i am not disputing anything you say above. so, when are we getting a baseline/standardized benchmark?
would ultimately help in evaluating gaming set-ups, right?
what i'm trying to say is for whatever amount of CPU workload AH2 requires, a cheap, lowly Athlon II is enough for a fluid, fast gaming system at whatever resolution tigger's friend plays on. the higher the resolution, the more GPU-dependent this game becomes(1080P and above)- in other words the GPU workload increases due to moar pixels needing to be rasterized. the CPU workload however, i observed, remain constant across all higher HD resolutions. having said that, i dont even notice the CPU workload significantly increase at 1024x768 or 1440x900.
and you probably meant "tenet."
The amount of time to generate and maintain a proper benchmark, based on the game is prohibitive.
And yes, I goofed my spelling of "tenet".
By the way, the Athlon II will stutter very badly if you toss 100 B17 bombers launching at one time in an FSO. It cannot get the data to the video card fast enough for each frame loop. Many think it is because the video card cannot handle it, and while that can impact things, it is not what causes the stutters. Data bound for a video card is, for the most part, completely asynchronous to the CPU/frame loop. As long as the data can get generated, the frame loop runs smoothly. Ther are a few situations where data will block and the CPU has to wait for it to process on the video card, but not many of those happen in the game.
Unusual situation? Sure, but it is a valid one. That is where you find out what is the bottleneck in your computer.
Flight is always complex to model. I remember when we upped the number of test points on the wings and how many people yelled and screamed how it was killing thier performance. Had absolutely nothing to do with the graphics.
I will be the first to admit there is no easy answer to this. The same exact hardware given to ten different people will all run the game differently due to everything else the player will use the computer for.
I recall posting a film of my computer at home running the game. With everything enabled it was butter smooth. Disabling vsync showed it was running over 200FPS, with all the shadows on, in the middle of a fight with a dozen planes around me over a filed with ack bursts going off.
A simple E8400 with 2GB of system RAM and an ATI4850. Well, maybe not so simple. It is very tweaked. Booting Windows XP in 4 seconds flat is not common. Point being, that someone else, whose computer has the same hardware, will not perform the same.
question :
Since AH is, as I understand all flight sim games are, CPU intensive due to the elaborate physics calculations needed to run the engine, my question is this.
Would it bode well for players with a single GPU card to purchase a lower-end (read: cheap) second GPU as a dedicated Physx processor?
Also, players that are already running an SLI setup with 2 cards, would switching the 2nd card from SLI to dedicated Physx, produce a greater overall net gain in performance in this particular application? Would this help alleviate issues of slower CPU builds?
Would this configuration be more cost efficient then going for a higher tier CPU, subsidizing a cheaper second GPU?
The game does not support the Physx processor, nor does it support using a GPU for floating point calculations. Someday, when there is single API that supports all that hardware (ATI or NVidia or...), the game might support it.
