My FX5200 - 4th 5200 on 3DMark03

[CPUMelting]

I think the reason FX5200 cards get such a bad rap is because most are 64-bit. Same thing with the 5600 series if you don't watch out. Just make sure the card has 8 ram chips total (rough, but pretty accurate) to get a 128-bit model, then overclock it.

The reason I wanted an FX card is simply because it has digital vibrance and sharpness sliders that make my 17" LCD look good. I got a retail MSI FX5200 (stripped of its temp monitoring of course, bummer). In 3DMark2003, after overclocking, it smokes all the top ATI 8500, 9000, 9100, 9200 series by a good margin, and certainly these are already overclocked too.

Actually I'm only a dozen or so points away from 3rd place in the 3DMark2003 FX5200 database, without even really trying. Just used StarStorm 53.03x drivers, and overclocked GPU (313/595) and CPU (XP 1800 > 2352Mhz, 224FSB). $85 Vid card, $50 CPU. BTW my "CPU Mark" is identical to Amd 64-bit CPU at 2000Mhz. But my Epox 8RDA+ was $85 and my 512MB stick was $80.

I mostly use my Xbox for gaming, hence my choice of the 5200. I like driving games, not so many good ones for the PC.

3DMark2003 - 2054
http://service.futuremark.com/compare?2k3=2209933

3DMark2001 - 9912
http://service.futuremark.com/compare?2k1=7651393
Everyone with this score and higher is "cheating" with the 44.03 dets. I don't think it's a coincidence that every score above mine uses 44.03's.

[Sazar]

Originally posted by CPUMelting
I think the reason FX5200 cards get such a bad rap is because most are 64-bit. Same thing with the 5600 series if you don't watch out. Just make sure the card has 8 ram chips total (rough, but pretty accurate) to get a 128-bit model, then overclock it.

The reason I wanted an FX card is simply because it has digital vibrance and sharpness sliders that make my 17" LCD look good. I got a retail MSI FX5200 (stripped of its temp monitoring of course, bummer). In 3DMark2003, after overclocking, it smokes all the top ATI 8500, 9000, 9100, 9200 series by a good margin, and certainly these are already overclocked too.

Actually I'm only a dozen or so points away from 3rd place in the 3DMark2003 FX5200 database, without even really trying. Just used StarStorm 53.03x drivers, and overclocked GPU (313/595) and CPU (XP 1800 > 2352Mhz, 224FSB). $85 Vid card, $50 CPU. BTW my "CPU Mark" is identical to Amd 64-bit CPU at 2000Mhz. But my Epox 8RDA+ was $85 and my 512MB stick was $80.

I mostly use my Xbox for gaming, hence my choice of the 5200. I like driving games, not so many good ones for the PC.

3DMark2003 - 2054
http://service.futuremark.com/compare?2k3=2209933

3DMark2001 - 9912
http://service.futuremark.com/compare?2k1=7651393
Everyone with this score and higher is "cheating" with the 44.03 dets. I don't think it's a coincidence that every score above mine uses 44.03's.

you realise the smokeage in 3dm03 is completely m00t since none of the other cards are dx9 and yours is.. when you factor in the large percentile of the score calculated using nature you SHOULD smoke the other cards anyways.. if the margin is small than thats just embarrasing...

to get a true gauge of the comparitive performance try out 3dm2001SE and compare your card there... also do a compare with gf4 ti's and you will see a similar pattern emerge...

[saturnotaku]

Originally posted by CPUMelting
I think the reason FX5200 cards get such a bad rap is because most are 64-bit.

Even the 128-bit 5200 cards pretty well suck. The 5700 non-ultra is a far better card for not much more money. You can pick them up for less than $130 now, and board members who own these cards absolutely rave about them.

after overclocking, it smokes all the top ATI 8500, 9000, 9100, 9200 series by a good margin, and certainly these are already overclocked too.

That's because those are DirectX 8 cards. If you want a better comparison, look to the Radeon 9500-9600 cards as those are DirectX 9 parts like that 5200.

Long as you're satisfied, though that's all that matters.

