• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Ati ~ Amd?

Also, ATI is heading in new directions. AMD stopped playing the gigahertz game with Intel and instead went to creating an onboard memory controller and more efficient design. That's just what ATI did here. Which is why, with less pixel pipelines, it can keep pace with the best nVidia has to offer. Some people may point out that ATI's cards are clocked much higher than nVidia's. Well, that is just a tribute to the efficiency and capacity of ATI's new memory controller, as well as their speed-binning process.
Click to expand...
Der, did you read the thread where it showed a benchmark of the 7800 gtx and the x1800 xt @ 16 pipes 450 mhz. The gtx stomped the xt in every benchmark. Including half life 2. Don't see how that's efficient.
 
delusion_2005 said:
yup ditto, i think ATI is more like Intel since they are ramping up the clock speed while Nvidia has went the other way with more pipelines but lower clock speed and they perform the same, just like a 3200+ performs almost the same or better then a 3.2ghz p4
Click to expand...
Video card architecture is an entirely different matter. "Oh, look, to increase performance, ATi has simply increased clockspeeds!" What about - "Oh, look, to increase performance, nVidia has simply added more dedicated shaders!"
It's an entirely moot point to compare how the work is done. Power consumption and heat output-wise, the X1800 XL/XT 256MB are incredibly close to the G70 equivalents (7800 GT/GTX). Don't say "But the X1800XL core is shown to reach up to 93C!" because this has already been shown as a calibration error in Catalyst Control Centre, which reports temperatures at least 20C too high at load.
Also, don't say "The X1800XT sucks down far more power than the 7800GTX!" because, while this is true, it is not due to the core, but the extra amount of RAM on the PCB that is running at exceptionally high speeds.
So, with all other relative variables being equal, i.e. heat output, power consumption, that brings us back to the performance of each core. Who cares if ATi manages the frames by being a bandwidth whore? It's effective, is it not? Who cares if nVidia jacks up the 3D Marks by being a wide-piped bitch? This is also effective, is it not?
The thing that people should worry about is the end performance, and as a close second, the architectural advancements and innovations (or "minor upgrades" as PRIME1 chose to label them).
To that effect, clap clap, nVidia, nice job on the G70. It's got some great tricks up it's sleeve, and does an excellent job within todays games. I applaud you for taking the NV40 core and making what works even better, and getting it out the door with excellent timing.
To ATi - Too bad on the late start. It was good of you to take the time, but being late to party means you miss the punch. However, congratulations on doing something truly different. Not only are you able to keep pace in today's games, but there is indications already that the powerful programmability of the R520's core will enable you to keep current customers future happy as well.
Honestly....how many times can it be said yesterday's architecture can keep a reasonable pace in tomorrow's games?
 
Intel_Hydralisk said:
Der, did you read the thread where it showed a benchmark of the 7800 gtx and the x1800 xt @ 16 pipes 450 mhz. The gtx stomped the xt in every benchmark. Including half life 2. Don't see how that's efficient.
Click to expand...
That article was what I often call, in the most objective manner possible, an utter load of shit.
Not only have both cores been crippled backwards from their optimal performance, the so-called "reviewer" seems to have conveniently forgotten that the 7800GTX has a 40Mhz delta-clock associated with the ROPs, and a 20Mhz delta clock associated with the vertex shaders, which is more than likely where the extra performance has come from. To achieve the fairness that the article professed, the X1800XT would have to be clocked 40Mhz higher on it's ROPs, and 20Mhz higher on it's vertex shaders.
This is impossible to achieve, so some other form of comparison is needed to determine which architecture is more efficient (a pointless quest for an altogether useless answer, anyway, the only use of which is for advocates to spurt inanely in rampaging forum wars with no idea of what they're really claiming).
 
Intel_Hydralisk said:
Der, did you read the thread where it showed a benchmark of the 7800 gtx and the x1800 xt @ 16 pipes 450 mhz. The gtx stomped the xt in every benchmark. Including half life 2. Don't see how that's efficient.
Click to expand...
The architectures are different. Kyle, Brent, and most of us agree that we should only really care about what IQ and performance it delivers in actual games. As others have pointed out, much of the X1800XT's extra power consumption is due to the 512MB of RAM, which reviews have shown the X1800XT can take advantage of (unlike the 512MB X800XL). With all the rumors running about a 512MB GTX/Ultra, NV probably agrees that it is time to follow suit with a 512MB part, and its power consumption will be higher than a 256MB GTX - even if they don't change anything on the GPU.
 
DougLite said:
much of the X1800XT's extra power consumption is due to the 512MB of RAM,
Click to expand...

I personally don't know how much extra power consumption another 256mb's of vRAM will do but the X1800's have more transistors than their nvidia halfs equating to more power consumption...
 
Terra said:
And their GPU's also vary in design, but the final output is very simmilar in 3D.
So is the output of AVIVO and Purevideo enhancements on ATI's X1x00 series and the NVIDIA 6x00 and 7x00 series 3D-cards...

Terra - But Purevieo was introduced over a year ago, AVIVO has just been released ;)
Click to expand...

AVIVIO is free though. zing!
 
Netrat33 said:
Wow...suprised someone had to point it out to you.

rolly eyes back towards ya. :D
Click to expand...
Once again Purevideo is a programmable hardware video accelerator. It's included on all NVIDIA GPU's from the 6xxx series on up.

What you are paying for is a MPEG2 video codec that works with it. This is generally used for DVD playback. Many video card manufactures bundle this application with their cards.

