PDA

View Full Version : NV35 to be like initial R350 reviews?


Solomon
05-11-03, 05:58 PM
I'm having a strange feeling that the NV35 will be like the 9800 reviews were they showed the 9800 tower over the 9700 Pro because of the drivers they used. Remember when that TT Hardware or something hacked or "forced" the drivers upon the 9700 Pro to show that the 9800 increase in performance was due to the R350 paths..

I'm thinking that the NV35 will be exactly the same with the 50.xx drivers. What I wondering is if the 50.xx drivers will have the same effect on the NV30.

I don't know about you guys, but I'm more interested in these 50.xx drivers then the actual NV35. Because we all know the NV35 is what the NV30 should of been in the first place. So for me, there really isn't anything to get excited about. IMO.

We all know what the NV30 is supposed to do, the NV35 finally does it. Heh.

Regards,
D. Solomon Jr.
*********.com

surfhurleydude
05-11-03, 06:18 PM
Well, by specs, the NV35 should be about 60% faster than the GeForce FX in bandwidth limited situations like Anti Aliasing just because of its massive bandwidth advantage. This isn't like the R350, where the differences were very small between the card and its predecessor.

Solomon
05-11-03, 06:24 PM
True... The 256-bit is geared for FSAA/AF performance increases. But it will be interesting to see how real world performance will be if not using the 50.xx drivers "or" how the NV30 reacts to the 50.xx series.

Regards,
D. Solomon Jr.
*********.com

bkswaney
05-11-03, 07:20 PM
Originally posted by Solomon
True... The 256-bit is geared for FSAA/AF performance increases. But it will be interesting to see how real world performance will be if not using the 50.xx drivers "or" how the NV30 reacts to the 50.xx series.

Regards,
D. Solomon Jr.
*********.com

Agreed... I've been saying sense I got the NV30 that the only thing holding this core back is the 128bit memory.
The core on this chip is a beast.
I know the NV35 is going to be a great card. With 500+mhz overclocks on the NV35 the scores are going to be all time highs. :D

From the looks of the first benches the NV30 is going to be a MUCH better card with the Det 50 drivers. "from looking at the nv35"

Nvidia is going to have some great cards in the future coming from the NV30 core.

The Baron
05-11-03, 07:26 PM
Solomon, you're right again (just like how you were with the R350 ;) )...

StealthHawk
05-11-03, 08:42 PM
You'll find out very soon ;)

Solomon
05-11-03, 09:09 PM
Originally posted by The Baron
Solomon, you're right again (just like how you were with the R350 ;) )...

:D

Once the drivers finally emerged, it proved that the ATi Radeon 9800 was identical to the 9700 Pro clock for clock. Granted some of it was optimized to give you a boost in pixel shaders. But overall sense from one card to another was identical :(

I got slammed hard for that little editorial post on the that. I guess the truth hurt some die hard fans! Hehehe.

Regards,
D. Solomon Jr.
*********.com

StealthHawk
05-12-03, 05:08 AM
Originally posted by CoWBoY
Well,

I hope we don't see something like the Tident fiasco in a couple weeks being that there is aloot missing in pic's when you compare the 5900 to the 5800.

Tident Fiasco (http://www.spodesabode.com/content/article/trident)

I am sure this is one of many articles to develope in the next weeks for sure. :(

well now that nvidia defaults Application to Quality which is enabled by default, I don't think they will plan on lowering texture quality to get high benchmark results. They already tried that, remember?

silence
05-12-03, 05:43 AM
Originally posted by Solomon
:D

Once the drivers finally emerged, it proved that the ATi Radeon 9800 was identical to the 9700 Pro clock for clock. Granted some of it was optimized to give you a boost in pixel shaders. But overall sense from one card to another was identical :(

I got slammed hard for that little editorial post on the that. I guess the truth hurt some die hard fans! Hehehe.

Regards,
D. Solomon Jr.
*********.com

well.....9700 <> 9800 might have been same clock per clock, but that same core....so it's not THAT much surprise. NV35 is basically NV30 with 256-bus and DDRI.....so difference is bigger. bandwidth given with 256-bus will make difference, exactly what was missing on NV30.

i dunno about dets 50.xx ,but if they are optimized for nv35, i see no problem there........

let's wait and see.....

MuFu
05-12-03, 07:22 AM
Inital reviews aren't with the Det50s, AFAIK. All the ones I know of use 43.xx, I think.

MuFu.

Evildeus
05-12-03, 07:44 AM
I beg to differ, 44.** :afro: I think we should learn more on what Det FX is at that time also :)

jbirney
05-12-03, 09:13 AM
Originally posted by Solomon
:D

Once the drivers finally emerged, it proved that the ATi Radeon 9800 was identical to the 9700 Pro clock for clock. Granted some of it was optimized to give you a boost in pixel shaders. But overall sense from one card to another was identical :(

I got slammed hard for that little editorial post on the that. I guess the truth hurt some die hard fans! Hehehe.

