Legit Reviews

EVGA, Palit and XFX GeForce 9600 GT Video Card Review

The NVIDIA GeForce 9 series of graphics cards has finally arrived! Today, NVIDIA is announcing the GeForce 9600 GT video card and while it may not be as anticipated as previous generations, it still marks a significant turning point. Usually, a new series is kicked off by the fastest card in the bunch, but this time around NVIDIA is launching the GeForce series at the heart of the discrete graphics market, the $169-189 price range.

The GeForce 9600 GT is an PCI Express 2.0 graphics card that is powered by the G94 core which can be seen above. The core is labeled G94-300-A1 and was made the first week of 2008. The DirectX 10 ready core is built on the 65nm process and features 64 stream processors. The GeForce 9600 GT will be above the 8600 GTS and below the 8800 GT in terms of both price and performace. The GeForce 9600 GT also has new enhancements to NVIDIA PureVideo HD making it ideal for use in home theater applications. The recommended minimum power supply for GeForce 9600 GT graphics cards is 400 watts with a +12V rating of 26 Amps for a single card, so make sure your system power supply can handle it...

Article Title: EVGA, Palit and XFX GeForce 9600 GT Video Card Review
Article URL: http://legitreviews.com/article/666/1/

nice review! i love that frog/man/futuristic thinggy on the palit card! that alone would make me buyit! really though, although the hsf is awesome, the color is terrible! might a looked good on like some old lanparty board or some super obnoxious case with bright green uv lights

Nice review ! Just one thing temp graph legend mixed around load/idle temps http://www.legitreviews.com/article/666/14/ ;)

does the palit one have a built in audio decoder or something? why does it have a audio output? and lastly, can it output both sound and video through the hdmi port? cause if it can, itll be a great all in one hometheater pc solution

nice review i am amazed that their midrange cards are performing so close to their high end cards of yester-month(its not really yester-year)

it looks like nvidia will yet again be the powerhouse this year

i also did not know that lightsmark works THAT much better on nvidia cards it makes the 3870X2 look like a bottom end card

martini161 wrote:does the palit one have a built in audio decoder or something? why does it have a audio output? and lastly, can it output both sound and video through the hdmi port? cause if it can, itll be a great all in one hometheater pc solution

not 100% sure... On the Palit box it says "2x Dual Link DVI, Display Port, HDMI, Optical SPIDF (For HDMI Audio)". We got these cards at the last minute, so didn't get a good chance to play around with all the HTPC connections. I just sent an e-mail to Palit to see if they can clear that up for us and got this response...

AMD put pass through via PCIe out of the chip through HDMI. Nvidia does not have the pass-through via PCIe. However the signals combine for HDMI and DisplayPort using the cable from the sound card to the graphics card. This is not true for DVI so we put the audio pass through out of the board so both connections are at the same location. No need to tie up the sound card port, just the one off the back of our card.

That was good review of what we have to look forward to in the budget arena.

Another great review as usual Nate!

Review 666!!!!

It's a sign, no-one buy these cards or bad things will happen! lol

Great review as ever nate. Nice to get a full suite of benchmarks.

Don't know if you want to add this to the display port section in the review, but it was designed to take over from DVI, so we should be seeing lots of it if things go to plan.

Dan

The whole review was great, but i feel that LIGHTMARK 2007 test used by legit was only
to letdown ATI HD3800 Series cards, and what was the need of showing the results when ATI
HD3870 & Others fails to post good frames, I wonder why don't they test ant HD2900/2600 Card
which could have possibly rendered cube maps smoothly as these were based on Dx 10.
In my opinion, HQV Benchmark would have been a better option over lightmark..

welcome to the forums. Lightsmark is good to show the difference between different ATI card or NVIDIA cards. It shouldn't be used to compare ATI to NVIDIA in terms of raw frame rates. I like Lightsmark for the single reason that ATI and NVIDIA don't optimize for it and the maker is not taking money from either company. All of the other benchmarks (HQV Benchmark included) are being optimized for.

HQV Benchmark serves a purpose, but on higher end cards like these they all score 100... The test is also subjective to the tester as you get to pick 0-10 in various areas as I'm sure you know about.

Yaa u people really give a superb review for almost every product. I enjoyed a lot from 9600GT review and also from privious 8800GT one. All tests were best of their kind and were choosin for the all upcomed GPUs in the market. I want to add sum words reguarding the lightsmark test for which the Muzu_x2 have expressed his thoughts. I think he havn't seen the privious tests of the GPUs likely the 8800GT. which included the ATI models 26 and 29's in comparision.And where the light mark test show even worst results for ATI. He must accept dat ATI dx10 and dx 10.1 card hav nt been successful in the 2007- 08 starting. It is really a blunder that the Dx10.1 cards like 3800's hav nt shown any huge scores in lightmark where its compititors (NVIDIA) hav ruled. If we see the comparision between the ATI's 38's and NV's 8800(G92)s. the 3800's must won all the scores but the results are upside down.It is really a huge blow for ATI cards.n reguarding testing of 26 and 29s in review,according to me a card which is dx10.1 must hold good even on dx9 and dx10 also.there is no question where a card which is dx10.1 will have per4mence issues on dx10 benchs. and thanks for Legit once again dat the site hav compared 2 GPU companies at their Cores. Hope they will Show the result to the public in future also,which are without interference of money as mostly all benchmarks do only for optimization of a company,nowadays...Thanks for picking up lightmark bench.

