FX 5800 Ultra video.

[dguy6789]

A while ago, nvidia released a video talkin about how they messed up bad with the 5800 ultra. It is really funny, but i cant find it. If any of you guys know where i can get it, i would really appreciate it if you could give me a link.

 

[titanmiller]

I dont know, but I'd like to know.

 

[nRollo]

A while ago, nvidia released a video talkin about how they messed up bad with the 5800 ultra. It is really funny, but i cant find it. If any of you guys know where i can get it, i would really appreciate it if you could give me a link.

nVidia said the 5800 Ultra was a mistake, but there was no video with them talking about it that I remember.

 

[oldfart]

I have a copy of it. Send me a PM ifyou want it. I'll set it up on my ftp server (~5 Meg file size). It was pretty funny.

 

[kylebisme]

the makeing fun of the cooling system video eh?

 

[dguy6789]

yea thats the one

 

[VIAN]

I got a copy too.

 

[Evdawg]

kinda sorda off topic... but didnt the 5800 get better after new drivers?

 

[SneakyStuff]

yes, but who cares, it was already outdated by then

 

[modedepe]

Originally posted by: Evdawg
kinda sorda off topic... but didnt the 5800 get better after new drivers?

Yes. It actually performs fairly well, somewhere between the 5900 and the 5700.

 

[SneakyStuff]

the 5700u is actually QUITE similar to the 5800u, in a lot of ways, 128 bit, DDRII, architecture.

 

[kylebisme]

the 5700ultra is actualy fast than the 5800utra in many of the shadermark tests. i don't think the drivers have changed much for the 5800s, it's just that the publics focus has been swayed from its weak points. it still has poor ps2.0 shader performace and aa and af are a bigger performace hit on all of nvidias cards, those were the the big complaints about the 5800 when it came out and the issues are the same today.

 

[nRollo]

it still has poor ps2.0 shader performace

Wow, that's tragic. My .....oh yeah...I don't have any PS2.0 games because there's only a couple out that no one cares about....

the 5700u is actually QUITE similar to the 5800u, in a lot of ways, 128 bit, DDRII, architecture.

Of course it only has half the fillrate due to half the pipelines, so it's really not very similar at all...

yes, but who cares, it was already outdated by then

You should care Sneaky, because it's WAAAAAAAYYYYYY faster than your 5700Ultra. If you had a 5800Ultra, you'd actually have a pretty fast card.

Check out the "outdated" 5800U stomping feeble "modern" 5700U flat at DX9 Halo
Ouch, I'd rather be outdated than have a 5700U, Sneaky. I could actually play my games at some decent resolutions and AA/AF settings then.

 

[modedepe]

Originally posted by: SneakyStuff
the 5700u is actually QUITE similar to the 5800u, in a lot of ways, 128 bit, DDRII, architecture.

Well, yes they are similar in some respects. Both have 3 vertex shader units, so there they should perform roughly the same there. Pixel fillrate is also about the same, but because of the 4x2 design of the 5800u it has more textel fillrate. The 5800u also will have more bandwidth, courtesy of its 1ghz clockrate, so it should handle higher res and aa/af better. In a lot of ways it's a 5900 with less memory bandwidth.

 

[SneakyStuff]

the 5800 would have been a revolutionary card if they could have made it cheaper, quieter, and 256 bit memory bandwidth. The Radeon 9800 mopped the floor with it.

 

[MDE]

Originally posted by: Rollo

Check out the "outdated" 5800U stomping feeble "modern" 5700U flat at DX9 Halo

Ouch, I'd rather be outdated than have a 5700U, Sneaky. I could actually play my games at some decent resolutions and AA/AF settings then.

Mistake #1) Your link to THG was broken
#2) You're using a THG benchmark in an argument .

 

[kylebisme]

Originally posted by: Rollo

it still has poor ps2.0 shader performace

yes, but who cares, it was already outdated by then

You should care Sneaky, because it's WAAAAAAAYYYYYY faster than your 5700Ultra. If you had a 5800Ultra, you'd actually have a pretty fast card.

Check out the "outdated" 5800U stomping feeble "modern" 5700U flat at DX9 Halo

Ouch, I'd rather be outdated than have a 5700U, Sneaky. I could actually play my games at some decent resolutions and AA/AF settings then.

fixed the link for ya.

it would be nice if those numbers came with some screenshot comparisons so we can see if there is anything funky going on with floating point precision on the 5800 in order to get it to benchmark as well as it does. is it possable than nvidia would do something as dirty and underhanded as that?

 

[nRollo]

Monkey:

Mistake #1) Your link to THG was broken
#2) You're using a THG benchmark in an argument .

I fixed the link, and I'd use Anand's site if he would include the 5800U in some benchmarks against the feeble 5700U.

Sneaky:

the 5800 would have been a revolutionary card if they could have made it cheaper, quieter, and 256 bit memory bandwidth. The Radeon 9800 mopped the floor with it.

I can see why that matters to me as I had a 5800, and have a 9800Pro, but why is that important to you? We're talking about you slamming a card that anyone with any sense would rather own than your feeble 5700U.

Snowman:

it would be nice if those numbers came with some screenshot comparisons so we can see if there is anything funky going on with floating point precision on the 5800 in order to get it to benchmark as well as it does. is it possable than nvidia would do something as dirty and underhanded as that?

This isn't much of an argument Snowman. Unless you have some links to some review site saying the 5800s run at lower precision than 5700s, and why they think so, you argument is baseless speculation. It would be just as accurate for me to say, "ATI has been caught cheating before on benchmarks. Therefore, they probably cheated on Shader Day HL2 benchmarks".

