Except that it blows cack
First of all, if you want a comparison of video cards, there was a post about it quite a while ago:
As for a good budget card, people have generally been talking about the ATI Radeon 5770 (~$160) and the Nvidia geforce gtx 460 (~$200-$230). The 460 came out pretty recently so it's not on that list up there, but you can do a google search for gtx 460 reviews in order to get a good idea of how it compares to other cards. Also, the 460 has a 768MB and a 1GB version; the 1GB version is about $30 more. You'd probably be best off if you look at the benchmarks in the reviews so that you can choose your card based on how much extra you're willing to pay for better performance.
As for A) I can't really say, the world is kind of moving on and accepting windows 7. It's possible that things could run marginally faster in XP, but you might face compatibility issues eventually.
As for B) it's like what was said above. ATI and Nvidia design the cards, but it's companies like ASUS that actually build and sell the cards. I can't really say which brand is best, there may be some minor differences in performance but unless they added some kind of custom design on their card, they're all about the same.
Getting a bit antsy waiting for my new PC (delayed another few days for a faulty part)..
I don't suppose anyone would have any idea of what it might get in the benchmark with no overclocking?
Phenom II 965BE @ 3.4GHz quad
4GB 1600MHz DDR3 RAM
HD5850
Gigabyte GA-890FXA-UD5
Any idea on High/Low scores all at stock?
A 4870 isn't that much slower than a 460 to begin with, but you should see SOME fluctuation on high. Possibly a CPU restraint, since my 470 was getting around the same score on a similarly clocked C2Q before I upgraded, then my score shot up.
Post pictures of the 460 runs though and I'll get it on the board, I know people were lookin for em.
Hmm, that makes sense. I was wondering why on 720p neither CPU nor GPU seem to be at 100%.
If on 1080p my GPU runs constantly at 100%, does a CPU upgrade still make sense?
To give a bit more info.
CPU: E5200 @ 3.5Ghz (333x10.5)
GPU: GTX260 @ 655/1412/1280
RAM: 4GB DDR2 @ 5-5-5-15 1109
I'd like to run at 1680x1050 with maybe MSAAx4 and DoF. What should I upgrade first? CPU? GPU? Both?
Benchmark 720p: 3820. (Both CPU cores around 60%, GPU 70% with a few 99% peaks)
Benchmark 1080p: 2980. (GPU constantly at 100%)
If your GPU is at 99% the entire time on High, it's safe to say you wouldn't benefit much from a CPU upgrade on those settings since its already working as hard as it can. You could most likely 100% confirm this yourself by backing off your OC and running High once at stock. I may be overestimating the 260, but I don't think you'd still maxed out with your CPU at 2.5ghz.
From there, you could extrapolate a little and guesstimate what you'll get @ 1680x1050. So like, you're 70% load on low(720p).. and some percentage above 100 high(1080p). If you drop your cpu to 2.5 and get sayyy, something like 80%, then run 3.0. If that gets 99%+ again, you know your 3.5 OC is a good deal faster than your 260 can handle. Whereas if you drop to 3.4ghz and your GPU usage immediately drops, you know you're just barely on the brink of maxing your GPU. Follow?
FYI- the radeon 5870M runs at 99% almost the entire benchmark on the G73JH-A1 asus laptop. Other than SLI or custom machines its the best laptop card out. The i7 seems to run at between 25-40% load while running the benchmark. Think i might save up and put in a SSD in a little while as its about the only thing i can think of that would further boost performance.
You could in theory experience a stutter/hitch with poor texture streaming on a shitty engine with too little RAM, but your framerate would still be unaffected. Exceptions being like, a 4200rpm drive or something equally ridiculous under constant pressure. Trust me, this arguments come up before here and I had to resort to peer pressure to hammer the point homeMajority of votes for 'possible' were for the above exception if you read commentary, probably should have clarified that at the time for the clean sweep lol.
Anything you did to your beta folder was probably done in between tests and rendered irrelevant/coincidence.
unless their engine is written on public tech from pre 2005, or the word unoptimized is being purely charitable and badly programmed flash is literally more efficient, yes.
remember, 720p is a *small* res for a pc. (1280x720? we've had videocards running 1280x1024 for over a decade) while 4xAA and depth of field require some firepower, they shouldn't require 10 revisions of hardware firepower.
to the guy asking about how to read videocard numbers: the first number is always the 'series' (which generation of tech) everything after is model within the series; higher numbers are generally better except when they reset naming conventions (like both companies did recently) but the second set of numbers are generally more important.
eg. a 4200 is a 4000 series, but a low end (200) a 4800 is a 4000 series, but high end. generally, last year's 800 is better than this years 600 or lower.
it gets complicated past this, but that's the 'basics'. you've got some months to research, start googling for 'graphics cards 2010 review' or similar topics, and read up.
I won't point you straight to tom's quarterly round ups because they shouldn't be where you start, or you might be too tempted to just blindly follow their advice..