Holy Crap This Must Be A Misprint!

[aphremen]

a geforce fx for 99 bucks? (http://www.newegg.com/app/viewproduct.asp?description=14-170-024&DEPA=1)

Really though, is it a misprint? Are they high?

[never mind, didn't realize this thing was the mx-ified FX chip...]

[sushi128]

it's less than MX-ified. an GF4 MX440 has better specs than this. crippled at a 64bit memory bus. the only reason someone would buy this is for Dx9 support.

[ReDeeMeR]

I wanna see someone run anything DX8 on this junk, let alone DX9... FUBAR!

[future man]

that was the stupidist 100 bucks i ever spent. I just want an FX card for the ol' museum

[sushi128]

this card can be had for 78$ on pricewatch. cheapest 128mb GF4 is 109$. this card is a steal as far as Dx9 goes. i see this goin in the OEM market (dell, maybe). and quite honestly, i believe it'll play any game out there (probably not high quality, but it'll be playable).

[aphremen]

geforce4 Ti4200 is pretty good as far as lower end cards go. I picked up mine for $120. I tend to shop at Newegg 'cause they may not be the cheapest but they are VERY cheap and returns are no prob with them. They even took back a mobo I returned that had a few scratches. :D

[Rhenna]

I just installed one of these FX5200s in someone's computer this afternoon! It was an eVGA brand, a non-Ultra 5200, clocked at 250/200. (core/memory)

This was a road-trip, and all I had with me were the latest 4-in-1's from VIA. System consisted of:

Chaintech KT-133 Mainboard
AMD Athlon T-Bird @ 1.1GHz (11 X 100)
256Mb PC-133 CL2 SDRAM
GeForce2 GTS (32Mb DDR)
CMI 8738 PCI Audio
Maxtor 'DiamondMax 8' 40Gb HDD (7200rpm)
LiteOn CD-RW, 32X-12X-40X
Win98SE

This friend of a friend wanted to, "get a video card with something more than 32Mb of RAM". He didn't like the GeForce4 MX cards on the basis of them just being DirectX 7 tech. He isn't a rabid gamer, so he thought that a GeForce4 Ti 4200 would be wasted on this box. He spotted the 5200 on the web, liked the price, ($99), and thought this was the way to go. DirectX 9, 128Mb of DDR, DVI and S-Video outputs, etc.

So far, I was thinking, "Yeah, I can understand his logic, but I probably would have sprung for a 64Mb Ti 4200".
I check out his BIOS and system and find he's got no issues, and I'm pleased to see there's nothing running "behind the scenes" slowing the system down. His 4-in-1's aren't THAT old, so, for now, I leave well enough alone. I see he has 3DMark 2000 on his drive, and think, "Let's get some numbers for a little before and after comparison".

After a re-boot, we run 3DMark 2000 with the default settings and get a score of 6280.

Do the uninstall/re-install of the video cards and drivers, (Version 43.03; very nice installation routine, BTW), along with the supplied DirectX9. Everything looks good. Re-boot and run 3DMark 2000. Get a score of 5750. Hmmm...

"OK", I'm thinking, "the old card skips the environmental bump-mapping, the new one doesn't, that's going to affect the score a little, I guess..." The friend of a friend isn't happy. "I figured it would be higher, not lower", he says softly.

"Well", I say, "let's hit the Microsoft site and get DirectX 9.0a, and then see if there are later drivers for the card." We do that. Install DirectX 9.0a, AND the 43.45 drivers, AND the VIA 446 4-in-1s. Re-boot. Run 3DMark 2000. New score of 6210. (I'm, of course, mentally kicking myself for running this bitch in the first place.) I give him my best smile and a little eyelash action and say, "Well, you're within 75 points of the old card; that's a negligible difference. When they refine the drivers a bit more, I'm sure it will get better". "Where's the door?", I'm thinking.

So, we run the MadOnion demo at 1280X1024, 32-bit color, etc. and the card doesn't skip a beat. (???) Fire up Serious Sam 2E with 2X AA and AF, and Sam is lookin' pretty sporty. Same settings for UT, the only other game on the drive; it looks and runs great... ...but it's a 1999 game. Card comes with America's Army and Ghost Recon, (full), so that's something, I suppose. The friend of a friend seems a bit better, so I bolt.

Tonight, I learn that MY friend goes to HIS friend's place to check out the new 5200, and takes the BloodRayne demo with him. The bump-mapping murders the framerate at 1024X768 in Direct3D. There's a lot of tension in the air over there, and it's not because of Rayne's cleavage.

OK, this is NOT a high-end rig, but it's not toothless, either. A faster CPU and better memory bandwidth would help, but the true gamer isn't going to opt for a $99 card in his/her XP2800+ equipped DDR mobo, so where's the niche for this card?

My advice: Pass on this, and the 5200 Ultra, as well. Personally, I wouldn't be too quick to order a 5600 Ultra, but I'd like to hear about the card from someone who rolled those dice.