Apoptosis wrote:welcome to the forums. Lightsmark is good to show the difference between different ATI card or NVIDIA cards. It shouldn't be used to compare ATI to NVIDIA in terms of raw frame rates. I like Lightsmark for the single reason that ATI and NVIDIA don't optimize for it and the maker is not taking money from either company. All of the other benchmarks (HQV Benchmark included) are being optimized for.

HQV Benchmark serves a purpose, but on higher end cards like these they all score 100... The test is also subjective to the tester as you get to pick 0-10 in various areas as I'm sure you know about.

@ apoptosis

oh thanks! i like this forum & ofcourse the site. This site forced me to include it among
top five tech sites in my Bookmarks..Anyway, comeback to review

I think the failure of HD3800 Series & even HD2000 Series could be because
of a driver issue. As ATI said, they don't know about Lightmark & how was it built, possibly
means that Catalyst drivers are almost blind to recognize the lightmark..Let you clear me guys!
am i right? Is this all becoz of Driver issue?

Furthermore, i'd like to request Regit team to have an article on " Image & Video Quality
Showdown" to review ATI & NVIDIA's Image & Video Quality of their latest GPU's. This will help us to know how
both of these fare each other in this test & surely would be the best review..

NV_AG wrote:Yaa u people really give a superb review for almost every product. I enjoyed a lot from 9600GT review and also from privious 8800GT one. All tests were best of their kind and were choosin for the all upcomed GPUs in the market. I want to add sum words reguarding the lightsmark test for which the Muzu_x2 have expressed his thoughts. I think he havn't seen the privious tests of the GPUs likely the 8800GT. which included the ATI models 26 and 29's in comparision.And where the light mark test show even worst results for ATI. He must accept dat ATI dx10 and dx 10.1 card hav nt been successful in the 2007- 08 starting. It is really a blunder that the Dx10.1 cards like 3800's hav nt shown any huge scores in lightmark where its compititors (NVIDIA) hav ruled. If we see the comparision between the ATI's 38's and NV's 8800(G92)s. the 3800's must won all the scores but the results are upside down.It is really a huge blow for ATI cards.n reguarding testing of 26 and 29s in review,according to me a card which is dx10.1 must hold good even on dx9 and dx10 also.there is no question where a card which is dx10.1 will have per4mence issues on dx10 benchs. and thanks for Legit once again dat the site hav compared 2 GPU companies at their Cores. Hope they will Show the result to the public in future also,which are without interference of money as mostly all benchmarks do only for optimization of a company,nowadays...Thanks for picking up lightmark bench.

@ NV_AG

I agree with you, ATI HD3800/2900 series have been badly beaten by G80 & G90 series.. but i'd like to add my views
here that R600/670 competes almost well with G80 & we saw it beats 8800GTS 640 & 320 versions.. G90 series was a big
blow to R600/670 & i believe it makes no sense to compare G90 with R600/670 becoz the competition
has been going around past as

G40 vs R200 = ATI + NVIDIA

G50 vs R300 = ATI

G60 vs R400 = NVIDIA

G70 vs R500 = ATI + NVIDIA

G80 vs R600 = NVIDIA

G90 vs R700 = ????

it makes sense to compare G90 with R700.. i'm right or wrong?

Ok i m impressed dat the Muzu_x2 has reviewed the history to us.But i m nt conserned wid dat. i was talking about the compitition for dx10 cards by nvidia and dx10.1 cards of ATI. reguardless of the fact that they have different core generations.ok let me ask a question. if ATI has a dx10.1 card which has advancement for DX10.1 features then y did it failed in compiting wid the cores of nvidia which are still in dx10. so, a layman can understnd that the ATI cards only perform well only wid their homemade apps.That is nt good news for the company.Accordin to me , A Graphs card must perform well in almost all apps.reguardless of the apps. And also, if muzu_x2 thinks that these results r due 2 immature drivers of the company, then it is nt headache of the customers. Wenever ATI releases a card the drivers r immature for long time and during dat period the nvidia comes up wid a new monstor card.and if u still think that lightmark is in favour of nv drivers then u must once again c the review and wat the developer has to say about dat.. And lastly. reguardin tests of image quality and others. I think administrator hav told in reply dat all new cards perform 100% in dem.so, no big difference. And last i must also add that the technology which u want to compare in tests is nt ATI's own 1(AVIVI) but for nvidia(purevideo) is their own technology.Hope u no dat..

