Idiotic review

[Typedef Enum]

http://www.vr-zone.com/reviews/NVIDIA/FX5800U/page12.htm

However NV30 is meant to compete with R300 not R350 so it would be more fair to compete NV35 with R350 and not NV30 vs R350 so hopefully NVIDIA will not delay NV35 beyond May if not it will spell further troubles ahead for the company

Is this guy on the payroll or what?

Not only does his crap review lack virtually any decent AA/AF metrics whatsoever, he then updates the thing with this utter crapola.

Let me see if I understand this correctly...You actually think that because nVidia doesn't seem to be able to execute a plan anymore, that NV30 should not be compared with the 9800 because it was _meant_ to be released last year?

OK...I just got back from the local Frye's, and they did receive a shipment of 9800's this weekend. Meanwhile, you cannot so much as find _any_ NV3x product on any retail shelf...AND this guy wants to compare the 9800's to NV35's? What crack is this guy smoking?

In addition, I also found this quote interesting.

However, the fact that NVIDIA was delaying NV30 constantly till they almost missed one cycle wasn't entirely due to the difficult GPU design but other contributing factors as well such as low yield rates from the TSMC's 0.13 micron process as well as stringent PCB requirements to support NV30 and GDDR2 proved more challenging than initially thought.

Again, I don't know where this guy has been for the last...oh, year or so...But the last real product nVidia delivered was the GeForce4. That was more than a year ago. In other words...for all intents and purposes, nVidia has failed to essentially bring any new products out for an entire calendar year (I don't count those AGP8x boards either), and this guy has the audacity to say they _almost_ missed just _one_ cycle???

[Pete]

If that's the review I think it is, they had screenshots comparing nV's "16x" AA to ATi's 6x AA--I just can't understand how a supposed 16x AA algorithm can look worse at any angle than a 6x one, be it OG or RG.

Funny how PC Magazine managed to come to the logical conclusion, and a purported enthusiast site like this can still be so deluded/clueless.

Also, do people know that run-on sentences are bad form, or do they find proper punctuation too tiring? :P

[K.I.L.E.R]

Let me see if I understand this correctly...You actually think that because nVidia doesn't seem to be able to execute a plan anymore, that NV30 should not be compared with the 9800 because it was _meant_ to be released last year?

AFAIK, there are NO rules that say you have to compare this with that. I can compare an R300 to an NV1 if I wanted to. :)

[demalion]

Well, I had problems with the review, but I disagree with your criticism of the first quote.

The NV30 was meant to compete with the R300 and not the R350, and I don't see a problem with holding the opinion that it is fairer to compare them, as long as price and availability comparison information is provided to the user along with the opinion so they can judge for themself. So, what I'd disagree with is his implication that comparing the R350 versus NV30 (ultra) is unfair by his ignoring how equivalent pricing and availability are and by completely ignoring the 5800 Ultra versus 5800 non Ultra performance and availability concerns when doing so.
If he'd included a non-slanted mention or recognition of that, the first statement would only be an opinion instead of something that seems to be trying to mislead....what he said instead was:

It seems like some Radeon 9800 Pro cards has already shipped out to those who have pre-ordered the cards but as for mass availability in the retail market, it should still be towards end of April. There are also a few GeForce FX 5800 Ultra cards sold to end users already and it seems like distributors here are getting ready to bring in some FX cards for retail.

Classic spin, where something just recently announced is "but as for mass availability...it should still be towards the end of April" and something that is half a year late is "a few...cards sold to end users already", used to describe equivalent availability situations (with actual indications being that the 9800 is more available, AFAICS).

Also missing is a discussion of Ultra versus non Ultra 5800 availability, which does give the article a "trying to stay on nVidia's good side" aura.

The "almost" missed one cycle comment does seem to further support that impression.

My own problems were along the lines of the apparently now established "nVidia 'told' me to compare 'Aggressive' (honest), but don't worry users I'm using 'Balanced'/'Quality' to give you the real scoop" type of word games. No mention of Application mode is made, or whether the R300 is using trilinear (they're both using default settings, so they must be comparable, right?)...almost like the reviewer has been living under a rock for a year (or is trying to pretend for some reason that he has).
No one seems to see through the nVidia PR enough to actually try and perform a real comparison of the strengths of Balanced/Quality and simply compare it to bilinear filtering and try to establish if it has advantages, instead depending on presumed ignorance and deception to make comparisons.

It seems an article perfectly suited to nVidia's bait & switch tactics.
Having difficulty believing someone's ignorance is so thorough regarding these issues that they aren't even mentioned, along with included admission in the article of nVidia's input into the article's benchmarking and other evidence about the type of input nVidia is offering, it makes it awfully easy to believe in a conspiracy to directly deceive consumers.

[Cuervo]

How could anyone compare the 5800 Ultra to anything? You can't get one. Well, I guess you could if you really wanted to. Gainwood? Right.

On the other hand, the NV350s aren't even in preview mode and I will get a 9800 Pro from the Fed-Ex lady tomorrow.

[Tahir2]

Gainwood?

I like that... that's funny :P

[K.I.L.E.R]

Gainwood?

I like that... that's funny :P

I thought it was Gainward? ;)
Grainwood also works. ;)

[OpenGL guy]

If that's the review I think it is, they had screenshots comparing nV's "16x" AA to ATi's 6x AA--I just can't understand how a supposed 16x AA algorithm can look worse at any angle than a 6x one, be it OG or RG.
It's quite easy. If you look around the Beyond 3D forums, you can find screenshots showing the positions of nvidia's 16x AA samples and the edge AA quality they give. If you compare nvidia's 16x AA edge quality to the 9700's 4x AA mode, you'll see that 16x AA is marginally better. However, the lack of gamma correction is a big strike against nvidia's implementation.

16x AA, as implemented by nvidia, doesn't compare to ATi's 6x AA.

[Pete]

I knew that RG + GC would trump higher-level OG (as the screens plainly show), OGLg, I was just trying to be dramatic. :)

[OpenGL guy]

I knew that RG + GC would trump higher-level OG (as the screens plainly show), OGLg, I was just trying to be dramatic. :)
Sorry, your drama went over my head ;)

[Tagrineth]

On the other hand, the NV350s aren't even in preview mode and I will get a 9800 Pro from the Fed-Ex lady tomorrow.

Well, I'd hope not! :shock:

[Ante P]

How could anyone compare the 5800 Ultra to anything? You can't get one. Well, I guess you could if you really wanted to. Gainwood? Right.

On the other hand, the NV350s aren't even in preview mode and I will get a 9800 Pro from the Fed-Ex lady tomorrow.

Since when can't you get a 5800 Ultra?
I've seen Leadteks, Gainwards and MSIs cards in retail. I've also seen a review of a PNY card.
That's four pretty large manufacturers.
On top of that I've spoken to people at various forums and Terratec have their cards for sale in germany.
Asus have their cards out by now too. Abits are arriving within the week. (Next Monday)

5800 Ultra is and will continue to be available, though in small quantities.

Not that I can think of any reason that a sane person would buy one, but that's a different dilemma. ;)

[Tagrineth]

Seems like they're hellishly rare stateside, I haven't seen any in any local places...