Slashdot Mirror
Rumors of a GeForceFX 5800 Ultra Cancelation?
chris_oat writes "It seems that nVidia's GeForceFX 5800 Ultra may never see the light of day after months of super-hype and annoying delays. This article on megarad.com suggests that poor manufacturing yields are causing nVidia to rethink plans for its (new?) flagship part. Lack of an "Ultra" type solution from nVidia would leave ATI's Radeon9700 uncontested as the defacto performance part."
I see lots of comments predicting doom and gloom for nVidia already. The GFFX has been somewhat of a disappointment, both for consumers and for NV - it's too slow, too hot, and too hard to make. nVidia is not going to go into bankruptcy because of this however - they will still sell a few and will work madly on the next generation aimed for smaller design rules and will learn from their mistakes this time around. The GFFX isn't the death knell for the company, it's just an unpleasant reminder of what minor manufacturing difficulties can do in a nasty business like video card manufacture. They're already hard at work on the next-gen part, and I'm sure they've learned a lot with this one.
Meanwhile ATI will enjoy higher profits and will have a bit of breathing room. Hopefully, they will use this time to extend their product offerings viz the R350 core, continue pouring money into driver development, and keep working on R400 or whatever their next-gen core ends up being called. In any event 6-9 months from now we will see these next-generation parts coming to market, and they will be just that much better.
High-speed Road Trip (18.000KPH)
Disclaimer - I'm not a gamer, in any way, shape or form...and could care less about 3D acceleration
However, I have been personally predicting the fall of nVidia for quite a while now, ever since my Diamond Viper 770 Ultra died, to be honest. I replaced it with a used Asus TNT2 Ultra Card (V3800 I think), and had the same kind of issues, which were mostly the system locking up due to heat buildup.
Then I built a machine for a friend of mine, and we put a Hercules GF2 MX440 card in it, with the same kind of issues. Not only that, but the power requirements were what I was considering to be obscene, for the time.
When we replaced that MX card with a Radeon VIVO, all of the stability issues went away, same thing when I replaced my Asus Card with a Matrox Millenium G550. All of a sudden XP Pro was...well..."Stable"
Then, six months ago I finally got so completely sick of The stability issues that I was living with, mostly due to a rather ecleptic mix of hardware, that I switched XP out for Debian. Found that not only did all of my stability issues disappear (which I knew would happen) but, I had better driver support for my Matrox card than I did in windows.
This got kind of ranty, and I forgot where I was going with this...oh well...
nVidia's plans for the FX were greater than what actually happened. If this had been released with support for 256 bit memory, I think it would've stomped ATi big time.
Sometimes it takes a brilliant failure like this to catapult R&D to the next level. Let's hope that happens here.
"...Well, there's egg and bacon; egg sausage and bacon; egg and spam; egg bacon and spam; egg bacon sausage and spam..."
[H]ardOCP has confirmation that GeforceFX5800 Ultra graphics cards will not make it to retail, and are available as pre-order items only, for a limited time. However, the GeforceFX5800 non-ultra model *will* make it to retail, sans the elaborate cooling mechanism, and running at 400MHz GPU / 800MHz RAM.
Additionally, it seems the "Radeon9900" information at Xbitlabs might be less accurate than it appears.
This isn't the greatest news for Nvidia, but it doesn't exactly break the bank: Nvidia still has the lion's share of the graphics market, and will probably continue to keep that market simply due to Tier 1/2 OEM sales, as well as their reputation - even though ATI has faster hardware, Nvidia has had a history of rock-solid drivers 4 generations back. Although ATI's driver quality has improved significantly in recent times, they're still not up to par with Nvidia's. And be sure that Nvidia will capitalize on that, since they don't have bragging rights for their hardware currently.
------- "From bored to fanboy in 3.8 asian girls" ----------
It's said that you stop seeing the difference at around 30 fps
:-) And then there's FSAA to remove the pixelation even more.
And 16 million colours is more than the eye can see, and 44,100 samples per second is more than the ear can hear. Throughout the march of technology we've heard these ridiculously arbitrary "limits" of our senses, and invariably they are discounted at a future time. In essence you can consider them a sort of justification.
but so far I've only heard comparisons with movies and TV
Actually I've been paying attention at movies having heard the "well movies are 24fps and they look perfect": MOVIES LOOK LIKE TRASH. Seriously the next time you go to the movies pay close attention to any large movements on the screen and you'll be surprized how horrendous 24fps really is. For instance, my wife recently dragged me to see "Two Weeks Notice" and there is a scene where the camera pans laterally across a shelf full of shoes at a rate of about a screen width per 1/2 second-- It looks absolutely atrocious. For fast action most filmmakers either resort to the action taking a small portion of the screen, or they use slow motion effects, again because the action simply looks terrible at 24fps.