[RAY16]

lol, i got a GeForce 3 for $79 total at Newegg.com. It performs far better then that GeForce FX 5200 you just talked about. Too bad GF3's are DX8 cards, but all the games i've run so far have no problems, and its way faster then my old GeForce 4 MX440 64Meg DDR. Splinter Cell is so cool, it looks great, and it runs fast too. Bumpmapping doesn't seem to bother my GF3 either. I'm glad i decided to go with my instinct and get the GeForce 3 instead of the GeForce FX 5200, i was thinking about getting it.



Full specs:

Athlon XP 2000+ (1.66Ghz)
Asus A7N8X mobo (Non-Deluxe)
256Megs PC-2700 DDR RAM
Gainward GeForce 3 Ti-200 64Meg DDR
7200RPM 20Gig HD
Bunch of other crap, too lazy to type the rest.


*Edit*

Heres my scores for all the 3DMark versions.

3DMark2000: 11025
3DMark2001SE: 8235
3DMark2003: 1023

[WhoGivesARatsAss]

ray finally got rid of the fx huh.. good for u son.. but u would have been a little better with a gf4 ti 4200 64 meg at that price.. damn newegg i guess.. compusa had the asylums for 80 bucks after rebates and such.. dunno abt now but this was a week ago.. but anything is better than the shitty MX.. good for u dude!

[sushi128]

silly old bear..........my Ti4200 didn't score well in 3Dmark2000 either. these cards aren't meant for a Dx7 environment. a GF2 ultra will beat a geforce 3 in 3Dmark2000. the GFFX 5200 is capable of about 1300 marks in 3Dmark03 (u don't believe me, check the ORB), which is more than i can say for the GF4 MX, which will run only ONE test. in comparison, the GF4 Ti4200 scored an average of 1700 with only three tests (compared to the GFFX running 4). so, while the Ti4200 remains the undisputed bargain card in the realm of performance vs. price, the 5200 should at least be given credit for price, if not performance :)

btw, RAY16.....1300 beats 1023 hands down. and the reason i say it is not as a joke, but to point out that the cpu speed seems to not affect 3Dmark03 AT ALL. a person with a 1.2ghz Tbird and a 5200 is getting 1295 :D

i overclocked from 1730 up to 1890, gaining 350 pts in 3Dmark2k1, and 3Dmark03 dropped 5 pts.

this card IS kinda cheap, and since i'll be building a rig for someone, i'll stuff them with one of these. that way i can play with it a little, see how bad it really is. i'll throw some Unreal 2 at it for a test.

[ReDeeMeR]

sushi128:

What's the friggin point of having DX9 and runing at 3-5fps???? (And dont tell me it runs faster on 3DMark03-Battle of proxycon/nature.

Geforce 3 is by "FAR" better investment. We aint gonna see DX9 games for a while anyway.

But this is a good move by Nvidia, DX9 for everyone(even if they wont be able to play anything :p) will make sure SX9 games will ship faster.

[sushi128]

i used two similar scores in my searching. proxicon and trolls lair is Dx8, so we'll use that. it seems they score just about equally in both (7.5, 7.8 vs 8.1, 6.5). so how is it that the Geforce 3 is the better buy now? i'm comparing this to a Gefroce 3 Ti500, with mature drivers and the like, btw. cheapest Geforce 3 Ti500 is $100. like i said, GFFX goes for $80.

[ReDeeMeR]

Still Geforce 3 is faster card and 20$ isnt much. Yes I know they are both DX8(proxycon and trolls), but I was just making my point that you wont be able to play anything DX8-9 with that card.

[RAY16]

This $79 GeForce 3 was all i could afford, so i couldn't get anything else. And even if i did get a GF4 that had rebates i wouldn't be able to pay for it up front, plus i hate waiting 8 goddamn weeks for a rebate to arrive. And about 3DMark, i really don't care what my score is. 3DMark is crap, all it does is generate eye-candy. DOOM III looks just as good, if not better then the battle of proxycon and proxycon gave me low ass FPS (10FPS average). I know DOOM III will run at playable speeds, eince a GeForce 3 is what you'll need to run it with all features enabled.


3DMark = T3h Sux0rZ!

Real Games = T3h OwNaGe!!

[sushi128]

the point of 3Dmark is not to give u a good representation of game performance (not anymore, anyway). it's just a standard in which to compare your performance. and redeemer, i'm not a GFFX or GF4 MX fan at all; horrible excuse for even budget cards. but people must begin to understand that nvidia is not stupid. you'll notice that the GFFX 5800 ultra is NOT WORSE than the 9700 pro. to make a card like that would be suicide for the company. sure, it may have anywhere near the expected performance, but then again, ask yourself; does it beat the Ti4600?

