Here's something I don't get about drivers, especially Nvidia, since they seem to release new ones so often.
The product (GPU) is made. The company writes drivers for it. Then later, they tune and tweak it a bit for some gains. I can see this happening a FEW times as they continually fine tune the driver, but after a certain point you have to wonder who are the idiots in charge of making these drivers. Because every time they come out with a "New And Improved" driver, that's essentially the same as them saying "We screwed up the last time and forgot some stuff, but THIS time it's better and more complete and you'll like it. Really."
Just how many driver revisions do they NEED to get it right?
Here's something I don't get about drivers, especially Nvidia, since they seem to release new ones so often.
The product (GPU) is made. The company writes drivers for it. Then later, they tune and tweak it a bit for some gains. I can see this happening a FEW times as they continually fine tune the driver, but after a certain point you have to wonder who are the idiots in charge of making these drivers. Because every time they come out with a "New And Improved" driver, that's essentially the same as them saying "We screwed up the last time and forgot some stuff, but THIS time it's better and more complete and you'll like it. Really."
Just how many driver revisions do they NEED to get it right?
Clearly you've never written software - especially as part of a team. There is almost always a better way to write code to make it faster, add a feature, or make the overall code base more stable. Sometimes one of these comes at the expense of the other two (most frequently stability is the driving factor). Make it stable first, then optimize. In almost zero cases will your first version of the product be the 100% best version - especially when you have a large team working on the code as almost every project does.
I've been a programmer for 15 years, and I still don't get it perfectly on the first try. If it was easy, they wouldn't pay so well.
Comments
If it ain't broke, don't fix it. Huh? Obviously, who ever coined that expression didn't know me.
The product (GPU) is made. The company writes drivers for it. Then later, they tune and tweak it a bit for some gains. I can see this happening a FEW times as they continually fine tune the driver, but after a certain point you have to wonder who are the idiots in charge of making these drivers. Because every time they come out with a "New And Improved" driver, that's essentially the same as them saying "We screwed up the last time and forgot some stuff, but THIS time it's better and more complete and you'll like it. Really."
Just how many driver revisions do they NEED to get it right?
Clearly you've never written software - especially as part of a team. There is almost always a better way to write code to make it faster, add a feature, or make the overall code base more stable. Sometimes one of these comes at the expense of the other two (most frequently stability is the driving factor). Make it stable first, then optimize. In almost zero cases will your first version of the product be the 100% best version - especially when you have a large team working on the code as almost every project does.
I've been a programmer for 15 years, and I still don't get it perfectly on the first try. If it was easy, they wouldn't pay so well.