performance. We dont know how well it could have really done.
Just like today, we hear all this hype about AMDs Zen. If AMD goes under, this hype is all people will know about. That is how it will be remembered. See, 3DFX was a long way from launching their rampage. They just got the Voodoo5 x2 5500 chip out and went under. If you ask me, the rampage was at least a yr away from feasibility when the company went under, it wasnt ready for manufacture. It is hard for me to see how it could have saved them at all.