View Full Version : The Inquirer: Getting lamer every single article they post
Bah... The Inquirer can't even keep itself updated on OFFICIAL information!
As for the name we wont spoil a big plans to Nvidia to surprise the world and will keep you waiting for that for sometime next week, when NV30 will debut with a new mobile baby NV28M. It might not be the same day, OK?
That was AFTER Anandtech & others posted their NV28M review.
As for that "48GB/s effective" figure.... Anyone wanna bet they're wrong? :)
Only way I can see that making any type of sense is if FSAA 6x memory cost on a NV30 is the same as FSAA 2X memory cost on a NV25.
And compared to the R300, which already got a few basic compression systems for AA, it's going to be a lot more like FSAA 8X on the NV30 for the cost of FSAA 4X on the R300 ( this is based on that whole 1.4x color compression for the R300 )
So that would be more like 32GB/s compared to the R300's 20GB/s
Which would thus give a 60% memory advantage for the NV30 when in FSAA.
And I refuse to trust that. It just seems too much... But I guess we'll see that soon.
Uttar
Originally posted by Uttar
And I refuse to trust that.
I am shocked, shocked to find someone not believing the Inquirer.
After NV30 is released, I think it would be interesting to put together all the rumors the Inquirer posted about it and backtrack to see if they were at all correct.
Originally posted by DadGT
I am shocked, shocked to find someone not believing the Inquirer.
After NV30 is released, I think it would be interesting to put together all the rumors the Inquirer posted about it and backtrack to see if they were at all correct.
Well, they were correct for DDR-2... And they were the first to say the NV30 will use DDR-2. So, that's quite impressive - they actually got ONE thing right for once.
What would be even more fun is taking ALL rumors. They all seem to be very wrong, and, as strange as it sounds, it currently seems that The Inquirer got the most correct information ( but that doesn't mean it's correct )
I'm guessing they're also right about the 400+ core clock. I'm ready to do some serious crow eating if the best NV30 model doesn't have at least 400Mhz core clock.
But then again, maybe they just invented all of that crap and got the DDR2 & core clock things right by mistake.
Uttar
Originally posted by Uttar
That was AFTER Anandtech & others posted their NV28M review.
As for that "48GB/s effective" figure.... Anyone wanna bet they're wrong? :)
Well the origin of that figure is in the CineFX whitepapers (I believe - you'll have to ask over at Beyond3D for specifics) so it could well be accurate - how nVidia arrived at 48GB/sec who knows?
The Inquirer were right on a lot of the R300 details pror to it's launch. That's the annoying thing about them - they tend to get most things horribly wrong but get it right occasionally; enough to warrant a place in many people's favourites folder and a daily browse. :D
MuFu.
Originally posted by MuFu
Well the origin of that figure is in the CineFX whitepapers (I believe - you'll have to ask over at Beyond3D for specifics) so it could well be accurate - how nVidia arrived at 48GB/sec who knows?
MuFu.
nVidia NEVER arrived to such a figure. No CineFX paper talk about Memory Bandwidth beside AGP.
This is a rumor created some time ago by a *cough*french*cough* source. The post saying that figure for the first time actually was on nV News I think.
No offense intended for every french which might frequent this forum. It's just that it doesn't make any sense for nVidia to give such a figure to a french programmer so much before NV30 launch ( that rumor arised a few months ago AFAIK )
Uttar
Ah, ok. My mistake, thanks for the correction. :)
MuFu.
budd_wm
11-15-02, 06:30 PM
Heh, I'm betting those figures stated by the Inquirer are pretty accurate. But what do I know....;)
RipItup
11-15-02, 07:02 PM
They're probably mostly right, though there are lies, damn lies and then statistics ..for instance the 48GB might be right but then only when measured as X, when measured as a more real world Y then it's not really relevant.
I like the inquirer .. you can't complain about the PR departments of the manufacturers and then complain about the non PR orientated web sites as well.
F**king geeks ..never happy :D
SavagePaladin
11-15-02, 08:20 PM
I say lies, damn lies, statistics, and The Inquirer...who wouldn't know the truth from their ass.
***CENSORED***
11-15-02, 10:23 PM
:D. LOL!
budd_wm
11-15-02, 10:51 PM
Originally posted by SavagePaladin
I say lies, damn lies, statistics, and The Inquirer...who wouldn't know the truth from their ass.
You'll be eating some crow once nVidia releases the specs at Comdex...I shouldn't say anymore. :)
imtim83
11-15-02, 11:20 PM
budd_wm do you work for nvidia?
SavagePaladin
11-15-02, 11:33 PM
Originally posted by budd_wm
You'll be eating some crow once nVidia releases the specs at Comdex...I shouldn't say anymore. :)
hmm and how many times have they contradicted themself on various issues? I really don't care if they're right occasionally
Originally posted by Uttar
nVidia NEVER arrived to such a figure. No CineFX paper talk about Memory Bandwidth beside AGP.
