Slashdot Mirror
ATi Radeon 9800 Pro
ATi is bringing out their new card, the Radeon 9800 Pro, and all of the hardware review sites which depend on ATi's generosity for pre-release hardware have released their necessarily favorable reviews. Here's a few: Hothardware.com, Hexus.net, HardOCP.com, Anandtech, Tom's Hardware, Extremetech, PCWorld.
I hope to bring to the attention of ATI developers, if they are reading, that it would be nice to release official driver support for the R200 models (Radeon, Radeon 7500 etc) and only the latest 8000+ models.
These cards are partly supported by the DRI project on dri.sourceforge.net since they lack important features as texture compression making them useless for games as DoomIII.
Thanks.
ps. Or at least, please help the DRI guys complete the great job.
So which link did you click first and why? Personally, I went for Tom's Hardware.
Just because your a Nvidiot, doesn't mean you can go bash ATI for making a better card or the websites that review them.
I know I speak for everyone here when I say "michael, JUST SHUT UP!!!!!"
seriously though - was it like last week 9700PRO became available? what's up with this break-neck card-releasing? I didn't think it was christmas yet...
My life in the land of the rising sun.
Keep it up, ATi. Competition is good. I'm really lovin' what I see in the 9X00 series. Keep hammering on improving those Linux drivers while you're at it, because nVidia still has the edge on non-Windows platforms. The day that you release Linux drivers that are on par with those under Windows (as PowerVR and nVidia have done) is the day that I fork out $400 for your car. Rest assured that I will, as long as you back the product.
Quoting: ATI will call the extended set of DX9 features the DX9++, although we suppose it could add just as many ++++++ as it wanted to. ... ... Nvidia should perhaps call its own DX9 extensions DX9## or DX9.NET.
the sad thing is, though - I would not be surprised if Nvidia did release a DX9# or something stupid like that. I mean, look at Athlons naming themselves AthlonXP. ack
My life in the land of the rising sun.
... but I have yet to see ~realtime rendering of my 3D Max projects. When are these supposed advances in video cards going to trickle down to the hardcore 3D artist?...
The GeForce FX has some horrible Pixel shader performance using ShaderMark v1.7 as shown by HardOCP:
"In ShaderMark the GeForceFX pretty much terrible when it comes to pixel shader 2.0 performance compared to the 9700Pro and 9800Pro. Performance of the GeForceFX is horrible compared to what these cards are showing us. The 9800 Pro improves up to 50 FPS in some cases compared to the 9700 Pro. There is no doubt that the 9700 Pro and 9800 Pro have very strong pixel shader speed.
This benchmark also does give some credence to the 3DMark03 PS2.0 numbers.[my bold face] More PS2.0 coming next week that will really get you asking questions."
I wish hardware sites would talk about more interesting things: serial ATA, 10Gbps ethernet (yes you heard me right... that's what's next...), giant LCD screens (or plasma), 7.1 channel sound, not a graphics card that gives me a 3% edge on directX 9.0 games of which there AREN'T ANY. Okay, rant over :)
Karma: pi (Mostly due to circular reasoning in posts).
1. So can one truly notice the difference between say 45fps and 100fps?
2. How many games will be out within the next six months to take advantage of this cutting edge technology?
I understand this is the business practice of these times. To always wait about 6-8 months before hyping up the next release of something. Why so many changes to squeeze more fps? Is it like trying to add 10 more HP to your Honda? How many people on this place can actually look at a screen shot or video and name what type of graphics card is being used and what options are set like AA and such?
This SIG pulled due to lack of funding. (This damn war is costing too much!)
There must be a very interesting formula at work for early release reviews. The product suppliers want good press and a wide audience. The reviewers want a larger audience for their web site, and possibly fame or a chance to try out the next big thing --first! The readers want interesting, informative reviews they can believe, use for purchases, and quote with authority. These forces pull early release reviews to a common middle. The product suppliers won't provide their product to a site that reports credible, but consistently unfavorable reviews. Readers won't keep reading reviews that are favorable, but consistently boring, unhelpful, or not credible --then the product supplier drops the review site for lack of audience, anyway. So, the review sites that get the chance to review new products are the ones that produce consistently interesting, informative, and favorable (or at least, not UNfavorable) reviews.
Of course, confounding this formula is PT Barnum's line "I don't care what they write about me as long as they spell my name right." Some suppliers may continue to release early products to unfavorable, but popular reviewers, just to increase the overall level of press coverage. Worse yet, since the early product is provided by the product supplier, it may have been specifically modified from the "retail" version to work better on benchmarks, just for the review. For that matter, the reviewer may be tempted to soften a review for the sake of a site advertiser's new product.
