Radeon 9500PRO or nVidia 4600Ti

[Ben88]

I would get a ti4600. Looks like i'm the only one.

 

[Ben88]

I just looked at some prices and for a good ti4600 you are looking at around $230 bucks which is $30 more than the 9500pro for around the same performance. If they were the same price I would go for the ti4600 but the cheapest ti4600 I found was $210 for an EVGA and you might not get very good 2d with that brand.

 

[EglsFly]

Seems to me, after reading more of this review. That the 4600 Ti is faster than the 9500 PRO in games when not using any AA.

Since I rarely use AA, in favor of faster FPS...
I am going to go ahead and get the 4600 Ti.

I can pick one up for $215. (free shipping)
I know it will be a rock solid performer, and not have any weird bugs or quirks.

 

[Mustanggt]

I would get the TI 4200 overclock it to 4600 speeds And save $100 bucks. the 9500 is very nice but I will never pay $200 for vid card, The prices fall on them to quickly.

 

[Rand]

I wonder if its thier usual crappy drivers, or is it the hardware?

Personally I'm leaning towards it having nothing to do with the 9700Pro at all... it seems to be there is something VERY wrong with THG's testing. He has the Ti4600 performing at well over 2X the minimum frame rate of the Ti4400 in every istance, and in some cases managing nearly 3X the frame rate.

Definitely seems like something went wrong with the Ti4600 tests, as there is absolutely no way the Ti4600 could even theoretically outperform the Ti4400 by that much.
Even assuming it were 100% bandwdith limited and 100% bandwidth effective the Ti4600 would still only peak at a meager 18% faster.... an awfully far cry from the consistent 200%+ faster scores THG is claiming. Not to mention it's quite clear that in the real world, the difference would never hit 18% at the GF4 most definitely isnt 100% bandwidth efficient.

I'd ignore the Ti4600 results entirely, as it's quite clear they are completely impossible relative to the rest of the GF4 line. Not to mention the results are at considerable odds with what other websites show.

 

[Bovinicus]

The Ti4600 will lag behind if you like anisotropic filtering as well. I will no longer play a game without anisotropic filtering. It makes a large difference in enahncing the realism of a game.

 

[mamisano]

Just Installed a 9500 Pro last night.

Before with Albatron Ti4200P-Turbo @ 291/635 - 12123 3DMarks
After with ATI Radeon 9500 Pro @ 270/275 - 11881 3DMarks

The average Ti4600 on 3DMark Results page is around 11700.

So, I lose 300 3DMarks going from Overclocked Ti4200 to 9500 Pro, gain Much better AA/AF Performace, DX9 Support, Better Graphics. The card cost me $197 shipped, the Albatron is being RMAd to NewEgg for a $180 refund. Worth it to me for the $17.

I just hope there is a way to OC this card!

My .02

 

[nardvark]

would you mind running a more "real-life performance" benchmark for us? (UT2003 perhaps?)
I know there are those on this site that consider 3DMark to be pretty much worthless.
Or better yet, play games for 12 hours and tell us your impressions (everybody wins in this case, especially you)
Plus, this will be the only way to see if the card shows the minimum frame rate problems being discussed.
enjoy the new card! (I'm jealous)

 

[mamisano]

Originally posted by: nardvark
would you mind running a more "real-life performance" benchmark for us? (UT2003 perhaps?)
I know there are those on this site that consider 3DMark to be pretty much worthless.
Or better yet, play games for 12 hours and tell us your impressions (everybody wins in this case, especially you)
Plus, this will be the only way to see if the card shows the minimum frame rate problems being discussed.
enjoy the new card! (I'm jealous)

I did run HardOCP's Ut2003 Test Suite. At High-Detail, 1024x768 the R9500 Pro was faster than the Albatron (291/635) in 5 out of 7 tests by an average of AT LEAST 20fps and the other two tests were higher by about 1fps.

Don't have the #'s in front of me, I will post when I get home this evening.

 

[overclocksomemore]

Many of yous seem to depend on DirectX 9 way way way way too much. Are you gonna ignore DX8 based games? Are you gonna trash all the present and older games and move on to DX9 optimized games just for R9500pro??? that's pretty stupid.

GF4 ti4600 is faster than R9500 in many games including UT2003. UT2003 is DX8 optimized. Using the DX9 will not boost the performance for you, instead, it will de-celerate it for you. Many many many people play UT2003 and it's gonna last pretty damn long. Go with the GF4 ti4600.

 

[Sxotty]

direct x 9 is useless right now and will be for quite awhile. From my experience I have found it pointless to worry about support for the next directx incarnation, b/c by the time games use it regularly you will have a new video card anyway. The 9500 will go down in price I think shortly as well, then it would be good b/c of AA results.

Also UT2K3 is what everyone is talking about, but past Epic game engines have not scaled well anyway so in fact it could make the 9700pro look worse and not better than a ti4600.

