The Official RV630/RV610 Rumours & Speculation Thread

[Jawed]

I'd note that the dual-slot version and this one we're seeing now look to have the same (or very similar) PCBs, too. That's somewhat telling.

Yeah, this is why I assumed the board was below-XT spec. If it's merely a 256MB GDDR4 board, then that's much closer to the "top" than expected.

Perhaps the 512MB GDDR4 version is higher-clocked and has a dual-slot cooler. Or maybe the idea is to allow the AIBs to offer "overclocking" versions, so the PEG power socket will come into play.

Jawed

 

[vertex_shader]

I realised just now those flames are supposed to be ice expanding across the cooler guess the graphical concept needs a rethink.

Hmm, the power is down, but 80W would still require extra power, wouldn't it? Unless the person saying 80W is guesstimating from measurements at the wall or something.

Looking at the picture, there's a capacitor in the top right corner, right on top of the set of 6 holes for the PEG power connector. So, that board definitely doesn't have extra power.

I assumed this was a lower spec board when I examined it last night. If it's GDDR4, then maybe not. Is the 512MB GDDR4 version the only one requiring extra power?

Jawed

80 watt not possible without extra power connector, according the rumors all rx6xx card pci-e 2.0 competible this can be used for 150watt power, but no pci-e 2.0 motherboards released yet, and when its use more than 75watt and pci-e 2.0 compatible still need extra power connector for the old motheboards, so i think the hd2600xt 256mb gddr4 consume less than 75 watt.
When its consume close to 75watt than OC can be limited do the lack of enough power, so i think AIB's will sell the card with extra 6pin connector.

Possible 512mb version need extra psu connector when the 256mb version power usage close to 75watt, old rumors said 512mb version consume +7watt as the 256mb, but this rumors said the 121/128watt TDP too :smile:

 

[Sound_Card]

I really hope that ATi does not stick 512mb of ram on a 128bit interface. What a waste. Just like the x1600pro 512mb. I hope that dual slot cooler is a 256bit part.

 

[Sound_Card]

Don't know if this was posted yet but looks good to me.

http://www.fx57.net/?p=581

- 3DMark 2005: 12500
- 3DMark 2006: 5600


The only problem I have with these scores is that they dont state their system spec. For all I know they could be using a quad core clocked at 3+ghz which in the end, means nothing to me.

 

[jamis]

So... 2105 in 3dmark's SM3.0 test. That doesn't indicate better performance than 8600gts. :/

Screw 3DMark, unlike 8600GTS it might actually deliver it at games! Look at the NFS Carbon test:
http://www.hardspell.com/english/doc/showcont.asp?news_id=414&pageid=517

51@1280X1024

G8600GTS = 33,9
X1950Pro = 45,1

Maybe it's somehow more balanced architecture and gives more steady performance than G8600GTS? So unlike with it's nvidia counterpart that 3DMark number might actually somewhat reflect the real life prformance. We need more benchmarks!

Edit: Are there any Infernal benchies with 8600GTS? Couldn't find one.

 

[INKster]

Screw 3DMark, unlike 8600GTS it might actually deliver it at games! Look at the NFS Carbon test:
http://www.hardspell.com/english/doc/showcont.asp?news_id=414&pageid=517

51@1280X1024

G8600GTS = 33,9
X1950Pro = 45,1

Maybe it's somehow more balanced architecture and gives more steady performance than G8600GTS? So unlike with it's nvidia counterpart that 3DMark number might actually somewhat reflect the real life prformance. We need more benchmarks!

Edit: Are there any Infernal benchies with 8600GTS? Couldn't find one.

Let's see if this can compete with a 8800 GS.

 

[jamis]

Let's see if this can compete with a 8800 GS.

Why should it compete with 8800GS? That's for RV660/670 or whatever coming sometime H2.

 

[Jawed]

51@1280X1024

A single FRAPs number for an instantaneous frame rate like that is even worse than a single 3DMk number.