I can tell you now that the LR team are snowed under with new hardware to review, so if there is time to do an image quality review, it'd have to be in the near future (not being on the reviewing team, I can't even promise one could be done).

NV_AG: Can you make a bit more effort to spell out your words, it makes reading your posts that much easier, and also helps people to comprehend them whose first language isn't english. Thanks!

And lastly. reguardin tests of image quality and others. I think administrator hav told in reply dat all new cards perform 100% in dem.so, no big difference. And last i must also add that the technology which u want to compare in tests is nt ATI's own 1(AVIVI) but for nvidia(purevideo) is their own technology.

I'm sure muzux does know that, he probably wants to see which one performs better at things like antialiasing and antistropic filtering. It used to be the rule that ATI had better image quality, even if nvidia did have slightly better performance, but I think that in current gen nvidia cards they have managed to bring the image quality up to ATI's standard. I agree with muzux that it would be nice to know which does better, but it isn't going to come from LR any time soon.

@muzux2: While it might be fairer to compare the G90 to the next (unreleased) ati card, in reality it depends on what is available to buy as to what you compare it with. THat is what this is all about after all, what is the best card for a price people can buy.

Dan

DMB2000uk Thanks for advice. I will try my best to make it as complete plane english. reguarding comparision of lightmark results for both the graphics companies according to u who have better hold on the lights enhancement given by this private software bench, As both companies are not optimised for this bench.And both havn't drivers very well optimised of this type of next generation benchmark. Does it mean that nvidia have utilised the unified processors which was not case with the ATI when the technology was revealed. Correct me,i might be wrong.

Do you realise that image quality wouldn't just be about that one benchmark (lightsmark), but would take samples (probably screenshots) of games/benchmarks and look close at them as to how they display the more complicated things. Here is one such example review: http://www.firingsquad.com/hardware/ati ... efault.asp

Re image quality in Lightsmark, I don't know what either company has done with the drivers so can't comment. It being unoptimised, it might be unfair to compare image quality in that benchmark, as Nvidia might have had some time to optimise drivers when ATI hasn't. The author of lightsmark is aware that there is a performance deficit with current ATI cards and is working on fixing it.

I don't know what all that adds up to though

Dan

NV_AG wrote:Ok i m impressed dat the Muzu_x2 has reviewed the history to us.But i m nt conserned wid dat. i was talking about the compitition for dx10 cards by nvidia and dx10.1 cards of ATI. reguardless of the fact that they have different core generations.ok let me ask a question. if ATI has a dx10.1 card which has advancement for DX10.1 features then y did it failed in compiting wid the cores of nvidia which are still in dx10. so, a layman can understnd that the ATI cards only perform well only wid their homemade apps.That is nt good news for the company.Accordin to me , A Graphs card must perform well in almost all apps.reguardless of the apps. And also, if muzu_x2 thinks that these results r due 2 immature drivers of the company, then it is nt headache of the customers. Wenever ATI releases a card the drivers r immature for long time and during dat period the nvidia comes up wid a new monstor card.and if u still think that lightmark is in favour of nv drivers then u must once again c the review and wat the developer has to say about dat.. And lastly. reguardin tests of image quality and others. I think administrator hav told in reply dat all new cards perform 100% in dem.so, no big difference. And last i must also add that the technology which u want to compare in tests is nt ATI's own 1(AVIVI) but for nvidia(purevideo) is their own technology.Hope u no dat..

Dx10 vs Dx10.1 ? you mean to say there should be performance advantage of Dx10.1 over Dx10..
but i don't think so, Dx 10.1 adds only new shading & AA Improvements as i hope u know it.
means it adds only Visual improvement or can say image quality. dat is why i request legit to
review image quality of both..
well,AVIVO is ATI's own technology but not UVD..u may also know then SLI isn't NVIDIA's own
technology but Crossfire is ATI's own..

DMB2000uk wrote:Do you realise that image quality wouldn't just be about that one benchmark (lightsmark), but would take samples (probably screenshots) of games/benchmarks and look close at them as to how they display the more complicated things. Here is one such example review: http://www.firingsquad.com/hardware/ati ... efault.asp

Re image quality in Lightsmark, I don't know what either company has done with the drivers so can't comment. It being unoptimised, it might be unfair to compare image quality in that benchmark, as Nvidia might have had some time to optimise drivers when ATI hasn't. The author of lightsmark is aware that there is a performance deficit with current ATI cards and is working on fixing it.

I don't know what all that adds up to though

Dan

i think u have taken it in the second direction.Let me clarify. I was asking about the fps results.Not wether the benchmark showed image quality or not.Ok i don't think u understood what i want to convey here.Thanks for replys. :mrgreen:Lastly i saw the firingsqurd reviews and conclusion was a draw between the two and that too in late 06.