R9 290 or GTX780 ??

[YBS1]

I suppose, I have both of those. Truthfully, I keep waiting to see a flood of custom/water cooled 290s assaulting the leaderboard on 3DMark, but that ain't happening either. :sneaky:
Don't worry, when/if Mantle hits...then it's on! It's the 29th now so it should be any minute now.

 

[BallaTheFeared]

Not even with tessellation disabled? :hmm:

 

[YBS1]

I think that only counts for "World Records"....

 

[BallaTheFeared]

Metro LL

"Stock"

1320/1775 (I hate you Classified)

 

[YBS1]

You're a tad faster there despite lower clocks, perhaps the Haswell advantage?
1359/1900

 

[Nintendesert]

Dang, these threads are volatile! I'll add this for the OP, if I could get an aftermarket cooler 290 say the Triwhatever at MSRP I would get that over a 780. If I couldn't and the price is closer to the 780 I'm getting a 780 and overclocking the pizzle out of it.

 

[blackened23]

Not even with tessellation disabled? :hmm:

Metro: LL would be a better benchmark as a "neutral" testing ground wouldn't it? Square Enix games favor AMD, generally speaking, while Metro LL is largely even without favoring either side. From what I've seen anyway. Also the result webpage generated from Metro :LL will show what game settings specifically were used. Which is kinda nice in these GPU showdowns.

And yeah, tessellation cheats. Must be nice to have that checkbox in the control panel to instantly increase benchmarks by 10-20%. Still can't believe HWBOT allows that nonsense with AMD cards. At this point, i'm not even sure why AMD still has this "feature" in their control panel - it was necessary with Cayman since Cayman was a piece of junk in terms of tess performance. But AMD corrected that with Tahiti and Hawaii so I don't even see why AMD sees fit to include the "disable tessellation on a whim" checkbox. Pretty silly if you ask me. But the world record guys love it since they can get 20% higher benchmark scores than otherwise possible.

 

[BallaTheFeared]

Haswell advantage?

Boy, I hope so

Metro: LL would be a better benchmark as a "neutral" testing ground wouldn't it?

It favors AMD slightly, which is the goal here to take AMDs best showings and see what happens OC vs OC.

Obviously if we take the tests were Nvidia has a clear advantage it's just going to open it up more.

Really more interested in seeing if Nvidia can use it's better OC headroom to close the gaps in games it lags at stock.

 

[ICDP]

So typical of the pro Nvidia crowd.

• It starts with "GTX780 overclocked will beat the R9 290 cards easily". They are proved wrong.
• Then it moves to "oh they are using the AMD tessellation cheat". My tests were done using a RadeonPro profile that explicitly sets tessellation to App controlled. Not that they would believe I want fair tests to compare my own GTX780 and R9 290X with.
• Or to "still waiting for the AMD guys to obliterate us", even though that was never the claim.
• Oh, Tomb Raider is a pro AMD game, even though Balla is the one who chose that game to prove GTX780 was better. I suppose he wanted to beat AMD at ther own game to prove his agenda.
• Now we have "Metro Last Light is a neutral game", even though it is TWMTBP.

No matter what anyone says, or does, or proves, the Nvidia faithful will move the goalposts to whatever metric makes their favourite brand look best.

 

[ICDP]

Sorry, was running a game profile in RadeonPro

Stock: 74.8

1150/1500: 88.3


Having trouble getting it stable at 1200/1500. I don't think my Elpida vram will let me, so I'll probably have to drop memory down a lot. I'll try later tonight, have to step out right now. Did not expect to be playing nerdball with you guys at 10AM on a Sunday

The VRAM past around 1450 barely makes a difference anyway.

 

[BertolomeoDiaz]

i don´t think ballathefeared is generally pro nvidia/anti amd
in numerous benchmark threads he used to post scores of cf 7950 until very recently or even now ?
i just think he likes the gtx 780
given the posts i so far have seen from him i also think he will fight to the death for what ever product he currently uses and is convinced of it being supreme, but i could also picture him telling people that something he tested/owned was crappy, if he thought so

 

[Pandamonia]

i don´t think ballathefeared is generally pro nvidia/anti amd
in numerous benchmark threads he used to post scores of cf 7950 until very recently or even now ?
i just think he likes the gtx 780
given the posts i so far have seen from him i also think he will fight to the death for what ever product he currently uses and is convinced of it being supreme, but i could also picture him telling people that something he tested/owned was crappy, if he thought so

He used to be in the AMD camp and complain about the CF Drivers. Then he moved to Nvidia and just as he did that the 290 came out and the drivers are fixed lol.

The 290 is better than the 780 GTX especially at MSRP. I dont see any tech website claiming otherwise.