[CPUMelting]

Originally posted by Sazar
you realise the smokeage in 3dm03 is completely m00t since none of the other cards are dx9 and yours is.. when you factor in the large percentile of the score calculated using nature you SHOULD smoke the other cards anyways.. if the margin is small than thats just embarrasing...

to get a true gauge of the comparitive performance try out 3dm2001SE and compare your card there... also do a compare with gf4 ti's and you will see a similar pattern emerge... DX9 doesn't matter to me, or this comparison. The other cards are in the same price range, and I said the margin was not small in 3DMark03.

Look at my post, I did use the latest 3DMark2001. Top GF4 Ti's are around 20K, and guess what, all use the "cheating" 40-44.03 drivers. If I used them, I'd probably get 12K or so, to be expected due to the chip differences. Everyone knows a GF4 TI can smoke most FX's in DX8.1, that's old news.

[CPUMelting]

Originally posted by saturnotaku
Even the 128-bit 5200 cards pretty well suck. The 5700 non-ultra is a far better card for not much more money. You can pick them up for less than $130 now, and board members who own these cards absolutely rave about them.
[b]
That's because those are DirectX 8 cards. If you want a better comparison, look to the Radeon 9500-9600 cards as those are DirectX 9 parts like that 5200.

Long as you're satisfied, though that's all that matters. Haven't seen a 5700 for $130, perhaps you have, it wouldn't have mattered to me, as I've stated my reason for a $85 GFX card. PC games, to use your wording "mostly suck" to me. With a few exceptions.

BTW, again, everyone knows the DX9 on the low FX series is a joke/marketing thing. And, again, I was using the ATI cards as a price comparison for low-end buyers.

I was trying to help the people who do buy low-end cards and can't figure out why they score less than a Geforce 2. But I'm just getting bashed.

[Sazar]

Originally posted by CPUMelting
DX9 doesn't matter to me, or this comparison. The other cards are in the same price range, and I said the margin was not small in 3DMark03.

Look at my post, I did use the latest 3DMark2001. Top GF4 Ti's are around 20K, and guess what, all use the "cheating" 40-44.03 drivers. If I used them, I'd probably get 12K or so, to be expected due to the chip differences. Everyone knows a GF4 TI can smoke most FX's in DX8.1, that's old news.

since 3dm2001 is so cpu dependent you should be able to get a decent score.. as it stands a score of 6-10 k and up is generally good enough to play most games out there today (w/o heavy ps2.0 usage) unless their coding is a little off :)

[CPUMelting]

Originally posted by Sazar
as it stands a score of 6-10 k and up is generally good enough to play most games out there today (w/o heavy ps2.0 usage) unless their coding is a little off :) I'd agree.

To the person who raved about the 5700, I did compare my '01 and '03 results to the 5700 non-ultra. Even if I could find it for $130, I'm not impressed. But then again, as I've said, I like my Xbox quite a bit (been playing PC games since the 8088, 4Mhz), so I wouldn't be that impressed as a total FPS person would.

[NickSpolec]

DX9 doesn't matter to me, or this comparison. The other cards are in the same price range, and I said the margin was not small in 3DMark03.

You don't understand.

The reason you get 2,000 points and the other cards get like 500-1,000 points is because they DO NOT run all the game tests. The Mother Nature test REQUIRES DX9, so that game test is not run on those other cards because, meaning they cannot get points for that test. Also, the Mother Nature test accounts for a lot of points.

So, it's not wonder the margin of difference isn't small.

It's very deceiving. Because in regular gaming, the 5200 sucks.

[CPUMelting]

Originally posted by NickSpolec
You don't understand.

It's very deceiving. Because in regular gaming, the 5200 sucks. BTW, I'm not a noob here. I'm MCSE, CCNA, A+, etc... I'm aware the DX8.1 cards omit most of '03s testing; old news... Okay, let's forget 3DMark2003 altogether and assume it "sucks" and is good for nothing (What were they thinking?).

