AMD Radeon R9 290 Review: Fast And $400, But Is It Consistent?

[ojas]

@Chris: Maybe try using the Arctic cooler on the retail 290X?

Hope you have a 290 and 290X roundup planned! I think this would be the first time where a good cooling solution would have a direct effect on performance, perhaps even more so than overclocking.

 

[cvearl]

Dumb move. At $400, dropping noise levels a tad and coming in just below GTX780 performance would have STILL been amazing and all of these reviews would have been "OMG" performance/value card of all time! Then the partners could have released exotic cooling versions in the $429 - $449 zone that would curb stomp the 780. Why AMD? I might just get a $500 EVGA GTX780 Superclocked and wait a few years to care again.

C.

 

[coolitic]

I stick with my belief.
AMD is cheap, but it sucks at everything else.

 

[Mousemonkey]

Well as a long time AMD fan this is both great news and cause for concern. That disparity between retail and press cards is disturbing to say the least. I am also curious if a proper cooling solution would rectify that problem. As to why I like AMD well much like their CPU's the Hawaii line looks to be power hungry and hot but will shine if power is not a concern and pampered by someone who knows what their doing. I'm looking forward to what I can do with one of these with a proper water cooling setup.

The retail version is slower? Whats with the discrepancies?

"Clearly there’s something wrong when the derivative card straight from AMD ends up on top of the just-purchased flagships. So who’s to say that retail 290s won’t follow suit, and when we start buying those cards, they prove to underperform GeForce GTX 780?"

Something ain't right here, so the one you buy in the stores are slower than the ones AMD gives Tom's to benchmark and review?

Isn't it amusing how the fans are all overlooking that bit! :lol:

Now if Nvidia were to do the same with the 780Ti I bet they would have something to say.

 

[ojas]

The only thing I hate in your reviews Chris , is because you over exaggerate about noise. I have seen people run GPUs like 7970s on ref cooler with full speed with little complaint of noise. I always see you complaining about noise on AMD's ref coolers. Firstly , people have cases at a distance of like 1m from the ear in many ways. Can be from front , back , bottom or sideways. Consider the nearby noise , since not everyone lives in a place with no noise around , right ? Then noise reducing solutions like the AcoustiPacks which many people prefer for reducing noise. They are available for like 30$ and thats with 3 layers coating. Then again , your reviewing a 400$ card performing at par ( in some way ) with a stupidly price 1000$ card. Do you think its fair to say its noisy ? I'd say mention it can be noisy , very noisy , but dont overexaggerate it to make a huge reason to not go for the card.
"I have two issues with this. First, at 47% duty cycle, the fan is too loud. It’s obviously not as bad as the 290X’s Uber mode, but I don’t see any compelling reason to compromise acoustics when quieter solutions exist" Stick to them. And spoil the beauty too.
IMO you should think more of the average case too.
Next time if you could provide acoustics from a distance of 100 , 150 cm like , I would appreciate it.
Also Kudos to AMD. It should awake NVIDIA that AMD can fight em.
Waiting for the custom coolers to hit on those beasts,

Well, AnandTech actually recommended against buying this card simply because of noise.
http://www.anandtech.com/show/7481/the-amd-radeon-r9-290-review/17

 

[ojas]

Dumb move. At $400, dropping noise levels a tad and coming in just below GTX780 performance would have STILL been amazing and all of these reviews would have been "OMG" performance/value card of all time! Then the partners could have released exotic cooling versions in the $429 - $449 zone that would curb stomp the 780. Why AMD? I might just get a $500 EVGA GTX780 Superclocked and wait a few years to care again.

C.

What you're saying actually makes a ton of sense. But other sites are reporting that this was supposed to be competition for the 770 (which i think we all were expecting), before the price drop...which is probably why AMD ended up panicking and doing this.

 

[rolli59]

Well looking forward for the follow up review with retail cards. Until then looks like the performance is just a rumor.

 

[Jordan Nwokolo]

NV has send a full battalion of troller to all around the tech web to bashing this card,even tough they know that their effort is useless..they shouldn't to do this,coz better to them to focus and concentrate their energy to repair their FAILURE in mobile Soc

 

[fat_panda]

I want to see a review of 290 and 290X water cooled.

 

[nikoli707]

290x waterblocks are already hitting 1300mhz on the core with temps well under 70c... look to see a 290x watercooled amd solution to answer the 780ti. as for the 290.... what a beast

 

[GBJudgeEhud]

@Chris-

Which version of the 770 did you use (2gb, 4gb?). You don't list it in the setup with the other cards. Sorry if you have that stated somewhere else in article and I missed it. Since all the other cards listed in setup have >2gb of vram, I wonder if the 770 was having frame buffer challenges at the higher resolutions, if it was a 2GB version.

Thanks for the nice review.

 

[nforce4max]

As bad as the thermals and acoustics appear there has been worse that few have bothered to take notice from cards in the past like the 590 ect. This card is easily fixable provided it is in a decent case and the crappy stock compound is replaced. Wished that it came with a back plate which would shave off like 3c but would make it look better.

They could very easily redesign the cooler to use copper fins and a single connected heatpipe to loop over the top from the vapor chamber which would make the card run cooler thus hold clocks better.

