GeforceFX=surprisingly slow

[Hammer]

20 odd days aint goona get a huge performance increase in all likelyhood, is it? (and yes I know its been done before)

The drivers v42.70 was still in beta form...

[Minion]

Hammer wrote:

Is that relevant?  (several times)

I do own MSI built VIA KT class chipset and have access to ASUS built to VIA KT class chipset test machines.

Is that relevant?   Is that relevant?   Is that relevant?   Is that relevant?
I now see how all of these arguments erupt on A.org.  YOU'RE NOT ALLOWED TO AGREE WITH ANYONE!  Or its not relevant or a waste of time.

 and the Nforce2 is worth about another 200+ rating on them.

What do you  mean by "Nforce2 is worth about another 200+ rating on them"?[/quote]
OK i'll reword it.  The Nforce2 Mobo's appear to be up to 10% faster in many tests than the VIA based ones.  Therefore, on a 2000XP, they are worth another 10% i.e. It then performs like a 2200XP or has an extra 200+

I checked your link - personally I dont care if the R300 ONLY has 96 bit pixel shader precision, that still allows 7.9x10^28 different positions.  I think thats plenty.
The point is the GeforceFX is a crock of sh*te.

[ksk]

Not sure if this was alredy posted in this thread, but anyway. Here's a anandtech view to the Radeon vs Geforce debate:
http://www.anandtech.com/video/showdoc.html?i=1779&p=16

It would be nice to see a 0.13 micron version of Radeon9700PRO running at 500Mhz or faster....  

[Phoenix]

OT Did I log on to [color=990000]Amiga[/color][/b][/i][/u].org?

I will never trust ATI's drivers completely - makes life difficult running a day to day PeeCee...

[Ilwrath]

@Hammer

Ilwrath's claim was 1 market segment = success. This case was not true for 3DFX.

No, that wasn't at all my claim.  My claim was don't neglect the market you're good in for a chance to try your hand in a market you're not so good in.

Pre-expansion nVidia prided itself on making shipping dates.  Huang (nVidia's CEO) quote -- straight from Wired magazine

Fact is, Huang knows there's little room for even one mistake in his business, much less the same one twice. It's the nature of the graphics-chip industry: A company rises to leadership only to miss a delivery window and rolls over for an upstart with a better technology. Cirrus Logic, 3dfx Interactive, Pseng Labs, s3, Rendition, Chips and Technologies - they once were all leaders; now they're all gone.

Nvidia has sidestepped the boom-and-bust cycle by hewing to a simple philosophy: Technology matters, but the production calendar rules. "The first breath of success for Nvidia came when we recognized that the PC market has a pulse that's regular and predictable," says chief scientist David Kirk. PC manufacturers ship machines to resellers twice a year - in April and August. That means Nvidia has to have a new chip ready each February and June.

(For the rest of the very good interview, go to
Wired's nVidia interview - July 2002 - It's worth a read)

Now, for a review of the facts of what's happened:
1) they're blowing ship dates.
They're late on the Feb ship date for the NV30, (Feb 03) and neglected to put anything interesting out for the June 02 ship date.  

2) While the NV30 is powerful, it's more of only a tweak above the 9700Pro, rather than the generation jump that was originally marketed...

3) Their financial security is highly debatable.  While they have diversified, it's cost them money.  Money they may not have, or that they SHOULD have applied to shipping the NV30 faster / better.

[Hammer]

No, that wasn't at all my claim. My claim was don't neglect the market you're good in for a chance to try your hand in a market you're not so good in.

Are you implying “nForce 2” chipset is not competitive with VIA KT class chipset?

Now, for a review of the facts of what's happened:
1) they're blowing ship dates.

Their chip fab contractor has blown the dates.

From recorded history, fab companies has missed their time schedule during their shift to .13 process.

2) While the NV30 is powerful, it's more of only a tweak above the 9700Pro, rather than the generation jump that was originally marketed...

Did they (i.e. nVidia)?

I recall, their official comparison was with the GeForce 4-4600 TI. Majority of the so-called hype was magnified by the press.

[Hammer]

 YOU'RE NOT ALLOWED TO AGREE WITH ANYONE!  

Your words.

OK i'll reword it.  The Nforce2 Mobo's appear to be up to 10% faster in many tests than the VIA based ones.  Therefore, on a 2000XP, they are worth another 10% i.e. It then performs like a 2200XP or has an extra 200+

A little bit of effort doesn’t bug anyone.

I checked your link - personally I dont care if the R300 ONLY has 96 bit pixel shader precision, that still allows 7.9x10^28 different positions.  I think thats plenty.
The point is the GeforceFX is a crock of sh*te.