 

[ICDP]

i don´t think ballathefeared is generally pro nvidia/anti amd
in numerous benchmark threads he used to post scores of cf 7950 until very recently or even now ?
i just think he likes the gtx 780
given the posts i so far have seen from him i also think he will fight to the death for what ever product he currently uses and is convinced of it being supreme, but i could also picture him telling people that something he tested/owned was crappy, if he thought so

There is absolutely nothing wrong preferring one brand (or GPU) over the other. I favour AMD but will not do so blindly as some do, my first post in this thread was recommending a GTX780 over the R9 290 reference if price was similar. I know Balla or blackened are not so pro Nvidia that they would never consider AMD hardware. I just don't understand their stance trying to prove GTX780 once overclocked (with warranty breaking overvolted BIOS) simply destroys R9 290/X cards.

This claim has been posted many times and despite being proved wrong on many occasions it still crops up from blackened or Balla to prove GTX780 are far better cards.

P.S.

MLL results from R9 290X 1200/1250

 

[BallaTheFeared]

So typical of the pro Nvidia crowd.

• It starts with "GTX780 overclocked will beat the R9 290 cards easily". They are proved wrong.
• Then it moves to "oh they are using the AMD tessellation cheat". My tests were done using a RadeonPro profile that explicitly sets tessellation to App controlled. Not that they would believe I want fair tests to compare my own GTX780 and R9 290X with.
• Or to "still waiting for the AMD guys to obliterate us", even though that was never the claim.
• Oh, Tomb Raider is a pro AMD game, even though Balla is the one who chose that game to prove GTX780 was better. I suppose he wanted to beat AMD at ther own game to prove his agenda.
• Now we have "Metro Last Light is a neutral game", even though it is TWMTBP.

No matter what anyone says, or does, or proves, the Nvidia faithful will move the goalposts to whatever metric makes their favourite brand look best.

I think it's a little soon to be going full defensive mode.

I never said easily, I simply said faster. I believe the 780 is overall faster than the R290, possibly even the 290x at commonly achieved clocks.

Is it faster that it matters? Probably not so much, in the same way the R290x isn't fast enough to matter vs the R290.

The banter there was directly related to 3DMark11 and score submissions by AMD users. Upon further review it is clear the 3DMark offical Hall of Fame doesn't accept user modified settings, which is why the R290 series isn't doing well - compared to places like hwbot which fully accept modified settings via control panel options such as but not limited to, tessellation.

I didn't choose TR. That was a R290 user who decided to actually post something. It's been over a month now and it's so rare to see any user results outside the few people getting golden R290 chips that do 1200+.

I look at individual games when making such a comment, compared to overall performance differences. I don't really care about TW or GE.

Take Metro, I said it favors AMD slightly.

I put stock vs stock R290 about 6-7% ahead of stock 780. So when a title such as Metro shows a 13% difference I say that title favors AMD.

On the flip side when I look at say Crysis 3, BF3, Rome 2, Bioshock from the same review I say those titles slightly favor Nvidia.

It's not like we can run a single test and proclaim a winner. What I'm looking at are gains vs stock to basically add x to y from stock to try to figure out which card is gaining more from OC and thus either producing similar results to stock or perhaps slightly tipping the balance in the other direction.

 

[Gloomy]

Balla, your FPS graphs look worrisome... there is no way that is playable at ANY framerate. Too many dips and spikes. Same goes for YBS1. You should get that fixed before comparing performance, it is a total mess.

 

[PPB]

So typical of the pro Nvidia crowd.

• It starts with "GTX780 overclocked will beat the R9 290 cards easily". They are proved wrong.
• Then it moves to "oh they are using the AMD tessellation cheat". My tests were done using a RadeonPro profile that explicitly sets tessellation to App controlled. Not that they would believe I want fair tests to compare my own GTX780 and R9 290X with.
• Or to "still waiting for the AMD guys to obliterate us", even though that was never the claim.
• Oh, Tomb Raider is a pro AMD game, even though Balla is the one who chose that game to prove GTX780 was better. I suppose he wanted to beat AMD at ther own game to prove his agenda.
• Now we have "Metro Last Light is a neutral game", even though it is TWMTBP.

No matter what anyone says, or does, or proves, the Nvidia faithful will move the goalposts to whatever metric makes their favourite brand look best.

They are not quite there yet, when they start talking about the FPS-Frametime-dB-ºC-stockvalue-marketshare-featureset metric, you know they have hit rock bottom.

 

[blackened23]

He used to be in the AMD camp and complain about the CF Drivers. Then he moved to Nvidia and just as he did that the 290 came out and the drivers are fixed lol.

The drivers aren't fixed. Crossfire/eyefinity and DX9 crossfire still suffer from microstutter on the 79xx cards. Which is what Balla owned. I should add, this bug existed for more than 2 years. Does that give you confidence in AMD's driver and software engineers?

The 290 is better than the 780 GTX especially at MSRP. I dont see any tech website claiming otherwise.