Windows Media Player 10 has built in support for Purevideo for accelerating HD content.

Most people have to buy dvd player software for their computer if it did not come with it.
 
Bona Fide said:
This thought just struck me while reading one of the [many] benchmarks that compare the X1K series to nVidia's 7 series.

ATI is reminding me more and more of AMD with each passing day. First off, the obvious. They are the underdog. No question about it. nVidia, like Intel, holds a considerable share of the graphics market. Not to mention they are, on the whole, a much LARGER company than ATI.

Also, ATI is heading in new directions. AMD stopped playing the gigahertz game with Intel and instead went to creating an onboard memory controller and more efficient design. That's just what ATI did here. Which is why, with less pixel pipelines, it can keep pace with the best nVidia has to offer. Some people may point out that ATI's cards are clocked much higher than nVidia's. Well, that is just a tribute to the efficiency and capacity of ATI's new memory controller, as well as their speed-binning process.

Does anyone else see this connection? Or am I just crazy...

I think I want to trade in my 7800GT for an X1800 XL. Think it'll happen? :D
Click to expand...

I thought nVidia was the one that was similar to AMD. Like you said, both of them stopped playing the megahertz game and now going for more work per clock than simply raw speed.

And ATI's been around long before nVidia and with an even bigger marketshare. ATI only started taking 3D gaming seriously when nVidia appeared and started eating through their OEM market. If anything, ATI's the one with the advantage time and resource wise.

thore said:
do we not remeber how badly nvidia sucked during the nforce 1 days... it was slow.. buggy and generaly a bad move, this is ati's first try at a chipset, its hard to get it right on the first try (hell pakard bell never got it right) not saying be kind and by there product , what im saying is that the next gen (if they release one) will have the lessons learned, and (most of) the bugs worked out.

thore
Click to expand...

Huh? ATI's been making chipsets for years. They actually had more to work with than nVidia.

delusion_2005 said:
yup ditto, i think ATI is more like Intel since they are ramping up the clock speed while Nvidia has went the other way with more pipelines but lower clock speed and they perform the same, just like a 3200+ performs almost the same or better then a 3.2ghz p4
Click to expand...

AMD chips are also cheaper than Intels. Any chance nVidia will get cheaper as well? :D

banGerprawN said:
Video card architecture is an entirely different matter. "Oh, look, to increase performance, ATi has simply increased clockspeeds!" What about - "Oh, look, to increase performance, nVidia has simply added more dedicated shaders!"
It's an entirely moot point to compare how the work is done.
Click to expand...

What has basically defined AMD is that their chips do more work per clock than their competitors. That's what everyone's relating to, and why nVidia's being compared to them.
 
delusion_2005 said:
yup ditto, i think ATI is more like Intel since they are ramping up the clock speed while Nvidia has went the other way with more pipelines but lower clock speed and they perform the same, just like a 3200+ performs almost the same or better then a 3.2ghz p4
Click to expand...

Um... you know, that wasn't their strategy. They were going to have 32 pipes, but they had to cut it to 16 due to yield issues. Pipe-lines are not equivalent to the length of the CPU's instruction pipe. MHz for MHz both company’s clocks do about the same amount of work, when all the tertiary stuff is the same. I give ATI kudos for actually increasing the core speeds, and not pussyfooting around with a bunch of features. Pipes will come later. They are not part of the core architecture.


Intel doesn’t count because they make next to nothing on integrated graphics. Nvida and ATI make much, much more money on video cards than Intel.
 
PRIME1 said:
Only the WMV support was broken
Click to expand...

So that makes it ok???

I'm sorry but you wait 8 months (count 'em) for the decoder to come out and then the NV40 and NV45 has broken WM9 Decoder Acceleration..
 
tornadotsunamilife said:
So that makes it ok???

I'm sorry but you wait 8 months (count 'em) for the decoder to come out and then the NV40 and NV45 has broken WM9 Decoder Acceleration..
Click to expand...
ATI did not even have one, what do you care?
 
swatX said:
innovative? care to explain ?
Click to expand...

The memory architecture for starters. The boost Q4/Open GL got from a driver update
due to specific optimizations blew me away. :)
 
Why bother drawing analogies if they are going to be so blantantly innacurate?

Just look at the facts.
 
PRIME1 said:
ATI did not even have one, what do you care?
Click to expand...

ATI has had one since they named their cards "radeon"... they just didn't name it...

The memory architecture for starters. The boost Q4/Open GL got from a driver update
Click to expand...

that's without even using application specific optimisations, just a memory controller bug fix... i'm personally waiting for ATI to release an official driver, since you all know that there are none for x1k series, even the latest betas from the ATI support site don't work, you need to go with the drivers in the retail package for the X1800XLs to work...
 
LyCoS said:
even the latest betas from the ATI support site don't work, you need to go with the drivers in the retail package for the X1800XLs to work...
Click to expand...

O RLY? I was under the impression that the now fixed 5.10a's (or 5.10b's if you want) were functional...
 
PRIME1 said:
Not really, and not even near the scale of Intel vs. ATI


You mean like SM3.0, HDR, SLI....those are not very new.


Actually ATI is playing the megahertz game by using fewer pipes and a much higher clock speed.
Click to expand...
Be glad that the filter caught the first two words of the sentence that was here. - DL

I'm pretty sure hes talking about ATis headroom for optimation on there drivers. I lean towards nvidia for graphics card but i'm more than sure ill lean to ATi once they get a few more drivers out for there new generation.
 
You must log in or register to reply here.
Back
Top