Very centric response.

[Hammer]

It would be nice to see a 0.13 micron version of Radeon9700PRO running at 500Mhz or faster....

Both sides has their own advancement strategy.  

Refer to
NV35 plans for post NV30 era

[Minion]

Hammer wrote:

 YOU'RE NOT ALLOWED TO AGREE WITH ANYONE!  

Your words.

You are the one that said pointless when I agreed

A little bit of effort doesn’t bug anyone.

Thats why I explained what I mean otherwise I wouldn't have bothered, would I?

I checked your link - personally I dont care if the R300 ONLY has 96 bit pixel shader precision, that still allows 7.9x10^28 different positions.  I think thats plenty.
The point is the GeforceFX is a crock of sh*te.

Very centric response.

So you dont disagree, then?

[Ilwrath]

@hammer -

Are you implying “nForce 2” chipset is not competitive with VIA KT class chipset?

No...  I'm implying that they gave up the clear-cut lead in the fairly high-margin GPU market to create that chipset.  Basically the nForce is a good chipset, but it was a poor business decision, and the fact that it's horribly mis-marketed surely can't help matters.

First off, chipsets don't have the profits of high-end graphics processors.  Especially not chipsets that only work with AMD processors, which are currently suffering because very few OEM companies are producing computers supporting them.  In fact, about the only way to get a computer with an AMD processor is to purchase the poorest of the HP offerings -- which wouldn't use an nVidia chipset, or go white-box or self-built. White box and self-built machines don't sell in the same quantities (as say Dell) and have even LESS of a profit margin.

So, from that we can assume that nVidia is targetting the white-box and self-built market.  That market is mostly budget workstations and gamers.  The chipset still doesn't make sense, though.  The high-end version with firewire support has a semi-poor built-in sound card that no gamer would want.  All versions sport the on-board GeForce4MX, which isn't a good fit for gamers or graphics workstations, either.  (Gamers would rather have a GeForce4Ti, workstation users would rather have a Quadro, so they're certified with ACAD, Catia, etc.)

Exactly what market is nVidia trying to hit?? :-?

From recorded history, fab companies has missed their time schedule during their shift to .13 process.

Hmm...  Perhaps, but then again, if they were counting on this as the boost, then it wasn't the best of planning, either...

If they weren't as busy with other less profitable markets, they could have had a contingency plan for problems with the new manufacturing process, and been able to push the .13 process back to the next shipping round, while still delivering a quality update in the meantime.

I recall, their official comparison was with the GeForce 4-4600 TI. Majority of the so-called hype was magnified by the press.

Very good point.  I don't think nVidia ever did make that comparison, but it was what was expected of the card.  Releasing a graphics card that falls far short of expectations is never a good thing, even if those expectations are higher than you meant to set them.

[Hammer]

No... I'm implying that they gave up the clear-cut lead in the fairly high-margin GPU market to create that chipset. Basically the nForce is a good chipset, but it was a poor business decision, and the fact that it's horribly mis-marketed surely can't help matters.

Mis-marketed? How?

Exactly what market is nVidia trying to hit??
.

As with Intel 845 (and E7) series, it should target at all price points.  

First off, chipsets don't have the profits of high-end graphics processors. Especially not chipsets that only work with AMD processors, which are currently suffering because very few OEM companies are producing computers supporting them.
.

Based on what?
 

 In fact, about the only way to get a computer with an AMD processor is to purchase the poorest of the HP offerings -- which wouldn't use an nVidia chipset, or go white-box or self-built. White box and self-built machines don't sell in the same quantities (as say Dell) and have even LESS of a profit margin.
.

Refer to http://www.simhq.com/simhq3/hardware/previews/nForce%202/

All versions sport the on-board GeForce4MX, which isn't a good fit for gamers or graphics workstations, either

not true... not all NForce2 boards have onboard video... most dont... the NForce2 supports AGP 8x... I got an EPoX 8RDA+ without onboard.



I agree entirely about Nvidia though their going t*ts up in a big way from the look of it... their chipsets are good but their AMD based and no OEM's are going to use them... OEM market penetration of AMD is very lacking right now...all because they didnt cap their cores....

Then you have the issue of their graphics card sucking in general... their Quardo's are about the only thing worth buying right now...thats because their price/performance ratio is hitting a sweet spot right now that ATI cant touch...but probably soon will....