Still, what's a consumer to do? I guess we have to take early reviews with a dose of skepticism. Before we make a purchasing decision, we have to wait for a reviewer to buy an off the shelf unit and test it. That's the best way we can be sure the review is more in our interests than the product supplier's.
Sometimes I worry that I'll develop Alzheimer's disease, but no one will notice.
Here's the text I'm refering to below.
----
Pipelines, pipelines... February 25, 2003
Hello, world.
Just wanted to write a word or two regarding the issue raised couple of days ago. Seems like the whole Internet community wants to crucify nVidia about the controversy of how many rendering pipelines GeForceFX realy has. Is it 8 pipelines with 1 texture unit, or 4 with 2, or ... uh... I don't know anymore. And it really DOESN'T matter that much!
The only thing that matters is how fast and how good it can render pixels. And both GeForceFX and Radeon9700 are great products, the kind of hardware that developers long for. So, personally, I don't care much what's "under the hood".
Don't get me wrong, I am into 3D-graphic hardware, but this pipeline thing really went out of proportion. Number of pipelines is a good hardware information, and that's all there's to it. It really doesn't need to reflect the speed of the hardware directly. Come to think of it... currently, there are no games that utilize even 1/3rd of nifty features these two boards have.
Oh, before I forget... I'm not "nVidiot" (and I'm not "fanATIc", either). I'm just a game developer who wants good and fast technology for the future. And both ATI and nVidia have it now!
Just my two cents.
Dean "3D" Sekulic
(Programmer)
P.S. Yes, I snapped.
"To confine our attention to terrestrial matters would be to limit the human spirit." -Stephen Hawking
> What they need is a standards based API for graphics engines.
Already done -- middleware such as RenderWare, Net Immerse, etc., already provide this, and are starting to be used more and more.
The reason there hasn't been a standard for graphic engines, is because the problem is an old one -- flexibility (abstraction) vs performance (hard-coded). Game engines that are flexible used to suffer a HUGE frame rate hit, which is completely unacceptable on consoles, where they needed 30 fps minimum.
e.g.
BSP Trees vs Sphere Tree. A BSP Tree needs to be processed off-line (meant for static data, not dynamic), but gives perfect sorting, in linear time. Sphere Trees can handle dynamic objects just fine, but can't be used for sorting.
As CPUs have been become faster, and the graphics work has been offloaded to a dedicated GPU, the CPU has more time for the "general" solution, that is "fast enough."
Cheers
I remember reading a review of dual cpu motherboards on Tom's Hardware Guide. Everything scored within a few percent of the others, but he kept going on about how the Asus board was cleary faster than the others. There would be a graph of, for example, encoding mp3s, and all of the motherboard were withing in afew seconds of each other, and yet this guy is raving about one board and comdemning another. I stopped taking him seriously at that point.
Of course, you also get things like the infamous NT vs Linux benchmark: Let's test a rollout of NT specially calibrated and tuned for us by the NT development team, versus some old Linux distro we found in a trashcan.
Whenever you see one of these reviews, just ask yourself: where's the money to do this coming from? That'll tell you whether you should believe it or not.
dave
Please implement a VGA BIOS disable switch on your videocards. Some of us are working on computer platforms that can't work with your VGA BIOS, yet their exists graphics drivers that CAN use your proprietary graphics-acceleration architecture chipset on your related products.
For example, disabling the VGA BIOS would allow users of Alpha/Sparc/MIPS/PPC/Power(3/4) platforms to use a wee-little standard VGA graphics card that we know works (like a S3, Permedia2, G200, or RagePro), then throw a hefty ATI Radeon 9800+ Pro XPERTONIA ++plutonia++ 256MB or nVidia GeForce FX 6000++BrownOut/cooker 256MB L24a adaptor into the AGP port or hopefully see a 64bit PCI model from ATI/nVidia and we could use your hardware!
Sincerily,
The Alpha Troll
But I'm sure you already Gnu that.
Is ATI going to continue to sell RV300 based boards? And if so at what price points? I _just_ (last weekend) bought a 9700 PRO at Circuit City on sale for $299. I realize now that it was to just get rid of it (Best Buy also is listing theirs for $299 presumably for the same reason.)
The 9800 is only marginally better than the 9700, and the 9700 is far far better than the new 9600. The new 9600 is supposed to be $219 and the new 9800 replaces the 9700 at $399. That leaves a big gap.
What I'm worried about is if ATI is going to continue producing 9700's, will they be under $300? Anything less than $299 and I'll feel ripped off. (Unless I can get a price adjustment from CC.)
Still, I got a good deal I suppose. I never would have spent $399, and if they stop making 9700's then I paid a fair price for it too.
--
My comments and opinions completely reflect those of anyone and anything I am remotely associated with.