NVidia question

[Stemin]

Are the Quadro and Tesla series cards better than the Geforce GTX series?

 

I see them listed as "professional" cards, but is that because they're better quality cards, or do they just serve a different purpose, like being more durable. I see a lot of them listed as for workstations, which makes me think quality isn't as big a deal as being durable and being able to run for a long time.

 

For me personally, I don't always need the most high quality graphics (within reason) as much as I want my programs to run smoothly and quickly. A lot of the so called high quality mods don't do anything for me. I see direct comparisons to the stock graphics and yeah, the stock graphics look a bit stretched, by comparison, but they also BOTH look like textures I would see in real life when looking at things like the straw on a house roof or the wood on the side of a house.

 

My last two cards were NVidia and I prefer them over the others.

 

If I have one complaint from all the computer part manufacturers over the last decade it's the way they've gone away from having progressive model numbers. I realize there are a lot of cards that do a lot of different things, the same with processors, etc. but I feel like there ought to be a better hierarchy structure for consumers.

Edited November 18, 2012 by Stemin

[Dan3345]

Quadro and Tesla cards are for servers or PC's that are doing lots of unique work that goes to the GPU. They would be excellent cards for Bionic or folding, but not necessarily the best for things like gaming and certainly no good for average computing. Though this is not because they lack the capability its simply a huge waist to spend 4 grand on a tesla so you can browse the web.

 

In short, GTX cards are for gaming and a bit of average rendering/computing. Quadro, and Tesla are for massive computational tasks that you wouldn't want your slow CPU to do, but you wouldn't want an average card to work on either.

[Erik005]

Most of the Quadro and Tesla cards are physically the same as Geforce cards, they differ in driver support for professional applications and amount of memory.

[nosisab]

Tesla is optimized for general parallel processing... kind for usage in super-computers.

 

Quadro excels in memory bandwidth, they are meant mainly to work in huge panels, most of times composed by many monitors and/or intensive professional graphic applications.

 

Both have high scalability meaning you can have many of them working together.

 

So, for a desktop, using one or two cards you are better served by the GeForce series.

[FMod]

Tesla processors don't have any video outputs. They aren't video cards at all.

 

Quadro is a soft racket extracting money from the users of aging professional software that lacks support for modern API and refuses to add it even with new versions.

 

It's pretty hard work, crippling gaming cards' capabilities in professional applications without ruining their game performance. Even so it didn't entirely work with the latest series - they had to cut PhysX capabilities, and will probably give it up completely in a few years, because the same modules can be used for computing. Radeons don't have a dedicated GPGPU counterpart, so their compute modules are unlocked, but they aren't allowed to have CUDA, which is a proprietary interface, and it takes time for programmers to move to OpenCL.