The "LOW END" wich many people talk about...that makes everyone all their money is having FIERCE competition right now...
Trident comming out with a new low-end GPU, you got the ATI low-end offerings...the Nvidia low-end offerings... you got Xabre 400 hitting at the low-end... you still have Intel putting its integrated gpu on boards.... you got Matrox hitting that somewhat .... its a hard market right now on the low-end... Nvidia could loose that market in a matter of months...if SiS,Trident, or someone else(VIA?) comes up with a much cheaper but competative low-end solution that dosent cost much head/etc...

I think Nvidia is in sad shape right now... in this market 8 months makes or breaks a company...

the GeForceFX reminds me of the Voodoo5 in more ways then one...and thats NOT a good thing... by in large the Voodoo5 is what sunk 3DFX and it'll sink nvidia to if they cant pull out of this slump.... looseing visiontek is a big-deal to them aswell.

[Hammer]

You are the one that said pointless when I agreed

I didn't say it was "pointless". Please find the word "pointless" in my post.

So you dont disagree, then?

Are there any reasons to agree in the first place?   The product is not even released in the market place.

I checked your link - personally I dont care if the R300 ONLY has 96 bit pixel shader precision, that still allows 7.9x10^28 different positions. I think thats plenty.

The "first cause" post was in regards to transistor count and the potential reason why relatively large number of transistors was included with GeF FX.    

You can’t get something for nothing in regards to hardware features vs transistor count.  

The point is the GeforceFX is a crock of sh*te.

You haven’t made any substantial evidence why that view is valid. Except for writing fan fiction.

Refer to http://www.nvnews.net/articles/geforce_fx_commentary/index.shtml for past driver improvements(within the GeF 4 TI line).

Refer to http://www.penstarsys.com/editor/Today/nvidia3/nvda_tdy_4.htm for more information regarding GeF FX, nForce 2, John Carmack and 'etc'.

[Hammer]

The high-end version with firewire support

So does the similar priced ~$150 USD Audigy 2 card...

has a semi-poor built-in sound card that no gamer would want.

It’s a similar chip that was included in a X-BOX btw… “No gamer wants” assertion is completely wrong.

Have you tried running +32 channel midi–channel midi file on emu10k series card yet (for HW accelerated not the buddled SW option)?  
   
Nvidia’s Sound Storm has access to bandwidth beyond the PCI limit (All CL’s DSP cards are limited by this standard). There was also a talk about releasing nVidia Sound Storm for PCI-Express enabled slots (PCI-Express enabled slots has the necessary bandwidth to support 200+ HW accelerated channels).    

It’s good enough to match Audigy 1 and SBLive 5.1 DE level markets. Sound Storm also includes Creative Labs’s style EAX user configurable control panel and Dobby Digital Encoding (for multi-channel DirectSound3D titles).

Dobby Digital Encoding feature has yet to be included in VIA's and CL's audio card add-ons solution.

Sound Storm is a 24Bit DSP/APU chip (just like CL’s emu10k series).  But this is dependant on motherboard vendor’s installed CODEC. CODEC stage can be bypass via direct digital SPDIFOUT ports.

NVidia’s “Sound Storm” is a Microsoft reference for hardware implemented DirectX 8.x class audio.  

For Sound Storm's Dobby Digital Encoding feature refer to
http://www.overclockers.com.au/article.php?id=134416

All versions sport the on-board GeForce4MX,

Completely wrong. I recall ASUS A7N8X Deluxe (SPP/MCP-T) doesn’t include MX4x0 level IGP…

which isn't a good fit for gamers or graphics workstations, either

It runs QuakeIII and relate titles well enough compared VIA /Savage 2000 solution.

[Hammer]

the GeForceFX reminds me of the Voodoo5 in more ways then one...and thats NOT a good thing... by in large the Voodoo5 is what sunk 3DFX and it'll sink nvidia to if they cant pull out of this slump.... looseing visiontek is a big-deal to them aswell.

At least nVidia hasn’t made those massive 4 chips in 1 AGP card.

The lesser/cheaper GeF FX 5800 (9700 non-Pro edition market segment) sibling doesn’t require such cooling add-ons.

Note that ATI has survived from being second to nVidia from the past history. 3DFX’s collapse is a combination of legal battles with vNidia, non-completive value/middle markets segments, reduced distribution access (i.e. switched to non-multi-vendor sourced model(e.g. 3DFX only vendor)).  

[Hammer]

the GeForceFX reminds me of the Voodoo5 in more ways then one...and thats NOT a good thing... by in large the Voodoo5 is what sunk 3DFX and it'll sink nvidia to if they cant pull out of this slump.... looseing visiontek is a big-deal to them aswell.

Refer to http://www.tech-report.com/sendto_friend.x/4679/
Gainward to offer quieter GeForce FX?