You're clearly not looking very hard then. The only reason any website recommended the 290, given the severe shortcomings of the reference design, was because it was cheap at the time. At 400$, it represented good value. Then some websites, such as this one, outright didn't recommend it based on the same reference design shortcomings despite the cheap price. Of course now the 290 isn't cheap due to mining price inflation, it is actually more expensive than the GTX 780 generally speaking - despite performing the same and being hotter and louder in reference form.

 

[Freddy1765]

So what's the bottom line? If the Sapphire 290 Tri-X is priced at the same level as a Gigabyte or Asus 780, which is worth getting, strictly for gaming purposes? There are too many conflicting opinions, impossible to make an informed decision..

 

[blackened23]

If it's at MSRP? Buy the Tri-X. If an asus DC II 290 is released at 420$ and you find it at MSRP? It's worth buying. Reference 290? Avoid at all costs. That's my take on it.

If it's higher than MSRP, I can't see many people buying it unless they're miners. The entire value proposition of the 290 is gone once the price is inflated to 100$+ above the original point.

 

[PPB]

Raise your hand who can tell sub-pixelar tesselation in your screen..




.... anyone? Oh, over there's one!!! Oh nope, just scratching his head.

Moving on.

 

[Pandamonia]

The drivers aren't fixed. Crossfire/eyefinity and DX9 crossfire still suffer from microstutter on the 79xx cards. Which is what Balla owned. I should add, this bug existed for more than 2 years. Does that give you confidence in AMD's driver and software engineers?



You're clearly not looking very hard then. The only reason any website recommended the 290, given the severe shortcomings of the reference design, was because it was cheap at the time. At 400$, it represented good value. Then some websites, such as this one, outright didn't recommend it based on the same reference design shortcomings despite the cheap price. Of course now the 290 isn't cheap due to mining price inflation, it is actually more expensive than the GTX 780 generally speaking - despite performing the same and being hotter and louder in reference form.

Except the 290 can be had in the UK for £80 less than the 780 GTX. There is no stock shortage here in the UK.

A 290 CF is £160 less. Stock cooler or not there are custom coolers out next 2 weeks which is still £40 cheaper than the 780 Reference.

Shortage is BS. Its just retailers cashing in on peoples stupidity.

 

[RussianSensation]

So what's the bottom line? If the Sapphire 290 Tri-X is priced at the same level as a Gigabyte or Asus 780, which is worth getting, strictly for gaming purposes? There are too many conflicting opinions, impossible to make an informed decision..

If you can find R9 290 Tri-X for a similar price as 780, it's the better buy imo.

It beats 780 by 18% out of the box and has another 15-20% overclocking headroom from 1Ghz, which means the 780 won't be able to match it unless you get one that overclocks to 1.3Ghz+.
http://www.computerbase.de/artikel/...-r9-290-tri-x-oc-im-test/3/#performancerating

You also get higher resale value on R9 290s right now, more VRAM for free in your case and the possibility of Mantle adding another 10-15% performance down the line.

Having said that if you play many games that favour NV (Blizzard, AC4) and care for features such as PhysX, 780 isn't a bad option either, especially since right now it comes with 3 free games.

 

[BallaTheFeared]

He used to be in the AMD camp and complain about the CF Drivers. Then he moved to Nvidia and just as he did that the 290 came out and the drivers are fixed lol.

The 290 is better than the 780 GTX especially at MSRP. I dont see any tech website claiming otherwise.

I complained about having to turn my monitor on and off, and having to disable ULPS for crossfire to function properly... Other than that not so much.

It's not like I complained about the stutter, because I was aware of it prior to the purchase assuming it would be fixed and we'd see the memory rewrite everyone was talking about.

It's really telling though, when my 7950 CF was whoopin some 670/680 b-u-t-t not one person from "Team Nvidia" moaned like a moron about bias.

Balla, your FPS graphs look worrisome... there is no way that is playable at ANY framerate. Too many dips and spikes. Same goes for YBS1. You should get that fixed before comparing performance, it is a total mess.

It's due to the benchmark, not an issue on hardcore ranger mode

 

[Leadbox]

You're a tad faster there despite lower clocks, perhaps the Haswell advantage?
1359/1900

Those over inflated boost 2.0 maximums are propping up the geforce averages some. Just compare the above plot and Ballas' to the one ICDP posted and those minimums. Where's that smoothness thread again :whiste:

 

[Freddy1765]

I'm in Europe where the same price gouging does not appear to be happening. Reference 290s are ~$100 less than the average aftermarket 780, and aftermarket GB/Sapphire 290s (so far the only listed manufacturers) are priced within 5% of 780s from the same brands.

Amazing with some clear and concise answers! I'll look forward to some more options on the 290 front, really want to try out the Asus DC2 cooler this time. If only it were possible to find 290 reviews of something other than the ubiquitous Tri-X..