| CaseyDDR said: you can't call it fanboyism if it's true :p |
AMD v. Intel is debatable, as Intel has the performance lead but AMD is often better value.
But AMD v. Nvidia? Clear cut. AMD has the single-card performance lead, single-GPU performance lead, lower temperatures, power and noise, DX10.1 and DX11, Eyefinity, and Crossfire on all GPUs (GT 240 doesn't do SLI) and all motherboards without license. AMD's GPUs provide more perf/watt and perf/mm^2 [the latter even comparing 55nm to 55nm].
Nvidia's new cards are massively delayed [late Q1 if you believe Nvidia, or Q2 if you don't; revision A3 hasn't even taped out yet and it'll be 3 months from when it does to retail] , even bigger than their current GPUs so will cost at least $400 for the top model if they are to be profitable, and released figures on FLOPS, shader counts and clock speeds as well as extrapolation from the current series [Fermi is better at GPU compute but not many improvements help game performance] suggest it will be 10 or 20% faster than the 5870 - in other words, epic fail vs. the 5970. And that will be 5-7 months later than the actual 5xxx launch, with lower profit margins.
AMD chips are also at least as reliable as Nvidia: look at the admitted G8x/G9x mass chip failures or the Microsoft-data-proven Nvidia Vista driver crashes issue.
Finally, their corporate ethics are questionable. Look up the denial of fake Fermi, or the Batman AA mode that Nvidia paid to be Nvidia-only, or the artificial SLI lock-in and tax, or the endless rebrands (GTS250 = 9800GTX+ = 8800GT = 8800GTS, all based on G80 just with die shrinks, or GTX280M being G9x rather than GT200). Or locking reviewers out of the GTS 250 when they didn't say nice things about PhysX.
So how can you say Nvidia is better, if not by performance or perf/watt or features or price or schedule execution or reliability or ethically or technology advantage or process advantage or profitability?
