2600XT benched

[Cookie Monster]

Link

Disappointed. Look at the FEAR score

10x7
Max settings
4xAA note sure about the level of AF enabled

Min:10
Ave:24
Max: 49

12x10

Min:6
Ave:16
Max: 85

 

[chrismr]

Why does the text for the settings in f.e.a.r look so crap compared to rest of image? I dunno, I am personally not putting any stock in this "review".

Also, are they really using ddr2 for these things?

 

[Bateluer]

Originally posted by: chrismr


Also, are they really using ddr2 for these things?

The system RAM is DDR2, according to CPU-Z. The 2600s come equipped with GDDR3.

The HD2400s will likely have some SKUs that have DDR2 to cut costs though.

 

[chrismr]

Originally posted by: Bateluer

Originally posted by: chrismr


Also, are they really using ddr2 for these things?

The system RAM is DDR2, according to CPU-Z. The 2600s come equipped with GDDR3.

The HD2400s will likely have some SKUs that have DDR2 to cut costs though.

Oh yeah.... oops. My bad.

Wonder if they will come with 512mb versions?

 

[soybeast]

Oh wow, they have a Chinese COH!

As for the text, I think that might be a consequence of forcing Chinese as the language for non-unicode programs. I've noticed when I've changed that setting in Windows, some font styles are changed in certain programs like Firefox.

I wonder though how loud that cooler is :-o

 

[Fenixgoon]

Any idea if they will be doing other benches? How does that FEAR benchmark stack up to other cards?

 

[ShadowOfMyself]

I will take a wild guess and predict it will shine in the same games the 2900 does and underperform in the same games as well, and if that happens its 100% proof its driver related

 

[Chadder007]

So how does it compare to the 8600GT??

 

[Lonyo]

Originally posted by: ShadowOfMyself
I will take a wild guess and predict it will shine in the same games the 2900 does and underperform in the same games as well, and if that happens its 100% proof its driver related

How do you work that out?
They are based on the same architecture. If the arch is weak at something, it's possible that coul dbe seen across cards.

 

[ShadowOfMyself]

Originally posted by: Lonyo

Originally posted by: ShadowOfMyself
I will take a wild guess and predict it will shine in the same games the 2900 does and underperform in the same games as well, and if that happens its 100% proof its driver related

How do you work that out?
They are based on the same architecture. If the arch is weak at something, it's possible that coul dbe seen across cards.

Yeah I thought about that too... Humm

Well... I cant explain it lol /fail

 

[Cookie Monster]

Well, the biggest weakpoint of the 2600XT is its 4 TMUs. Yes 4!, although it has 8 ROPs and 120 shaders, it has a weak texturing performance. So im kind of skeptical to believe that this card might out perform 8600GTS, but rather perform on par.. or worse perform around the GT level.

Although the R600 uses a 512bit bus, the 2600XT uses 128bit bus so bandwidth isnt enormous even with GDDR4.

 

[blckgrffn]

Originally posted by: Cookie Monster
Well, the biggest weakpoint of the 2600XT is its 4 TMUs. Yes 4!, although it has 8 ROPs and 120 shaders, it has a weak texturing performance. So im kind of skeptical to believe that this card might out perform 8600GTS, but rather perform on par.. or worse perform around the GT level.

Although the R600 uses a 512bit bus, the 2600XT uses 128bit bus so bandwidth isnt enormous even with GDDR4.

But like you said, it isn't likely that bandwidth is the issue, rather the texturing. Jeeze, would it kill ATI to just come out with a decent midrange product right of the bat? The X1650XT is a great example of this - and will likely out-perform the 2600XT at this rate. Now if only ATI had been able to really make the X1650XT instead of cutting down an X1950... (I am pretty sure that is what happened, anyway).

Nat

 

[Cookie Monster]

X1650XT was good in the sense that it outperformed its competition, didn't require a 256bit bus nor a complex PCB like the X1800GTO. However it was late. Actually the X1650XT was a cut down version of X1950pro (RV570) because the RV560 didn't quite go to plan unless it was already planned out to be a RV570 with a quad disabled.

When was the release date again?

 

[Fenixgoon]

Originally posted by: blckgrffn

Originally posted by: Cookie Monster
Well, the biggest weakpoint of the 2600XT is its 4 TMUs. Yes 4!, although it has 8 ROPs and 120 shaders, it has a weak texturing performance. So im kind of skeptical to believe that this card might out perform 8600GTS, but rather perform on par.. or worse perform around the GT level.

