Mobility Radeon 9000

Tom pulls through again...

Benchmarking the mobility 9000 against a yet-to-even-be-announce announced chip from nVidia?

And questioning "which chip will be available first" when all the other reviewers seem to know and point out (because it's unusual in the notebook sector) that mobility 9000 solutions are shipping "imminently". (Next week or two for tier 3 and tier 2 vendors)?

Oh well...
 
Joe DeFuria said:
Tom pulls through again...

Benchmarking the mobility 9000 against a yet-to-even-be-announce announced chip from nVidia?

And questioning "which chip will be available first" when all the other reviewers seem to know and point out (because it's unusual in the notebook sector) that mobility 9000 solutions are shipping "imminently". (Next week or two for tier 3 and tier 2 vendors)?

Oh well...

Even so the Radeon 9000 mobile part still outperforms in tests even the power consumtion tests.
 
True but again we are talking about 'proper' journalism, benchmarking a card/chip that has not even been announced yet is to say...very strange...one example is it takes away from the unannounced chip as it shows it in a bad light....I think Tom has flipped his noodle, or the Amonia in his perms has affected brain functions.
 
Yeah, though those power consumption tests were pretty meaningless. Just shows the power consumption under a full load. Neither chip had their "power saving" features enabled, which would make for a more interesting comparison IMO.
 
Doomtrooper said:
True but again we are talking about 'proper' journalism, benchmarking a card/chip that has not even been announced yet is to say...very strange...one example is it takes away from the unannounced chip as it shows it in a bad light....I think Tom has flipped his noodle, or the Amonia in his perms has affected brain functions.

Agreed.

BTW, Tom and proper journalism together? Or even journalism? C'mon, don't joke. He's just a disgusting pocket guy... ;)
 
Well, I agree with his decision to include the 460, actually.

The problem with it is the lack of technical sophistication of this comparison, coupled with value judgements made as if there was some technical sophistication or analysis of performance...similar to the Lars syndrome, but not nearly as bad. That, and a complete discounting, which even Anand had, of the benefits of unique features offered by the chip over any NV1x design outside of any speed increases that might occur due to them. The inappropriateness of treating the 460 as shipping is a side effect of the lack of sophistication in the article, as its use otherwise is justifiable (IMO).

In other words, for the shallow, superficial, and simple-minded standard to which the site aspires for graphics reviews, it was a good article, atleast in my opinion.

I find that I did like the last joystick comparison at the site, though. :p

As for Anand, it strikes me that either he has learned from his past mistakes to some degree, or he is preparing to turn like a jackal on weakened prey. I find the benchmarks good, but the emphasis on the UT 2003 benchmark is still a distortion in my view...though I think the analysis overall was creditable. It could just be that his bias is in the process of being shifted after being impressed by ATi and we can look forward to some better objectivity while it is in transition. Then again, I'm pretty jaded by now...

EDIT: Oh, it was done my Matthew Witheiler, not Anand...well in that light I'd have to say it was just a good review without the past context I had in mind.
 
The inapporiateness of treating the 460 as shipping is a side effect of the lack of sophistication in the article, as its use otherwise is justifiable (IMO).

Well, my problem isn't that the 460 isn't yet shipping and available in products (neither is the mobility 9000).

My problem is that the 460 is not even announced as a product from nVidia.

http://www.nvidia.com/view.asp?PAGE=geforce4go

For all I know, this might change tomorrow. But if that's the case, Tom should wait until tomorrow to publish the review...
 
Joe DeFuria said:
The inapporiateness of treating the 460 as shipping is a side effect of the lack of sophistication in the article, as its use otherwise is justifiable (IMO).

Well, my problem isn't that the 460 isn't yet shipping and available in products (neither is the mobility 9000).

My problem is that the 460 is not even announced as a product from nVidia.

http://www.nvidia.com/view.asp?PAGE=geforce4go

For all I know, this might change tomorrow. But if that's the case, Tom should wait until tomorrow to publish the review...

I understand and agree to an extent, but he had it in a laptop, so it doesn't seem completely inappropriate to me even if unannounced.

Another way to state my sentiment is I don't have a problem with the use of the 460, though I find Tom's use to be less than desirable. The 460 is after all just a higher clocked 440...it would be proper to overclock a 440 to achieve speeds to simulate the 440's soon to be announced successor if it is done in the proper context, wouldn't it?
 
I am always against using "unannounced" products for one primary reason:

Things have a tendency to change right up until the time official announcements are made. That's one reason why things aren't announced....they aren't finalized. ;) Particularly: clock speeds, memory speeds, availability.

Heck, sometimes, things change even AFTER official announcements. But at least an official announcement is the IHVs way of relaying some confidence in the expectations of the shipping product.

Tom may be testing a 460 based on the "planned specs" for the part, but it may very well end up that the 460 specs will change (for better or worse). And a comparison based on "the wrong specs" isn't particularly beneficial to anyone.
 
