Radeon 9700 reviews

[Joe DeFuria]

Nice summary Rev.

To answer your questions about the Anand / NV30 thing:

If they state :

Chevy has a new SUV coming that promises more horsepower?
Chevy has a new SUV coming that promises more room?
Chevy has a new SUV coming that promises more gas mileage?

After talking to Chevy, is that OK to do with out the car?

That depends.

Has Chevy publically released information regarding this new, upcoming SUV? Is Anand allowed to tell everyone exactly what Chevy told him, or did Chevy put him under "NDA" for all the "promised specs" etc, and will only allow vague "off the record" type comparisons?

The thing is, until the specs are at least publically announced, the specs have a fairly high probability of changing -- especially with respect to clock speed and availability. As we know, even AFTER public announcements, sometimes things change.

"Public Announcement" is akin to the manufacturer having SOME degree of confidence in the actual, final form of the shipping product.

Would it really be surprising to find out, when publically announced, that

* Chevy offers an OPTION to have more horsepower than the "standard" Ford, but you must buy an expensive optional engine to get more power, making the price much higher than the comparable ford...

* The Chevy does has more interior toom: 5 cubic inches....

* It gets better gas milage, but only with the lower powered engine....You dodn't think Chevy wanted anyone to assume meant better gas milage AND more power with the SAME engine?

* Despite higher power, It has lower torque, and lower towing capacity than the Ford...

See what I'm getting at?

In short, the "litmus" test, as far as I'm concerned, on whether or not someone should really talk about a "future" product, is whether or not the info on that future product has been ofiicially and publically made available.

 

[Sharkfood]

Rev, I only see a couple problems with your post/opinion. As they are simply your opinions, it simply means we disagree on these few points.

Also in my experience is that hardware companies do not use "dirty tactics" nor "offer incentives" in order for their products to almost-always receive "favourable" reviews.

I think the special case here is "favorable" reviews. I don't believe the same can be said of "unfavorable" reviews. There is enough case history involving ReactorCritical w/ GTS vs V5, Riva3D with the pulling of a V5 vs GTS article, Firing Squad being immediately asked to pull a "Win2K Shootout" to include some NV Win98 numbers on the Win2K shootout, down to Kyle at HardOCP being "gifted" a special Quake3 bin-parser and patcher from you-know-who, to about a dozen other situations that have occurred and continue to occur.

The very same politics occurred with Matrox and selective distribution of test samples. And there was even such discussion here whether or not R300 (and possibly NV30) test sample distribution is somehow hinged on past "judgements" of hardware.

It's a game really- but I don't believe anyone is going to buy that the game doesn't exist. The sites that will have to wait 2-3 weeks to go buy a shelf version of the R300 missed the opportunity to have the "early scoop" and their readership will look to other sources if this is a repeating trend. PR reps from companies deciding test samples can play whatever games they wish, as has been illustrated in the past. After all, you don't bite the hand that feeds you- even if that hand theoretically gives you rat feces, you are expected to at least portray the rat feces as shiny, glowing and of some value.

These companies do not use "dirty tactics" nor "offer incentives" to me - they only expect fair reviews.

The can of worms here is what one considers a "fair review" as well as one's definition of "dirty tactics." I'm sure you, Rev, probably dont have many problems with this as your writing style has never been one to focus on negatives, but instead keep rivalry to a minimum in product reviews. If you take your older GF3 review as an example, the gist of the article was basically: here's the GF3. Here's what it does, here's how it works, here's a ton of (full sized) screenshots of how it looks, and here's all the features.. The End. Dave's R300 review follows the same premise. Informative, concise, without rivalry.

Unfortunately, the "gist" of other reviews aren't as focused. The "gist" of Anand's R300 review is "see, it isn't much better than a GF4 from the specialized, uncomparable conditions I've chosen. Oh, and you should wait for the NV30" which is nothing but agenda laid thick. More honest examples are like FiringSquad's Win2K shootout, which had the gist of "As you can see, the NV chipsets are scoring well below even the V5 and some Radeons" but did so with pure honesty. Being asked to pull such findings and line the page with Win98 benchmarks defeats the purpose. By my definition, being contacted by an IHV to make these kinds of "revisions" or being asked to remove/add unrelated data falls into the category of "dirty tactics"... and the unsaid repercussions of not following instructions from an IHV with "suggestions" becomes the "incentives."

The number of sites with Serious Sam 45 degree rotated textures on the 8500 imbedded inside of GF4 reviews is another good example. The sites that were given Quake3 binary patchers from NVidia is also another good example.

What is considered "allowable" for a website to present is the game here. If a review is in the form of a "shootout" versus "flat product information," what is considered "allowable" by IHV A or B is the dirty tactics/incentives portion.