[RAY16]

I still don't beleave 3DMark is worth all the praise people give it. It can't tell you how your computer is going to perform in a real game. The test's in 3DMark don't give you an idea of how well your computer will perform in a 32 Player battle UT2K3 with normal or high settings and full karma enabled. It just doesn't seem to be worth downloading. The only reason i downloaded it is so i can look at some pretty graphics. But don't flame me, i'm just stating my opinion. And i don't claim to be an expert on computers or computer graphics either.


*Edit*

And even though 3DMark doesn't give you a good idea of how real games will run. That means its only purpose is to give bragging rights. And IMO, anybody who needs to flash a high 3DMark score to feel good obviously doesn't feel very highly of themselves. High 3DMark score = Virtual version of a big penis.

[Rhenna]

I can understand, I think, why a DX9-compliant card actually running under DirectX 9, might give poorer results on a DX7 benchmark than a true DX7 product. If it's attempting to render a given scene, simply put, more completely, it will fall behind it's DX7 counterpart. (?)

That being said, it's still disturbing that a new product with a higher core and memory speed is struggling to even keep pace with a 2 or 3 year-old predecessor.

I'll take sushi128's word on 3dMark 2000 scores not being representative of true FX5200 performance. And, I've long felt, as RAY16 states, that it's NOT an accurate representation of actual game performance. But don't you two feel that the 5200 might indeed NOT be as quick in the older titles? If it was my only PC, I gotta tell you, I'd have to go with the GeForce 3, given a choice between those two.

The boys called earlier and told me they were busy messin' with the Freelancer demo. I haven't seen Freelancer; they tell me that DX9 is recommended. And, to lend support to sushi128's view, they report that it's running well with 2X AA and AF. So, maybe this IS FX5200 territory.

The highlight of the call was when I was told that there's some Oriental chick in the game by the name of Julie. (?) Anyway, they claim they witnessed, and I'm quoting here, "INTENSE VERTEX-SHADER ACTION" on this character's torso. So, apparently things are more mellow over there tonight. I know two guys who desperately need girlfriends.

Oh! Almost forgot! I like your name the way it is, Ray16. ;)

[sushi128]

well, my point is that you guys speak as if the 5200 was meant to replace the Ti4200. it's 100$ retail. u know what the Ti4200 first retailed at? 200$ (128mb version), and it shows, doesn't it? the 5200's territory is in the low low end rhenna. i'm talkin about people who don't even check the video settings before starting to play a game. also, the 5200 is clocked at 250/400. so technically, u could argue that the Geforce 3 Ti500 had slightly higher clocks, hehe (or u could argue stuff like DDRII LMA and all that).

what older titles might u be speaking of? if it's comparable to the wings of fury test in 3Dmark03 (i use it only as an example), then you're really comparing 2 high frame rates.

and btw, 2X AF for nvidia is really 1X AF. AF, by definition uses 16 filter taps per pixel, and nvidia's 1X uses only 8, so doesn't count. that would explain the break in nvidia's trend of AF vileness.

oh, and RAY16, your name is retarded. change it ;)
(jk......)

[aphremen]

Hmmm...my initial post was stupid, but it is now cool. nods to himself proudly Yes, this is good.

CPU speeds can make a difference in 3dmark, but it needs to be a large difference in CPU. IE:when I upgraded from duron 1ghz (OC'ed to 1.33ghz) to athlonXP 1.47ghz (OC'ed to 1.6ghz) and went from ~7,500 3dmarks to ~10,000 3dmarks. That's not only higher clock but also more/faster cache, blah blah blah.

It doesn't mean much though, mostly I care about minimum frame rate in UT2k3, BF1942, and many other games with cool acronyms. I also agree with Rayman16 that a high 3dmark score is something akin to a flashy car or building the largest conventional weapon (not that it ain't cool, it most certainly is!).

[edit] I should also note that watercooling, overclocking, having a case window, cold cathode lights, and painting your computer the same color as a blue Honda might also be included in the list of making up for lack in another area.

[sushi128]

cpu speed does NOT MAKE A DIFFERENCE IN 3Dmark03. that's it. no argument. check the ORB. that's why i consider 3Dmark2k1se much better in comparing overall system performance. and if you overclock to get higher 3Dmarks......why wouldn't you expect to get higher frame rates? it's consistent. if you get lower marks through overclocking, it's probably because of heat, and will also affect your games. 3Dmark isn't much good at comparing say, a geforce to a kyro (very evident in the games vs. 3Dmarks issue). but if my geforce 3 gets 1000 more points than your geforce 3, i can certainly guarantee myself higher frame rates.

and what do u mean by your last comment? people overclock because of their lack of...."wanting to pay more"?

[unskinnybob]

My Ti4200 puts out around 12000 on 3dmark 2001 and around 1600 on 3dmark 2003 (haven't tried with DX9, but apparantley no card support).

In South Africa you can pick up a Galaxy Geforce Ti4200 8x AGP for around $199. That is considered cheap, since Galaxy is not exactly a world renowned brand. I paid around $210 for my MSI 64MB Ti4200 back in 2001 when they came out.

[aphremen]

Other specs?