FWIW X1300 benchmark scores, maybe?

maybe I missed something, but where in those pictures does it directly read 'x1300'? there are two with 'Radeon X' but the rest is cut off, and the one showing support for SM 3.0 doesn't show any details...
 
Last edited by a moderator:
Graham said:
maybe I missed something, but where in those pictures does it directly read 'x1300'? there are two with 'Radeon X' but the rest is cut off, and the one showing support for SM 3.0 doesn't show any details...

The chip has RV5xx written on it and the board looks low-end so it has to be the RV515.
 
aaronspink said:
If its real... Thats not bad for the low end. Extrapolating, a X1800 in the range of 8-9K seems very reasonable.

I calculated an effiency improvement of around 30% that would put the x1800xt in the 9-10K range.
 
It would be nice to know how many VS pipelines there are.

And, also, does 3DMk report the number of pipes? I'm wondering if it reports RV515 as 4 pipes, or as 8 pipes (4 shader, 4 texture).

Jawed
 
Jawed said:
It would be nice to know how many VS pipelines there are.

And, also, does 3DMk report the number of pipes? I'm wondering if it reports RV515 as 4 pipes, or as 8 pipes (4 shader, 4 texture).

Jawed

Even if it did, it certainly wouldn't know how to count them the "Jawed" way. ;)
 
Ailuros said:
The X600XT (500/370MHz) gets 2082 points here:

http://www.xbitlabs.com/articles/video/display/2005-17gpu_39.html

The X600 are in a relative sense 9600 derivatives with PCIe and thus single quad/2VS. Add a couple of more VS units and 50MHz more clockspeed and the almost 2900 points sound pretty reasonable.

That's a 39% increase. It would have to come from somewhere other than a couple VS units and 50MHz. (not saying it's not an X1300).
 
3dmark05 is vastly vertex setup limited; if the VS unit amount theory has any merit then it's X1300PRO@550 MVertices/s vs. X600XT@250MVertices/s. Not the only factor obviously if that should be true, but a quite significant difference in VS power.
 
Perhaps someone could correlate based on both 3DMk03 and 05, triangulate them as it were :LOL:

The 03 score is 6375, it seems.

Jawed
 
Jawed said:
Perhaps someone could correlate based on both 3DMk03 and 05, triangulate them as it were :LOL:

The 03 score is 6375, it seems.

Jawed

From digit-life:

x600xt - 4224
9800pro - 5885
6600 - 4951
6800le - 7289
6600gt - 8118
 
Man/Blimey, how old is the 9800P that a low-end card is outpacing it? Maybe I'm being unusually thick, but it seems it's been a while since that's happened.

Or maybe I'm not comparing the two on equal CPUs.
 
That 9800Pro score is a bit low - mine was doing around 6050-6100 before the fan died. While I was typing an email, I hasten to add - nothing to do with 3DMk.

Anyway there's a fairly solid looking 35-40% per-pipe and per-clock speed-up there, as far as I can tell.

Hey, all we need now is the transistor count to make our lives complete :p

Jawed
 
I was counting the seconds till someone brought up the FX series. :D I was thinking more along the lines of GF3 -> GF4--i.e., same feature set--but they had the MX in there to confuse things. For instance, I recall the MX 440 being slightly slower than the Ti200, yet it also has a lower featureset, which is not the case with the X1300 and 9800. I guess I can't draw any simple comparisons here.

/me thinks all the way back to Voodoo 2 SLI to V3 2000 ;)
 
Last edited by a moderator:
serenity said:
If these scores are true, then it is truly exciting in the sense that performance has doubled atleast in the budget category.
Only if the product goes on sale for $60 or so. I'd say that's highly doubtful, as ATI will probably want to keep selling as much of their current stock as possible.
 
Dang if this is true well then image how decently the RV530 is gonna preform after all it has supposedly 3 times as many pixels pipes.
 
Back
Top