The Fermi Thread - Part 3

[zalbard]

http://www.semiaccurate.com/2010/03/...tx480-benches/

Oh man, seriously... DX9?

I have seen the case with a HD 4850 Sapphire classic 2 slots cooler on a bad ventilated mid-ATX case with Furmark. Really massive heat

No chip is designed to operate at 100C+.
You're physically killing the silicon in this case. There is no way around that.

[kadozer]

It's OK because it's ATI, not Nvidia. It's also OK that eventually, the VRMs on that card will pop when running Furmark. Again, because it isn't nvidia.

It will be boring compared to Fermi. No mock card to make fun of. Will not be 6+ months late, wood screws. etc etc . Gonna miss the Twitter leaks Leaks will probably come 2 months before release something like this

http://www.pclaunches.com/graphic_ca...ecs_leaked.php

[eric66]

You guys are ignoring the fact that Charlie has been wrong about clocks , Shader count , Power Consumption and nearly every aspect of GTX 4xx series except for the launch dates !

i doubt even nvidia knew those things at that time lol

[Aerwidh]

You know what they say. Where there's smoke....

...someone's making hot dogs?

[96redformula]

You guys are ignoring the fact that Charlie has been wrong about clocks , Shader count , Power Consumption and nearly every aspect of GTX 4xx series except for the launch dates !

Was he really wrong at the time, or has nvidia changed it on several occasions?

[Hornet331]

I haven't heard of any cards hitting 120C though

Oh the VRMs on the HD4870 certainly did hit that temp, they even surpassed that.

And depinding on the cooling also the coor could reach ~75-80°C.

Charlies claims are as always just that: claims.

[Iconyu]

You guys are ignoring the fact that Charlie has been wrong about clocks , Shader count , Power Consumption and nearly every aspect of GTX 4xx series except for the launch dates !

Let's face it, they had not really been finalised until recently. They did another spin, and the reference cards they sent out had different clocks on them, so how could he get it right when nvidia weren't sure themselves?

As far as I'm concerned it doesn't matter how many shaders it has, what it's clocks are or its TDP (not power consumption), it's how it performs that matters.

[Macadamia]

Let's face it, they had not really been finalised until recently. They did another spin, and the reference cards they sent out had different clocks on them, so how could he get it right when nvidia weren't sure themselves?

As far as I'm concerned it doesn't matter how many shaders it has, what it's clocks are or its TDP (not power consumption), it's how it performs that matters.

Well, considering he basically registered on XS to say that...
Let's see if people like him or annihilat0r are still here after Fermi.

[Sly Fox]

Has anyone pointed out yet that ATI having had the 5800 series in market for ~8 months means they're more or less in refresh territory?

If the GTX 480 can just barely, ever so slightly edge out the 5870 (while using much more power)... Things aren't looking too good if ATI can just launch a 5890...

Not that they will for sure, but AFAIK 8 - 12 months is a pretty typical time for a refresh (at least since I've been following tech).

[H2O]

Has anyone pointed out yet that ATI having had the 5800 series in market for ~8 months means they're more or less in refresh territory?

If the GTX 480 can just barely, ever so slightly edge out the 5870 (while using much more power)... Things aren't looking too good if ATI can just launch a 5890...

Not that they will for sure, but AFAIK 8 - 12 months is a pretty typical time for a refresh (at least since I've been following tech).

We haven't seen any concrete evidence on a HD5890, but there has been talk of ATi releasing a new reference "OC" version, with a bump to the core and shader clocks.

[highoctane]

Has anyone pointed out yet that ATI having had the 5800 series in market for ~8 months means they're more or less in refresh territory?

If the GTX 480 can just barely, ever so slightly edge out the 5870 (while using much more power)... Things aren't looking too good if ATI can just launch a 5890...

Not that they will for sure, but AFAIK 8 - 12 months is a pretty typical time for a refresh (at least since I've been following tech).

Slow down gun powder, more like 6 months to the day, you might say you've added ~1 whole quarter to release schedule give or take a month...

[Vipeax]

Things aren't looking too good if ATI can just launch a 5890...

These kind of cards should have been planned for a long time ago.
Actually I had expected a 5870 2GB a loooong time ago (4870 512MB -> 1GB anyone? http://www.anandtech.com/video/showdoc.aspx?i=3415) with a refresh like the 4890 for when Fermi would come out to finish the line up and make ATi really shine for the first time ever since the release of the 8800GTX.