However when you get down to it the root of the "X FPS is more than anyone can see" is people's astoundingly self-centered claims that no-one else can see more than 30fps, or some other metric. This can be disproved instantly via the Q3 command cg_maxfps. Set it to 30 and it looks like a horrendous slideshow. Set it to 45 and it looks like a 1998 computer. Set it to 80 and it feels smooth with a bit of jaggedness. Set it to 90 and it feels nice. You'd think this would disprove the 30fps'ers in an instant, but amazingly they persist.
and then complain that a gfx card sucks since it doesn't perform good enough in 1600x1200. It's not like you have enough time to spot the microscopic pixels anyway.
1600x1200 on a 19" monitor is hardly "microscopic" pixels, however to consider this in a forward thinking manner consider the heavyweight video-card required to do 1080p resolutions on a HDTV set? 1920x1024.
FSAA, BTW, is tremendously difficult for video-cards to do (because they're actually rendering at 2x or greater resolutions): There is no current video card that could dream of doing even Urban Terror (a Q3 mod) at 1600x1200 with FSAA at acceptable frame-rates.
NVidia's future looks quite uncertain. It appears they might be headed for a free fall. You can blame their problems with some bad business decisions, like backing AMD, but the real problem is that ATI's tech team is pulling ahead. The 9700 simply had a better designed core. Their position is remarkably similar to that of 3dfx during the introduction of the tnt2. The handwriting was on the wall, and there was nothing they could do about it, having sunk millions into technology consumers were just not interested in.
My most recent video card purchase, an ATI OEM 8500LE just died a little over a year after I bought it. Add too that the fact that the Mobility M4 in my $3K laptop still doesn't do all the tricks ATI promised. Not very impressive considering my TNT2 has been chugging away for years.
"God fights on the side with the best artillery." - Napoleon, Marshal of France - speaking truth to power
Everyone and their brother seems to be bellyaching about "losing the PCI slot" to the cooling system, and how horrible this is.
How stupid.
On practically every motherboard out there today, PCI 1 and the AGP slot share resources, so you're crippling your system performance by putting a card in each.
As I remember it, PCI has four specific special IRQ channels allocated for it, and thus the original spec is for one IRQ for each. Modern motherboards get away with this by having different slots share the bus mastering, so that two devices can piggyback on one slot. Usually, the onboard IDE controller piggybacks on one slot, and the last two slots (usually PCI 5 and 6) are often coupled together. By the same token, the AGP slot often shares an IRQ with PCI 1.
So, in short, if you're going to complain about the cooling system, complain about it being loud. You weren't losing anything on your motherboard that you could even use to begin with.
nVidia are a larger company with a string of huge successes to date. They have a much more diversified income, including some very popular OEM chips, the successful nForce2 (and less-successful Xbox) chipsets, a well-regarded pro card line, and a significant share of the Apple market too. Not to mention quite a bit of cash in the bank.
A single high-end chip(which is a small % of their total revenue anyway), even if it failed completely, is not going to impact their bottom line that much. It'll have more impact on their image as graphics leader, but they have the resources to learn, move on, redesign and try again.
Why would anyone engrave "Elbereth"?
Just a simple question: What is the refresh rate of your manitor? If you can do 1600X1200 at more than 85Hz and notice the improvement, I'll be impressed. The difference you report between 80 and 90 fps almost certainly has to do with the fact that your monitor refresh was set so that you show exactly five "dead" screens per second. When you set it above 85, each of your screen refreshes shows a different frame, so of course it looks better, but it's not for the reason you think. Try running at constant 75 FPS and set your refresh to 75, and you will see it also looks great.
No. This is bogus. The Kryo2 chip, along with the GF1/2, had a form of FSAA where it basically rendered the image at a much higher res than it was going to be displayed, but these don't exist anymore (at least not in the R300/NV30/NV25/R200). I BELIEVE this is called supersampling--don't quote me on this, I'm not a coder and don't care too much about FSAA modes (I have a GF3. I can't use AA in ANYTHING but the oldest games.). Supersampling takes a much larger performance hit, but a lot of people regard it as looking better than the newer method. This newer method is called multisampling--it actually renders the image multiple times, offsetting it each time. This is why color compression has become so important. 4x MSAA COULD take up to four times the memory bandwidth of normal rendering, but with adequate color compression, you could get it down to two times or 1.5 times the bandwidth. This is part of the reason why nVidia went with a 128-bit bus on the GFFX--it thought it had good enough color compression.
Anyway, moving right along, there are two forms of MSAA (multisampling antialiasing)--ordered grid and rotated grid (once again, do not quote me on this).
So basically, FSAA ain't as simple as rendering at 3200x2400 and reducing that to 1280x960 anymore.
---
nV News
Back a few years ago, ATI had the lead in first sampling with the Rage 128 over the Nvidia TNT. ATI had an amazing launch in Toronto and blew away the press. Nvidia was quaking in its boots.
.25 micron technology (along with some interesting - read poor - design decisions) caused ATI a delay of 6 months before it truly got any production quantities out - and these were chips that weren't performing at expected clock rates.
However, the change to
This is when Nvidia seized the lead and never looked back - until now. They killed ATI with their rock solid 6 month new product launch cycles.
Now ATI has the chance to be in front for the foreseeable future.
Both Nvidia and ATI have great teams of people and this battle of the champions benefits everyone interested in graphics.