Jawed

 

[vertex_shader]

Screw 3DMark, unlike 8600GTS it might actually deliver it at games! Look at the NFS Carbon test:
http://www.hardspell.com/english/doc/showcont.asp?news_id=414&pageid=517

51@1280X1024

G8600GTS = 33,9
X1950Pro = 45,1

Maybe it's somehow more balanced architecture and gives more steady performance than G8600GTS? So unlike with it's nvidia counterpart that 3DMark number might actually somewhat reflect the real life prformance. We need more benchmarks!

Edit: Are there any Infernal benchies with 8600GTS? Couldn't find one.

That 51frame is not avarage, its jut a single frame, check the picture.

Edit: I was late again

 

[jamis]

A single FRAPs number for an instantaneous frame rate like that is even worse than a single 3DMk number.

Jawed

I know, that's why I said we need mor bencies. However looking at that NFSC benchamrk it seems it's a game where NV parts are generally performing poorly so it prolly doesn't tell that much..

Edit: Well they are saying that the benchmark gives 51FPS so I trusted that..of course it can be BS

Edit2: Actually it's propably very much real. 8600GTS(@33,9fps) seems to perform better than 7900GTX/GS, so it would be logical that RV630(@51fps) performs better than X1950Pro. But it's a game where NV generally performs poorly, so even if real it doesn tell that much.

 

[Plano]

Edit: Well they are saying that the benchmark gives 51FPS so I trusted that..of course it can be BS

If you are looking that "Infernal benchmark" it says 88fps and in that screenshot fraps says 88fps too. Same thing with that NFS screen. Looks more like they just took one fps number and posted it.

 

[jamis]

If you are looking that "Infernal benchmark" it says 88fps and in that screenshot fraps says 88fps too. Same thing with that NFS screen. Looks more like they just took one fps number and posted it.

Nah, and as I said it doesn't really matter anyhow. Dunno about Infernal but in NFSC NV cards generally seem to perfom poorly. So looking at previous generation Radeons that 51fps seems realistic for RV630. But even if it's real it doesn't tell much as it is a benchmark where the Radeons generally do very well.

 

[Evildeus]

Don't forget that NFS is an ATI favour bench (like there's some NV favour benchies )

 

[Hanners]

A single FRAPs number for an instantaneous frame rate like that is even worse than a single 3DMk number.

Jawed

Particularly when it's taken at a menu screen, rather than in-game. Performance in Need For Speed: Carbon with motion blur turned on is always lower by quite a large degree compared to performance while rendering the game's menus.

 

[vertex_shader]

Screw 3DMark, unlike 8600GTS it might actually deliver it at games! Look at the NFS Carbon test:
http://www.hardspell.com/english/doc/showcont.asp?news_id=414&pageid=517

51@1280X1024

NFS Carbon, same picture, 1280x1024 high detail, AGP x1950pro 256mb default, e4300 default, 2x1gb ddr400 memory.
70 frame/sec
59 frame/sec with 4xaa 16xhqaf

It's just a picture, not means anything, for deep conclusion we need to wait the launch :smile:

 

[jamis]

NFS Carbon, same picture, 1280x1024 high detail, AGP x1950pro 256mb default, e4300 default, 2x1gb ddr400 memory.
70 frame/sec
59 frame/sec with 4xaa 16xhqaf

It's just a picture, not means anything, for deep conclusion we need to wait the launch :smile:

Ok, now all we need to know is how the rest of the cards run the menu screen.
It allmost made me download the NFSC demo myself but it was > 650MB.

 

[mao5]

Radeon HD 2600XT VS. GeForce 8600 GTS in NFS: Carbon

http://r800.blogspot.com/

 

[Arty]

Let's see if this can compete with a 8800 GS.

Is that admitting that RV630 > G84 ?

 

[Sound_Card]

Is that admitting that RV630 > G84 ?

Sorry to burst INKster's bubble...:smile:

http://www.guru3d.com/newsitem.php?id=5223

 

[INKster]

Sorry to burst INKster's bubble...:smile:

http://www.guru3d.com/newsitem.php?id=5223

Well, it's all over for 8800 GS, but perhaps something else is in the the cards for the future.
Traditionally, NV never gives up on launching a "just below the high-end pair derivative GPU" before Christmas.
We've had the FX5900 XT, the 6800 GS and the 7950 GT so, maybe next Fall...