Joe DeFuria said:
I am always against using "unannounced" products for one primary reason:

Things have a tendency to change right up until the time official announcements are made. That's one reason why things aren't announced....they aren't finalized. ;) Particularly: clock speeds, memory speeds, availability.

Heck, sometimes, things change even AFTER official announcements. But at least an official announcement is the IHVs way of relaying some confidence in the expectations of the shipping product.

Tom may be testing a 460 based on the "planned specs" for the part, but it may very well end up that the 460 specs will change (for better or worse). And a comparison based on "the wrong specs" isn't particularly beneficial to anyone.

That's what I mean by proper context. With a statement like that about the the product being unannounced and without statements like the 9000 M is a trivial speed increase when compared to this unannounced product that he has and using that to reach conclusions about the 9000 M, the use of the 460 could have been done in an acceptable way.

To make this shorter, since I don't seem to be expressing what I mean clearly to you: Is there any circumstance under which an unannounced product could be used as a reference in a benchmark of a "current" (announced, shipping to manufacturers, launched, whichever you find suitable) product? If you answer yes, then you are just saying what I'm saying in a different way. If you say no, we disagree, and atleast we can be clear on that.
 
Is there any circumstance under which an unannounced product could be used as a reference in a benchmark of a "current" (announced, shipping to manufacturers, launched, whichever you find suitable) product?

No. IMO, if a product is not publically announced, there are no circumstances in which it should be used in any sort of comparison to another product that has been publically announced.

So we'll just agree to disagree on that then. 8)

Now...if you have two unannounced products, and wish to compare them...;)
 
Joe DeFuria said:
Is there any circumstance under which an unannounced product could be used as a reference in a benchmark of a "current" (announced, shipping to manufacturers, launched, whichever you find suitable) product?

No. IMO, if a product is not publically announced, there are no circumstances in which it should be used in any sort of comparison to another product that has been publically announced.

So we'll just agree to disagree on that then. 8)

Now...if you have two unannounced products, and wish to compare them...;)

So if someone has benchmarks of the NV30 before nvidia announces final clock speeds or a card based on it, I expect at that time you will refuse to read an article comparing it to the R300 on principle, or urge them not to publish it?...

:p

Let me know when you do that so I can try to convince them otherwise for the rest of us. :LOL:
 
So I can expect to gain/loss only a handful of fps between these products and only 5 more minutes of play time when unplugged. Doesn't seem to be a huge difference to sway someone one way or another.

I think it will come down to a factor of price.
 
DemoCoder said:
So I can expect to gain/loss only a handful of fps between these products and only 5 more minutes of play time when unplugged. Doesn't seem to be a huge difference to sway someone one way or another.

I think it will come down to a factor of price.

Pirce and FEATURES.

The R9000 is a DX8.1 card...The GF4 Go is a DX7 card.

I know what one i'll rather have.
 
Doomtrooper said:
True but again we are talking about 'proper' journalism, benchmarking a card/chip that has not even been announced yet is to say...very strange...one example is it takes away from the unannounced chip as it shows it in a bad light....I think Tom has flipped his noodle, or the Amonia in his perms has affected brain functions.
What is the big deal about if a chip is paper launched or not?

By your logic we shouldn't have seen any Radeon 9700 benchmarks until it was on store shelves, but I don't hear you complaining about that?

I appreciate him putting in the new 460 as well, since I'm looking to buy a laptop in a month or two and want to see what choices I have. Whether or not you think it's "unfair" is not really relevant, because they're both chips and they're both not actually on store shelves at this very moment, so what's the problem?
 
So I can expect to gain/loss only a handful of fps between these products...

The DX 8 support was already mentioned. So if you prefer to have to turn features off (that are at this time are finally starting to get implemented more and more in games), go for the GeForce4.

As for performance:

Doesn't seem to be a huge difference to sway someone one way or another.

I think it's safe to say that notebook buyers tend to "upgrade" to new systems less often than say, one upgrades a video card in a PC, right? So to buy a notebook now, with the consideration of how it will run Doom3 is a fair point. Do we need a refresher on which architecture is better suited for such a game:

Carmack on GeForce4 MX tech:
Do not buy a GeForce4-MX for Doom....GF4-MX will still run Doom properly, but it will be using the NV10 codepath with only two texture units and no vertex shaders. A GF3 or 8500 will be
much better performers.

Carmack on Radeon 9000:
"The M9 laptop part allows state of the art high end game development to be accomplished on a laptop platform for the first time really since we've moved to hardware acceleration."

"You know, on our current work at Id right now we're still pushing really hard to make Doom run well on various high end desk top cards. So it's pretty startling to be able to fire it up on a laptop and see it run at a really pretty startling good pace."

I know which chip I'd rather have in a notebook....
 
What is the big deal about if a chip is paper launched or not?

I repeat:

Things have a tendency to change right up until the time official announcements are made. That's one reason why things aren't announced....they aren't finalized. Particularly: clock speeds, memory speeds, availability.

Heck, sometimes, things change even AFTER official announcements. But at least an official announcement is the IHVs way of relaying some confidence in the expectations of the shipping product.
 
Back
Top