Sadly they did not go that route (now if you are wondering why I mention sadly: I am kinda stuck at the red side, thanks to Eyefinity).

[mindfury]

Actually I had expected a 5870 2GB a loooong time ago

Radeon HD 5870 Eyefinity 6 Edition 2GB will launch on 3/25.

[annihilat0r]

Well that card's TDP is something like 220W as far as I can remember, so the huge TDP advantage of 5870 (as compared to Fermi) vanishes there...

[Sn0wm@n]

cant wait to see what 32nm/28nm from amd brings on the gpu side

Well that card's TDP is something like 220W as far as I can remember, so the huge TDP advantage of 5870 (as compared to Fermi) vanishes there...

so the tdp now is magicly going down????

at first it was like.... but then it went to.... and then it was like.... starting to sound like an xzibit poster joke

[highoctane]

cant wait to see what 32nm/28nm from amd brings on the gpu side

Maybe you can start a thread about it...

[Sly Fox]

We haven't seen any concrete evidence on a HD5890, but there has been talk of ATi releasing a new reference "OC" version, with a bump to the core and shader clocks.

I suppose I would just hesitate to even call an OC version a "refresh."

Hence why I say 5890 (for lack of a more accurate name), think 4890.

It wasn't anything revolutionary, but it yielded some performance gains.

[Vipeax]

Radeon HD 5870 Eyefinity 6 Edition 2GB will launch on 3/25.

Yes, but that card will cost more over a 5870 than a 5870 2GB would, because of all the extra Eyefinity6 goodies that I don't want...


But this is waaay too offtopic.

[annihilat0r]

cant wait to see what 32nm/28nm from amd brings on the gpu side



so the tdp now is magicly going down????

at first it was like.... but then it went to.... and then it was like.... starting to sound like an xzibit poster joke

Man, calm down. You don't have to reply to everything that might be pro-Nvidia. TDP isn't going down, it's going up. I'm talking about 5870 here, its TDP is 188W but the E6 version is supposed to be 210 or 225W or something along those lines.

[mindfury]

Well that card's TDP is something like 220W as far as I can remember, so the huge TDP advantage of 5870 (as compared to Fermi) vanishes there...

Yes, previous boards are measured with a max of 3 simultaneous displays, this is done with 6.

http://forum.beyond3d.com/showpost.p...postcount=5700

[H2O]

I suppose I would just hesitate to even call an OC version a "refresh."

Hence why I say 5890 (for lack of a more accurate name), think 4890.

It wasn't anything revolutionary, but it yielded some performance gains.

It looks like the OC version is shaping up to be a HD5890, but there has been no mention of that particular name recently. ATi may keep the HD5890 name for a respin/tweak to combat the possibility of Nvidia doing another GF100 respin and releasing a 512SP part.

[spursindonesia]

All i know, reviewers like SKYMTL and Benchzowner have hinted that GF 100 cards will have AWESOME gaming performance compared to competitor, even on todays games, that GF 100 will stomp RV 870 to the thrash bin where it belongs, so i think i will put more weight on their cues rather than Charlie the satan himself for this launch occasion. We will see in the next few days, who will be flat out right, and the one who talked out of their rear.

I'm tired of believing Charlie only to be mocked up by nVidia fanbois, so this time i'll swear by their side, and see which side is the correct one after all of this unveiled.

[Sly Fox]

It looks like the OC version is shaping up to be a HD5890, but there has been no mention of that particular name recently. ATi may keep the HD5890 name for a respin/tweak to combat the possibility of Nvidia doing another GF100 respin and releasing a 512SP part.

Yea, it's hard to say precisely. My only real point was just that ATI would appear to be in a decent position to refresh if they have to.

Whether they will, what it's called, etc etc I have no clue about.

[annihilat0r]

http://forum.beyond3d.com/showpost.p...postcount=5700

uh, so the rise in TDP won't be that much when we still use one (or three) monitors? Nice.

[annihilat0r]

All i know, reviewers like SKYMTL and Benchzowner have hinted that GF 100 cards will have AWESOME gaming performance compared to competitor, even on todays games, that GF 100 will stomp RV 870 to the thrash bin where it belongs, so i think i will put more weight on their cues rather than Charlie the satan himself for this launch occasion. We will see in the next few days, who will be flat out right, and the one who talked out of their rear.

I'm tired of believing Charlie only to be mocked up by nVidia fanbois, so this time i'll swear by their side, and see which side is the correct one after all of this unveiled.

Why have I completely failed to get those hints? AFAIK SKYTML has been pretty tight lipped.