 

[SneakyStuff]

Nothing to argue about here, the 5800 ultra barely beat out the ti4800, let alone a 9700 pro, I never said the 5700u was "better" than a 5800u, just cheaper, and quieter, they share MANY common traits. It is hardly "feeble" at least it does what they said it would

 

[nRollo]

Nothing to argue about here, the 5800 ultra barely beat out the ti4800, let alone a 9700 pro, I never said the 5700u was "better" than a 5800u, just cheaper, and quieter, they share MANY common traits

I'd stay out of the hardware review business if I were you Sneaky. "Barely beats a Ti4800"?! Uh huh. Can you post some links to this?
Do you know what a Ti4800 is? I think it's a 4600 with 8X AGP added. Let's see how a 5800U "barely" beats a Ti4600:

http://www.anandtech.com/video/showdoc.html?i=1821&p=24
Hmmm the 5800Ultra is more than twice as fast as the 4600 on all the UT2003 benches on that page.

oops, twice as fast at Splinter Cell too

Big surpise, demolishes the 4600 at Jedi Knight too

Twice as fast at Commanche 4

You know what problem I have with guys like you Sneaky? You post info that not even close to true, and someone that doesn't know any better might think you know what you're talking about and take your "advice".

I'll tell you what: I've got a $115 and a barely used Ti4200 (or $175) for anyone that wants to unload their "worthless" FX5800 U. Seems fair to me with the 5900NUs going for $180 used, or you could add a few bucks and get a new 5900 or 9800. PM me if you have one you want to sell, it would be worth this much money to me just to post some benches and play with this card.

 

[SneakyStuff]

Whatever, I could seriously care less. I hope the 400 bucks you spent on the 5800u was well worth it....

 

[DAPUNISHER]

Originally posted by: SneakyStuff
Whatever, I could seriously care less. I hope the 400 bucks you spent on the 5800u was well worth it....

I have a 5800U I bought slightly used *just a couple months old* paid the same amount for as the 5600U flip chip cost at the time, and the used 9700pros for that price where nigh impossible to get because someone else always got there first. I've been using this card a good while now and it's just gotten better with age, like a fine :wine: Here's what it'll do with the UT2K3 DEMO bench- A64@2.3ghz 5800U@525/1080 DET. 53.03 1024x768x32 flyby 284 BotMatch 105 not too shaby I think. Combined with the A64 it plays Halo, Far Cry, Painkiller, etc. very smoothly even with AA/AF on in some of the games, I usually use 4xAA/AF but use higher too it just depends on the game, my LCD's native res is 10x7 so that is the res I tend to stick with.

The only negative of this card for the price I paid is that it is comparable to a 60mm Delta black label screamer in 3D mode, I wear Sennheisers when gaimng though so it's a non-issue for me Since the noise doesn't bother me it actually gives an advantage in that it blows hot air out of my system instead of into my case I can hit 549/1.13ghz stable thanks to the massive&loud cooling but with the AGP running out of spec on my 8KUSA I'm limited to 525/1080, I have a Shuttle AN50R with locked AGP coming though so I should be able to get a great overclock and max the card's overclock too=even better performance, I'm hoping.

My point is that I wouldn't have paid anywhere near the $400 they wanted for it new either, but for half priced used it was a good value and I've yet to have it do anything but perform well so please stop bashing my baby! BTW, Rollo bought the 5800 which was less expensive and overclocked to stock ultra speeds, so he didn't pay $400 either.

 

[kylebisme]

Originally posted by: Rollo
Snowman:

it would be nice if those numbers came with some screenshot comparisons so we can see if there is anything funky going on with floating point precision on the 5800 in order to get it to benchmark as well as it does. is it possable than nvidia would do something as dirty and underhanded as that?

This isn't much of an argument Snowman. Unless you have some links to some review site saying the 5800s run at lower precision than 5700s, and why they think so, you argument is baseless speculation. It would be just as accurate for me to say, "ATI has been caught cheating before on benchmarks. Therefore, they probably cheated on Shader Day HL2 benchmarks".

na the difference is nvidia is currently cheating on benchmarks and they claim they have no intention of stoping. furthermore, tests have shown that the 5800 falls behind the 5700 in many floating point shader operations. i know you don't want to hear it, but running the ps2.0 shaders at fx12 is the most likely explaination as to why the 5800 preforms so strongly in halo.

 

[nRollo]

Whatever, I could seriously care less. I hope the 400 bucks you spent on the 5800u was well worth it....

LOL, I point out all the incredibly off base information in Sneaky's post, so he says, "Whatever, I don't care" like an urban youth on a day time talk show.

tests have shown that the 5800 falls behind the 5700 in many floating point shader operations. i know you don't want to hear it, but running the ps2.0 shaders at fx12

LOL
1. Could you post links to these tests? I don't doubt this could be the case, BUT....
2. Which games are you playing that use PS2.0? ......................................................................(crickets chirping)..............................................Oh I see. So I should care about PS2 on a video card about as much as say, a retractable spoon. Nuts, wait a minute. Soup and ice cream exist, so I guess a spoon would have some use.

 

[kylebisme]

just about any information on the 5800 is hard to find at this point, even on nvidia's site. however, you can see some shadermark benchs as well as a discussion about the 5800s performance here. as for your question about what ps2.0 games i play, that is compleatly irrelivent to the discussion. but sense you asked, prety much just halo at this point although i did get the new tomb raider free with my videocard and have been meaning to play though it eventualy.