This is a rumor created some time ago by a *cough*french*cough* source. The post saying that figure for the first time actually was on nV News I think. Lol! I hope they're not using that as a source of info without doing some serious investigation! The only goal of the post was to note that ddr2 has the capacity for some good bandwidth, and nothing more... It wasn't even saying a number for the bandwidth, just remarking that N chips times the bandwidth per chip = "a lot of bandwidth to play with" (quoting from memory)... It was only potential theoritical bandwidth at the ram chip level - totally ignoring the bus, the memory controller and the cache(s). The only link with NVidia was the assumption made on the french site that they were using those 1GHz ddr2 chips (which is correct I think - well we'll see soon)...
Originally posted by budd_wm
You'll be eating some crow once nVidia releases the specs at Comdex...I shouldn't say anymore. :)
Can i eat crow too if the 48 GB figure is false? I'm presonally betting most of the other information from The Inquirer is correct, but 48GB/s effective is simply ridiculous. I could trust 30GB/s effective or something, but this is ridiculous.
And as I said, they're comparing all of this to a GF4. The R300 already got several bandwidth saving solutions that the GF4 doesn't have ( color compression in FSAA, ... ) so the 48GB/s number would likely be lower if it was compared to the R300.
Now, there is ONE thing I could suppose as true... The AA algorithm I've described in another thread. If you compared performance/quality with the GF4's multisampling, it could justify a so darn huge multiplication.
As for other ways to get tot that huge number... I see none.
Uttar
StealthHawk
11-16-02, 04:32 AM
Originally posted by RipItup
They're probably mostly right, though there are lies, damn lies and then statistics ..for instance the 48GB might be right but then only when measured as X, when measured as a more real world Y then it's not really relevant.
I like the inquirer .. you can't complain about the PR departments of the manufacturers and then complain about the non PR orientated web sites as well.
F**king geeks ..never happy :D
insanity. the Inquirer has been wrong on many accounts even when they claimed to personally have inside information. i don't even begin to trust them when they just report rumors that are on other sites. in the last 6 months alone we've seen the folks over at the Inquirer misrepresent news ripped from other sites because they just simply refuse to check the context or credibility or their sources.
TI talk so much crap that their "journalistic" asses are jealous.
Stavros
11-16-02, 08:53 AM
I've never tasted crow before. Is it nice?
does it taste like chicken?
Would I be better off having turkey, goose or crow for xmas dinner.
I can quite believe nvidias PR department saying it has 48Gb effective bandwidth, but only between 2 and 3 in the afternoon, when all the planets are aligned
sebazve
11-16-02, 11:28 AM
i dunno about the inquirer article but the nv30 will be one big mother...:D and itll sure be faster than r300 i mean its a later product (despite it is delayed) but then it will come the r350 and it will surely kick some nv30 butt and so on..
Will we see benchmarks on monday and retail cards before xmas or just a paper launch with retail boards on february?I cannot accept the statement "limited supply for 2002" since that means that they wont be available worldwide.
ATI might paper launch r350 today and have retail boards on spring ala gf3 does that mean r350 is the fastest card in the world for 2002?:rolleyes:
ATI LoVeR 9700
11-16-02, 01:46 PM
Originally posted by Fotis
Will we see benchmarks on monday and retail cards before xmas or just a paper launch with retail boards on february?I cannot accept the statement "limited supply for 2002" since that means that they wont be available worldwide.
ATI might paper launch r350 today and have retail boards on spring ala gf3 does that mean r350 is the fastest card in the world for 2002?:rolleyes:
I had a chat with ATi reps, they said "No more paper launches! They hurt more than they do good."
Originally posted by ATI LoVeR 9700
I had a chat with ATi reps, they said "No more paper launches! They hurt more than they do good."
Great!!! I've just about had it with paper launches ,especially with athlon xp 2800+ announced september coming 2003.Why doesn't intel announce prescott right now!!!:rolleyes:
SavagePaladin
11-16-02, 10:18 PM
the 2800+ is out now, dude
budd_wm
11-16-02, 10:33 PM
Originally posted by SavagePaladin
the 2800+ is out now, dude
WAY off-topic, but I just had to say....Though I'm a fan of AMD, I think currently it's best to go with Intel. For one, Intel's 2.8GHz processor and AMD's 2800+ are priced the same (actually, AMD's is just a bit more). Plus with dual channel DDR mobo's coming out, Intel's processors will be able to take full advantage of it (more than AMD will) as you will get performace that is equal or better than you would using Rambus.
Yeah!
We have a first test with a Granite Bay and DDR266 Dual Channel
http://www.ocworkbench.com/2002/asus/p4g8x/p4g8xp1.htm
Check this, the DDR266 have similar performance then RAMBUS PC800.
But for me, i will wait for Springdale at Q1 or Q2 2003, he support FSB 667 and Prescott Processor, DDR333 Dual Channel, AGP 8X, Serial ATA.
vBulletin® v3.7.1, Copyright ©2000-2012, Jelsoft Enterprises Ltd.