(I am comparing how big a difference vidcards and processors made on say Q3 vs. UT) <=if you saw the numbers it was kind of strange that the unreal engine never really got way faster.

 

[mamisano]

Originally posted by: overclocksomemore
Many of yous seem to depend on DirectX 9 way way way way too much. Are you gonna ignore DX8 based games? Are you gonna trash all the present and older games and move on to DX9 optimized games just for R9500pro??? that's pretty stupid.

GF4 ti4600 is faster than R9500 in many games including UT2003. UT2003 is DX8 optimized. Using the DX9 will not boost the performance for you, instead, it will de-celerate it for you. Many many many people play UT2003 and it's gonna last pretty damn long. Go with the GF4 ti4600.

With High Detail HardOCP UT2K3 Test @ 1024x768, my R9500 Pro MOPS UP my now returned Albatron ti4200P-Turbo @ 291/635. The Ti4600 is not much faster than the Overclocked Albatron. Like I said earlier, I will post my results this evening when I get home from work.

How is DX9 going to "decelerate" the game? It is backwards compatible, no?

So lets see....
The 9500 Pro is CHEAPER
The 9500 Pro is FASTER
The 9500 Pro has BETTER FEATURES

Sure, he should get a Ti4600.

 

[pillage2001]

Originally posted by: mamisano

Originally posted by: overclocksomemore
Many of yous seem to depend on DirectX 9 way way way way too much. Are you gonna ignore DX8 based games? Are you gonna trash all the present and older games and move on to DX9 optimized games just for R9500pro??? that's pretty stupid.

GF4 ti4600 is faster than R9500 in many games including UT2003. UT2003 is DX8 optimized. Using the DX9 will not boost the performance for you, instead, it will de-celerate it for you. Many many many people play UT2003 and it's gonna last pretty damn long. Go with the GF4 ti4600.

With High Detail HardOCP UT2K3 Test @ 1024x768, my R9500 Pro MOPS UP my now returned Albatron ti4200P-Turbo @ 291/635. The Ti4600 is not much faster than the Overclocked Albatron. Like I said earlier, I will post my results this evening when I get home from work.

How is DX9 going to "decelerate" the game? It is backwards compatible, no?

So lets see....
The 9500 Pro is CHEAPER
The 9500 Pro is FASTER
The 9500 Pro has BETTER FEATURES

Sure, he should get a Ti4600.

nuff said.

 

[mamisano]

#'s are closer than I remembered. A Ti4600 would score a FEW % better than the Albatron.

[/u]Overclocked Albatron ti4200P-Turbo @ 291/635[/u]
dm-antalus
0.698998 / 38.056850 / 336.901764 fps
Score = 96.873978

dom-suntemple
55.590397 / 120.845779 / 386.461151 fps
Score = 120.873360

dm-phobos2
35.408882 / 128.937057 / 477.225952 fps
Score = 128.987732

dm-inferno
37.592339 / 89.757370 / 294.773834 fps
Score = 89.865532

ctf-face3
30.281363 / 138.203644 / 481.037567 fps
Score = 138.278381

ctf-citadel
36.793030 / 122.353737 / 309.300781 fps
Score = 122.578117

dm-asbestos
82.629066 / 170.079483 / 488.801453 fps
Score = 170.221573

[/u]Stock Radeon 9500 PRO[/u]
dm-antalus
21.837744 / 131.019775 / 275.672821 fps
Score = 131.057770

dom-suntemple
19.852020 / 131.655441 / 368.113953 fps
Score = 131.687790

dm-phobos2
28.998465 / 123.705627 / 493.140625 fps
Score = 123.749176

dm-inferno
14.749454 / 98.575722 / 323.926575 fps
Score = 98.632866

ctf-face3
63.011379 / 138.748474 / 489.287323 fps
Score = 138.820541

ctf-citadel
45.275078 / 122.117668 / 313.953644 fps
Score = 122.306351

dm-asbestos
85.706772 / 176.254028 / 456.637695 fps
Score = 176.404587

 

[overclocksomemore]

oh my young one... compare it with the Ti4600. What did the topic say?... Ti4600.

That's what I call, biased bench/comparison.

Also, we can't trust you exactly what you did in the graphics setting before you run the bench..

Anybody can make a biased bench and make one deliberately better than the other.. what a bull **** that is.. ho ho..


Take a default Ti4600 and default R9500pro with same graphics setting for the bench.

 

[mamisano]

Originally posted by: overclocksomemore
oh my young one... compare it with the Ti4600. What did the topic say?... Ti4600.

That's what I call, biased bench/comparison.

Also, we can't trust you exactly what you did in the graphics setting before you run the bench..

Anybody can make a biased bench and make one deliberately better than the other.. what a bull **** that is.. ho ho..


Take a default Ti4600 and default R9500pro with same graphics setting for the bench.

Ok Junior....They have the SAME settings. SHEESH! Just because it doesn't jive with what you want...don't slam it.