Regards,
D. Solomon Jr.
*********.com

I hate to disagree with you but didn't HardOCP do a review where they used the same Clock speed and same drivers and it showed the R9800 was a bit faster???? HardOCP site is a bit slow atm (heheheh) so I am trying to remember this one only...I will check for linky later

Hanners
05-12-03, 09:23 AM
Originally posted by jbirney
I hate to disagree with you but didn't HardOCP do a review where they used the same Clock speed and same drivers and it showed the R9800 was a bit faster???? HardOCP site is a bit slow atm (heheheh) so I am trying to remember this one only...I will check for linky later

You're right I think, I'm pretty sure [H] did... Beyond3D certainly did one showing the performance improvements of the 9800 Pro when clocked at 9700 Pro speeds.

Solomon
05-12-03, 09:50 AM
Originally posted by jbirney
I hate to disagree with you but didn't HardOCP do a review where they used the same Clock speed and same drivers and it showed the R9800 was a bit faster???? HardOCP site is a bit slow atm (heheheh) so I am trying to remember this one only...I will check for linky later

They did, but the 9700 Pro wasn't using the 9800 threads. That is what TT Hardware was able to do. They basically tricked the drivers to thinking the 9700 Pro was a 9800 Pro. So this caused the 9700 Pro to use the 9800 data paths and it showed that the two cards were basically identical give or take 1 to 3 fps difference. :(

Regards,
D. Solomon Jr.
*********.com

saturnotaku
05-12-03, 10:17 AM
Solomon,

I agree with you when you say you're more anticipating the release of the Detonator FX drivers. I say that first because I'm a GF FX owner and I want to see how performance and/or image quality improve, if at all.

In the end, though, I think even with the new drivers the NV35 will be noticably ahead of the NV30 when you crank up the AA and aniso. Without AA or aniso, the cards will probably be comparable.

Solomon
05-12-03, 10:26 AM
True... What I think would be interesting is a Ultra vs Non-Ultra 5900 comparison. Either overclock one or under clock the other and compare the two. This would be cool to see how much impact the 256Mb is on the Ultra vs Non-Ultra. I know we seen it already with the 256Mb version of the Radeon 9800 Pro's today... So hopefully we see a comparison like that soon! Hehe.

Regards,
D. Solomon Jr.
*********.com

digitalwanderer
05-12-03, 11:42 AM
Originally posted by Solomon
They did, but the 9700 Pro wasn't using the 9800 threads. That is what TT Hardware was able to do. They basically tricked the drivers to thinking the 9700 Pro was a 9800 Pro. So this caused the 9700 Pro to use the 9800 data paths and it showed that the two cards were basically identical give or take 1 to 3 fps difference. :(

I thought I'd seen a benchy that had been run with the 9800 downclocked AND the 9800 pathways enabled on the 9700 Pro they ran it against and the 9800 core still averaged about 5-10% higher depending on the bench. Not a big dif, but still a measurable difference.

I can't find a link to back that up, but I thought that I'd read it. :eek2:

jbirney
05-12-03, 12:02 PM
Yea I thought that as well Digital. I was sure they forced them to use the same drivers to level the playing feild. Their goal was to see if any of the so call AA optimizations did much...and IIRC they saw a slight improvement..

StealthHawk
05-12-03, 04:07 PM
I'm pretty sure that when FSAA was enabled the r9800 showed improvements over the r9700 even when the r9700 used the r9800 driver pathway.

edit: discussion (http://www.nvnews.net/vbulletin/showthread.php?s=&threadid=9162)

ChrisRay
05-12-03, 04:11 PM
Originally posted by Solomon
I'm having a strange feeling that the NV35 will be like the 9800 reviews were they showed the 9800 tower over the 9700 Pro because of the drivers they used. Remember when that TT Hardware or something hacked or "forced" the drivers upon the 9700 Pro to show that the 9800 increase in performance was due to the R350 paths..

I'm thinking that the NV35 will be exactly the same with the 50.xx drivers. What I wondering is if the 50.xx drivers will have the same effect on the NV30.

I don't know about you guys, but I'm more interested in these 50.xx drivers then the actual NV35. Because we all know the NV35 is what the NV30 should of been in the first place. So for me, there really isn't anything to get excited about. IMO.

We all know what the NV30 is supposed to do, the NV35 finally does it. Heh.

Regards,
D. Solomon Jr.
*********.com


You know, you might be right here, I am really begining to wonder if I chose the right card now.