The highest resolution my monitor can do is 1280x1024. My TV can do 1366x768. I guess neither of those are really big resolutions. I only do light gaming, particularly Prince of Persia and Tomb Raider. I really would like my computer to be able to play the next games on those series that come out later this year. Also, considering I have a Turion ML-37 clocked at 2.4Ghz and 2GB PC3200 in my desktop, will there be any bottlenecking?
I was leaning more toward the 8600GTS. I was just curious as I have seen benchmarks that shows a 2600XT beating the 8600GTS and vice versa. And knowing that the Sapphire 3650 at newegg supposedly has the ability to clock up to the same speed as the 2600XT and has higher memory clock speeds, I just had to be certain. Thanks for the help.
Christopherr, stop trolling on these forums with all the misinformation about how to judge the power of components. Your comparing to unlike brands, and your technique is completely flawed. I assume you'd say that the GTX 280 is much less powerful than an FX5950?
From my past experience with newegg, if they can no longer get the same card for you, they will refund your money. Things may have changed, but that's what happened to me on an 11 month old 9800 pro that died. If they will do that you could put the $65 toward any card...maybe step up to an 8800GS or HD3850 if within budget and your system is up to it.
Cleeve, what makes you say the 2600XT is better then the 3650? From what I've seen, the cards are virtually the same.
Chris, haven't you been banned yet? I thought the ape told you to stick to your own thread?
@OP, if they are going to refund the money you paid, $65, thats not a bad deal. If you can get the 8600GTS, thats a better card then the 2600XT/3650. If you have any money to add to it, I to suggest getting the 8800GS. (the 3850 has also hit the $100, it too is a good card.) The 9600GT is also not a bad card, but they tend to be around $125.
2600XT has higher core clocks and comes with 700(1400) GDDR3 up to 1150(2300)GDDR4. The HD3650 lower clocked core (almost halfway between a 2600 pro and 2600XT) and mem from 500(1000) GDDR2 to 800(1600) GDDR3. (edit: OC'ed models to 900(1800) mem. )
In general (but not always) for gaming I'd say:
GDDR4 HD2600XT > GDDR3 HD2600XT >= HD3650 GDDR3 > HD3650 GDDR2 > HD2600 pro (GDDR2)
From a review I read on the 3650, they tweaked the core a bit to compensate for the lower core clocks. As you pointed out, there are 3650s out there with higher memory clocks. I need to get ready for work, but I'll see about posting a review of the 3650 that tells what they did to the core. Its close to the 2600XT, but tweaked a bit. (it was from an odd website, I'm not 100% sure I can find it.) You can always look at real world benchies, they should show the two cards very close to each other.
One of those reviews showed the 3670 against the 2600XT. I've heard rumors about the 3670, but have never seen one in the wild. (or for sale for that matter.) There were the reviews with the DDR2 versions against the DDR3 version, no surpise who wins there.
About to give up hope, I clicked on the Digital Life benchmarks and hit paydirt. (meaning I'm not crazy.) I clicked on several 1280x1024 benchmakrs, and noticed the 725/725/1600 3650 about equal with the 800/800/1400. Even the 800/800/2200 wasn't much faster. For example,
In Crysis at 1024x768, the DDR4 2600XT scores 9.1FPS, while the 3650 scores 8.8, a mere .3FPS difference. Again, as long as we are all talking about DDR3 or better versions, I don't find much difference between the 3650 and the 2600XT.
Yeah, that's why I posted the digitlife main page so you could check all games/resolutions not just the crysis one I linked to. Problem with digitlife, we are talking in most tests settings that are far from playable anyway on these cards. (like 9 vs 8 vs 7 fps.) For crysis, we need to see medium detail tests to see which is best. Also they now do too few actual games.
I'd like to see driverheaven do their actual gameplay of 10 games with both versions in GDDR3 format, but I doubt there is much of a difference nor feel it's worth the lab time to test.
Thanks for all of the input and the extra review sites. The cost to ship the video card back was less than the cost for a new fan ($9.80 to ship back vs. $25 for a new fan). I'd love to put any refund towards a 3850, but seeing that I just graduated college and have yet to find a job, I'm broke. My choices are quite limited. Also, last time I had to return something was when my hard drive died a few years ago. Since they no longer had that model in stock, they offered me a different. In turn, I suggested a different one then their suggestion and they gave that to me. I like newegg; I just hope they haven't changed that policy.