Although the R600 uses a 512bit bus, the 2600XT uses 128bit bus so bandwidth isnt enormous even with GDDR4.

But like you said, it isn't likely that bandwidth is the issue, rather the texturing. Jeeze, would it kill ATI to just come out with a decent midrange product right of the bat? The X1650XT is a great example of this - and will likely out-perform the 2600XT at this rate. Now if only ATI had been able to really make the X1650XT instead of cutting down an X1950... (I am pretty sure that is what happened, anyway).

Nat

i wish that had cut down the X19/X18 series to begin with, instead of the joke of a midrange X1600's that we got for the most part. the 1650's were way too late in the game, IMO.

 

[blckgrffn]

Originally posted by: Cookie Monster
X1650XT was good in the sense that it outperformed its competition, didn't require a 256bit bus nor a complex PCB like the X1800GTO. However it was late. Actually the X1650XT was a cut down version of X1950pro (RV570) because the RV560 didn't quite go to plan unless it was already planned out to be a RV570 with a quad disabled.

When was the release date again?

No, I think it was meant to be it's own PCB, they just didn't make it happen. Quite frankly, that should have been released about the time of the 7600GT. That would have saved ATI the entire $90 - $150 market, which is pretty huge.

Originally posted by: Fenixgoon


i wish that had cut down the X19 series to begin with, instead of the joke of a midrange X1600's that we got for the most part. the 1650's were way too late in the game, IMO.

I totally agree. What a waste of time. All it needed more TMU's (whatever it had 4 of) all along, an 8/16 part would have been great. The X1300 was already a 4/4 part, the 4/12 configuration didn't really seem to offer any value. Cutting down the X1900's sucked in its own right, too much power and not enough performance, not to mention that it really didn't help ATI's bottom line much.

Nat

 

[aznrice54]

Why...why does this generation's midrange have to SUCK so much...
8600s and 2600s both

 

[lopri]

Originally posted by: Chadder007
So how does it compare to the 8600GT??

Doesn't look very good to me. Sigh.. I don't understand AMD.. And is that thing dual-slot? Please don't tell me so.

 

[Zstream]

Originally posted by: Cookie Monster
Link

Disappointed. Look at the FEAR score

10x7
Max settings
4xAA note sure about the level of AF enabled

Min:10
Ave:24
Max: 49

12x10

Min:6
Ave:16
Max: 85

8600 numbers from Anandtech, 2600 numbers from Coolaler forum's, machines:
Intel Core 2 Extreme X6800 (2.93GHz/4MB)
P965 (nv6800) vs 680i SLI (Anandtech)
DDR2-HK-800 4x512MB with 5-5-5-18 timings (nv6800) vs Corsair 2x1GB DDR2 with 4-4-4-12 timings (Anandtech)
Results
Company of Heroes, 1024x768 scores:

Geforce 8600GTS: 45 FPS
Radeon HD 2600XT: 49.6 FPS

1280x1024 scores:

Geforce 8600GTS: 34.3 FPS
Radeon HD 2600XT: 39 FPS

 

[soybeast]

collusion ftw! (maybe?)

 

[Extelleron]

Originally posted by: Cookie Monster
Well, the biggest weakpoint of the 2600XT is its 4 TMUs. Yes 4!, although it has 8 ROPs and 120 shaders, it has a weak texturing performance. So im kind of skeptical to believe that this card might out perform 8600GTS, but rather perform on par.. or worse perform around the GT level.

Although the R600 uses a 512bit bus, the 2600XT uses 128bit bus so bandwidth isnt enormous even with GDDR4.

I don't know where you're getting that number from, but the 2600XT has 8 TMUs.

 

[rmed64]

yea, isnt it 8 TMU and 4 ROP?

 

[lopri]

Not sure how much that guy OC'ed the vid card, but I get a lot higher 3DMark05 score with my overclocked 8600 GT. (default setting 1024x768)

 

[Cookie Monster]

Originally posted by: rmed64
yea, isnt it 8 TMU and 4 ROP?

I think your right. Heh, i got it mixed up but 4 ROPs? Maybe texturing performance isnt that bad after all but the fillrate looks bad now.