I'm sure there are times of conflict where a "flaw" or other issue is discovered early in the game. We've seen it with multitexturing + trilinear on the VSA-100s, and more recent with HyperZ + Smoothvision on the R200s, down to anomolies with multitexturing fillrate + AF on the GF4. It becomes a personal decision of one's relationship with their trusted friends inside an IHV and their relationship with their readers. If, say, something that could have some impact on product release is discovered with the R300 or NV30, would this *not* be a tough decision for inclusion in a hardware preview? Every chip has it's flaws and limitations, and PR reps likely know of them and sit with crossed fingers that sites don't discover them. This is also the reason why IHV's strive to obtain test samples of competitor's products, in hopes they can find them and provide 'hints' in forums or to their reviewing circle of hardware testers. Anyone that doesn't think NVIDIA has a few R300's under intense study right now... I can assure is only fooling themselves.

When Anand, HardOCP or Toms suddenly produces a 5-page "revelation" of such an issue, does anyone actually think any one of these sites discovered this on their own?

There is absolutely nothing wrong with this reality. It's big business. AMD and Intel have gone at things this way for years, as do Sony and Mitsubishi, Quantum and Western Digital.. you name it. Competitive spirit is alive and well and a GOOD thing. I just dont think it's fair to suggest that suddenly things have changed when they most certainly have not. I'd like to think they have, but only time will tell.

 

[Reverend]

Joe and Shark, both of you have valid points.

Has Chevy publically released information regarding this new, upcoming SUV? Is Anand allowed to tell everyone exactly what Chevy told him, or did Chevy put him under "NDA" for all the "promised specs" etc, and will only allow vague "off the record" type comparisons?

Joe, CineFX is publicly available. NVIDIA allowed me to publish my CineFX article, exactly as they briefed CineFX to me. I was not NDA'ed on CineFX.

I do understand that until a product actually exists, all announced specs prior to that is "doubtable".

Shark, as believeable as it is, I haven't any proof that NVIDIA provided Kyle with the Quake/Quack bin parser. You guys can call me pro-NVIDIA for saying this but that's the truth. And, honestly, even if it came from NVIDIA, I don't really see a problem with that.

When Anand, HardOCP or Toms suddenly produces a 5-page "revelation" of such an issue, does anyone actually think any one of these sites discovered this on their own?

You simply want the revelations or you want who-actually-told-these-websites-the-revelations? [/quote]

 

[Reverend]

Thanks Rev.

I was going to write something in response, but that will have to wait till later.

Just one thing though.... I thought you were married? So whats with the date?

What, being married means I can't have dates?!

 

[Dave Baumann]

What, being married means I can't have dates?!

The bowel loosening fruit? Yes. Anything else? Probably not!

 

[Joe DeFuria]

Joe, CineFX is publicly available.

Yes, CineFX is publically available...things that one needs to know about "waiting for the next product, are not. There is no public info on such things as bus width, projeted clock speed, price, availability, etc.

In short, public information "about a language based on an architecture" is not a substitute for public information about a product, IMO.

 

[AzBat]

Reverend,

I liked your reply regarding sites, favoritism, etc. In the 5+ years I had been in the graphics business I dealt with respective hardware companies while doing work on my own independent 3D graphics web site(Dimension 3D) and a market research firm(Jon Peddie Associates). When I started my site it was my intention to keep it a neutral as possible. This meant reviewing all products from all companies. I found that it could be done even when I did review on freebies I received from the companies. I was never asked to provide favorable reviews in return. My character wouldn't allow it and these companies understood this after they got to know me. However, at that same time I did find that as time went by I was being excluded from the group of sites that received boards for the first looks. This somewhat changed as I became an employee of JPA. I was given as much respect as market analyst could get. We would get pre-release hardware and I could have done reviews on them, but I personally felt that didn't jive with the image I thought JPA had. So I decided to do overviews of trade shows and new benchmark applications. I figured this would keep me from being biased to one hardware company over another.

With that being said I do believe you're right in that in order to be independent or as close to it as possible, you can't seriously run a web site that is affiliated with a hardware company by either your content or your domain name referencing a company's name or other trade mark(s). But by going that route, it's hard to create a site dedicated solely to 3D graphics and keep it going. I believe the guys here at Beyond3D.com know full well what I'm talking about. If you want the traffic that Anand has then you need to provide more than graphics coverage(CPU, motherboards, etc.). If you don't, then you have to make sure that you provide something of worth that other sites are not providing. With Beyond3D they have to be more technically focused and it works great for them, but they have smaller traffic than Anand.

What can current fan sites do to become independent? That's hard to say since you may forever be known as promoting one company over another. The only thing I think you can do is to piss off your affiliated hardware company with a truly independent review comparing their product to their competition. There have been cases where this has been done, but I don't believe those sites stuck to their guns. After a few hard hitting reviews like that and you should be able to start fresh with a new web site. However, the best thing is not to start a fan site to begin with. It's a lot harder, but it is sure well worth the work. I know I can sleep better at night with the decisions I made even though it eventually got me out of the business completely.

EDIT: I'd like to note that even with the above opinions, I still believe there is value in a fan site or I would have never seeked those kind of sites for inclusion into Dimension 3D. They do have advantages by not having to be the "Jack of All Trades". But I will say they are not created equal and thank God.

Tommy McClain