Using 53.03 dets, I did manage to pull 9912 in 3DMark2001SE (without the 40-44.03 "cheating" dets) with an $85 GFX card that's:

1) Nvidia (I like the drivers & I even have the 3D glasses, ATI=good luck)
2) Digital Vibrance & Sharpness sliders (ATI=good luck)
3) Low-end DX9 support (ATI@$85=good luck)
4) VGA, TVout, DVIout

Any 3D card that's around 10K in '01SE doesn't "suck" for regular gaming. If you play at 1600x1200 on a computer monitor with 16x AA blah blah, etc, then the 5200 isn't for you, everyone knows this. As I've already said, I play Xbox games on a 51" HDTV on 5.1 home theater. :)

[ChrisRay]

So if something is good at 3dmark it must be good for gaming right ;)

[CPUMelting]

Originally posted by ChrisRay
So if something is good at 3dmark it must be good for gaming right ;) Mods, for the love of god, please lock this thread, it's nothing but flames. If you don't have better than a 5700 on here, you're garbage.

[Clay]

Originally posted by CPUMelting
Mods, for the love of god, please lock this thread, it's nothing but flames. If you don't have better than a 5700 on here, you're garbage. Settle down the bub, take a few deep breaths okay? These are not flames, they are opinions based on facts from some knowledgeable members. I see where you were coming from in your initial post but at the same time you kinda set yourself up for some flak. Sometimes people can be terse in their responses when they see the same misconceptions over and over again. Nothing personal about it and there was no flaming that I saw...you've never seen flaming if you think you were flamed in here just now. ;)

Also, spare everyone the acronym junk after your name/title. No one cares and it doesn't mean diddly-poo. MCSE, CCNA, A+, etc...all are a dime a dozen and I've known many that may have passed the requisite tests but are incompetent and I wouldn't pay them $5/hr to work with me on a project. Nothing personal towards you, mind you, just a little tip that trying to qualify yourself like that in most any forum won't do much for you.

So, just relax and move on. You started this thread with some good points and some not so good points, that's all.

:afro:

[CPUMelting]

Originally posted by MaxPower
Settle down the bub, take a few deep breaths okay? These are not flames, they are opinions based on facts from some knowledgeable members. I see where you were coming from in your initial post but at the same time you kinda set yourself up for some flak. Sometimes people can be terse in their responses when they see the same misconceptions over and over again. Nothing personal about it and there was no flaming that I saw...you've never seen flaming if you think you were flamed in here just now. ;)

Also, spare everyone the acronym junk after your name/title. No one cares and it doesn't mean diddly-poo. MCSE, CCNA, A+, etc...all are a dime a dozen and I've known many that may have passed the requisite tests but are incompetent and I wouldn't pay them $5/hr to work with me on a project. Nothing personal towards you, mind you, just a little tip that trying to qualify yourself like that in most any forum won't do much for you.

So, just relax and move on. You started this thread with some good points and some not so good points, that's all.

:afro: I used the word "flame" to reflect a rather big misunderstanding on my initial post, believe me, I have seen true flaming. I simply feel this thread is off topic.

Do you have a masters in EE?? Does it mean anything that I can reverse engineer cards and mainboards? Probably not "diddly-poo" right... $5/hour.... yikes. BTW, paper-MCSE's are a dime a dozen.

Let me ask you then, if you've worked on PC's for over 20 years, how would you qualify yourself as "not a noob"?

I merely intended to bring up several points at once in the initial post to start a lively/meaningful conversation. Take a good look at what follows, everyone saying I don't know the first thing about how '03 works yet I was programming computers before the PC came out.

And you're right about one thing, I did set myself up for flak, I had no idea the type of members in this forum (terse responses from experienced members seeing same misconceptions was how you put it; though I still fail to see any misconception in my post).

[Rytr]

Thread closed.

[Clay]

Originally posted by CPUMelting
Do you have a masters in EE??Nope, bachelors in CS from Purdue University and majoring in common sense for past six years. ;) Does it mean anything that I can reverse engineer cards and mainboards?Sure, that's cool but isn't directly related to gaming video card benchmarking.

At any rate, I guarantee you that there are guys here that can't even drive yet but know more than you and I do combined (or ever will). :p