 

[Tonim89]

A high end Arctic coler would turn this card into a Titan killer for 500$ less money, and completely smoke the GTX 780 for less money too. My only concern was the power consumption, but ut seems to be at the same levels of the 7970 GHz Edition.

 

[softplacetoland]

Wow, I was reading comment section and literally I could have seen the moment when NVidia trolls started to crawl out to the light. Creepy XD

Simple, keep cheering. Doesn't affect people who chose wisely, no matter which color ends in their hands.

 

[kenh536]

Gah all these variances between beta drivers and cooler performance causing throttling. I don't care about numbers until the true story is revealed: Non-reference cards with nice coolers and optimized drivers AMD vs Nvidea. Only THEN will I believe numbers.

 

[clonazepam]

"The only thing I hate in your reviews Chris , is because you over exaggerate about noise..."

Bah, silliness, he's reporting on what people care about.

Kudos to Tom's for slapping their own cooling solution onto a board. You're one step closer to slapping water on reference cards for review time. =)

 

[Trompochi]

And here I am still rocking a freakin HD 5770.... lol

 

[Novuake]

"The only thing I hate in your reviews Chris , is because you over exaggerate about noise..."

Bah, silliness, he's reporting on what people care about.

Kudos to Tom's for slapping their own cooling solution onto a board. You're one step closer to slapping water on reference cards for review time. =)

Agreed, noise tolerance is relative but for most users, this GPU at 47% fan speed will annoy ANYONE.

 

[Trompochi]

And here I am still rocking a freakin HD 5770.... lol

 

[stevethejew]

im no nvidia fanboy. in fact I have 2 280x cards in my rig as I write this. but those benches between retail and sample 290x look extremely sketchy. I wouldnt trust those 290s until I saw more retail benchmarks.

 

[logainofhades]

I sure hope the friend I bought my 7970 from plans some kind of aftermarket cooling for his 290x after seeing this.

 

[somebodyspecial]

According to Tom's Benchmarks Nvidia's price drop just became meaningless

As long as you completely ignore the benchmarks from the RETAIL cards tom has that were basicaly 770's.

If 290's follow the TWO 290x RETAILS he had here these cards suck and AMD cherry picked reviewer's cards. I tend to believe RETAIL purchased cards when there is such a difference. I'm thinking they can't get anywhere near the perf of the PRESS cards that were received or why are the two retails so far below what all press cards show? AMD will have a crap xmas if people around the web start benching retail cards and getting what tomshardware has shown here. I'm sure we'll have a few retail cards benched at review sites before the end of this month. That's not going to be good for AMD. That performance is TERRIBLE and those should be BETTER picked gpus as they are 290x cards! You would get WORSE chips from a lower model, as you cherry pick the top bin.

Spoken like a fanboy who ignored the review you're commenting on. Did you see how bad the benchmarks were for RETAIL in the charts? None of them are good.

 

[somebodyspecial]

And here I am still rocking a freakin HD 5770.... lol

I don't feel so bad now...I'm rocking a XFX Radeon 5850 and dominating your butt...ROFL.

 

[somebodyspecial]

A high end Arctic coler would turn this card into a Titan killer for 500$ less money, and completely smoke the GTX 780 for less money too. My only concern was the power consumption, but ut seems to be at the same levels of the 7970 GHz Edition.

I don't think you understand what a titan is bought for usually. Sure a few buy them for gaming (and lets not forget it's been out for 8 months), but most buy for PRO features that your radeons get smoked by (and all other NV cards that aren't Tesla). They fly off the shelves at $1000 because a TESLA is $2500 and basically is a Titan with full driver support for pro apps. So for pro people, no matter what you think, Titan is still an awesome deal today. If you are a GAMER only and had the cash for a Titan, you clearly are unaffected by this months news from AMD anyway...LOL. If you were a gamer only, you should have bought 780 instead if you have any worries about $$$.

 

[DrKlahn]

According to Tom's Benchmarks Nvidia's price drop just became meaningless

As long as you completely ignore the benchmarks from the RETAIL cards tom has that were basicaly 770's.

If 290's follow the TWO 290x RETAILS he had here these cards suck and AMD cherry picked reviewer's cards. I tend to believe RETAIL purchased cards when there is such a difference. I'm thinking they can't get anywhere near the perf of the PRESS cards that were received or why are the two retails so far below what all press cards show? AMD will have a crap xmas if people around the web start benching retail cards and getting what tomshardware has shown here. I'm sure we'll have a few retail cards benched at review sites before the end of this month. That's not going to be good for AMD. That performance is TERRIBLE and those should be BETTER picked gpus as they are 290x cards! You would get WORSE chips from a lower model, as you cherry pick the top bin.

Spoken like a fanboy who ignored the review you're commenting on. Did you see how bad the benchmarks were for RETAIL in the charts? None of them are good.

If you look at enthusiasts sites people are getting the performance of the press cards (there is a 43 page thread on Overclockers.uk). The review states the clockspeeds fluctuated on the retail 290x's. The cards may need to have their fan profiles adjusted or it could be a driver bug that is limiting the 290X. Perhaps the 47% limit is being inadvertently being applied to the 290X (which should be 55% in uber mode). It's to bad the article alarms the reader, but doesn't do any real digging into the cause. I'm betting setting the fan cap to 100% (just as an experiment) would mirror what people have been seeing.