ASUS nVidia GTX590 Review
Introduction and Technical Specifications
Published: 24th March 2011 | Source: ASUS | Price: £579.99 |
Introduction
Putting two GPUs onto a single circuit board is a fairly recent trend for the highest models in a particular graphics card series. Although it had been tested before with varying success it was the GTX295 and ATI HD4870X2 that really proved it was both possible and beneficial to go for SLI/Crossfire but on a single card.
Of course the primary benefit is that you can buy two, and go for a Quad-GPU setup, but there are also energy savings over having two single cards.
Heat however is the largest hurdle to overcome. This has been dealt with either by reducing the performance of the GPUs used, or resorting to something that sounds like a 747 taking off.
Today we're looking at the latest attempt from nVidia, the GTX590. It comes at the perfect time as we've only just reviewed the AMD HD6990 and so we can really get to grips with which of the two is the King of single card performance.
Without further ado let's get to it.
Technical Specifications
Product Name | GeForce® GTX 590 |
GPU | Dual NVIDIA® GeForce® GTX 500 series |
Engine Clock speed | 607 MHz |
Unified Shaders | 1024 (512 per GPU) |
Shader Clock | 1215 MHz |
Memory Clock speed | 3414 MHz |
Memory | 3072MB DDR5 (1536MB per GPU) |
Memory interface | 768-bit (384-bit per GPU) |
Display Outputs | Triple DL-DVI-I, mini-DP |
HDCP | Yes |
Cooling | Active (with fan) (dual-slot) |
DirectX® version | DirectX® 11 with Shader Model 5.0 |
Other hardware features | 8-channel Digital Surround Sound, HDMI 1.4a compatible, HD Audio bitstream capable, hardware accelerated Blu-ray 3D ready, Quad NVIDIA® SLI™ ready, NVIDIA 3D Vision™ Surround ready |
Software Features | nView® Multi-Display, Hardware Video Decode Acceleration Technology, NVIDIA® CUDA™ technology, OpenGL® 4.1, |
Windows 7 capability | Windows® 7 with DirectCompute support |
Most Recent Comments
+rep
No, cos you've been around so long that you already know how much of a pain in the arse, opinionated, waffle-spouting, member I can be.
Really only trying to advise how the respective document system works, and I attract the arguments probably by the way I write things. Sue me, it took me ages to get an english qualification, whilst at the same time I got math, electronics and physics, and went on to work in the field where we get these document releases.
And despite how members may take things, or increasingly tend to over the last so many years, I honestly don't mean any offense by any of it. If someone takes it, I'll be the first to appologise.
This post pretty much says it all, pain in ares is right. You say you have an English qualification, so no excuse there. You claim something about math, electronics and physics, so no excuse there. You claim that you will be the first to apologize if someone is offended.
Well I am offended, everyone who has suggested that you are wrong has been responded to with off point often rude, sarcastic and hyperbolic comments. So all these people don't know what they are talking about? You think you are the only person who understands the way things work? My how lucky we all are that you are here to inform the rest of us. What would we idiots do without you.
What a load of SH*T!
The company I work for is a member of the PCI Special Interest Group and I am a consulting engineer. So when I say that you are wrong I think everyone on this forum will appreciate my meaning.
So here goes, you are wrong.
Now anyone can argue for the sake of argument. It takes real character to admit when you are wrong. So what's it gonna be, do you have the stones to admit when you are wrong or are you that other guy?
This post pretty much says it all, pain in ares is right. You say you have an English qualification, so no excuse there. You claim something about math, electronics and physics, so no excuse there. You claim that you will be the first to apologize if someone is offended.
Well I am offended, everyone who has suggested that you are wrong has been responded to with off point often rude, sarcastic and hyperbolic comments. So all these people don't know what they are talking about? You think you are the only person who understands the way things work? My how lucky we all are that you are here to inform the rest of us. What would we idiots do without you.
What a load of SH*T!
The company I work for is a member of the PCI Special Interest Group and I am a consulting engineer. So when I say that you are wrong I think everyone on this forum will appreciate my meaning.
So here goes, you are wrong.
Now anyone can argue for the sake of argument. It takes real character to admit when you are wrong. So what's it gonna be, do you have the stones to admit when you are wrong or are you that other guy?
I'm betting he is the other guy.
The company I work for is a member of the PCI Special Interest Group and I am a consulting engineer.
So you should be aware of how the system works and the possibilities of future statements/papers/addendums that may come on the back of technology developements.
And I do appologise, unreservedly if any offense is taken.
Excellent research, and best of luck with your ATX standard mobo with the agp for advanced graphics and isa lanes for the 7 expansion slots. Let us know how you get on with that, I assume your current mobo fits these standards.
There is no debate as to whether PCI-SIG mention 375w as the operational limit, but as I'm try to explain to you all about how the system works, this is/was dependant what is available at the time of testing and writing of the said document. i.e. 2x 8pin atx pcie power connectors - it would not be envisaged that anyone would go beyond that at the time.
Back with the base 2.0 specification they would have stated 75w, for the slot, plus *whatever additional power could feasibly and realisticly be supplied*. They added molex, upped the power, 6x pin, upped the power, 8x pin, upped the power .. til the documents now read 375 .. another change comes along and so the documents will/may change.
As a statement of intent, for sure both parties stated they were sticking within this, and they have. They're all nice and friendly behind the scenes, even tho tho poke tounges out at each other when the other's not looking.
This isn't a political battleground or anything buddy, I'm merely trying to explain to you how the system works.
At the current moment in time, the PCIe 2.0 (latest 2.1) 'should' (I can't confirm this) be being used within modern up to date mobos being release. 2.1 is (should be) the final stepping stone to 3.0, it is *practically*, for all intensive purposes, the same in everything except data usage (for arguments sake). Whether this itself carries the same electrical properties of 3.0 I can't tell you, as the documents within PCI-SIG, which you'll need membership for to look at maybe, do-not give specific numbers on what the power req for base-3.0 will be. I did speculate the other week, knowing that both nvidia and amd have release these cards that so fragrantly play the "I'm not touching you" game with 375w, amd especially as they have a bigger hand in mobos these days, that they are infact in-house using "3.0" and are gearing up for it for future releases.
The reasoning behind many of the failures of going beyond 375 in testing gfxcards is down to how these 8x pin sockets are supplied. If you go down the standard route of conventionally hooking up your psu and using 8x pin designated cables, your system can and usually will, shutdown as the psu trips out. BUT if you supply the 8x pin socket with a combination of power sources, using adapters in the main, much more than 375 can be there if the card requires it - i.e. what overclockers will tend to do if they're intent on breaking the boundries.
For sure, there will be disclaimers all over the websites of these manufacturers explaining to you how awful it will be for your system if you go beyond 375, they'll probably even use it as a means to refuse warranty. But hey - it's a disclaimer - Intel have disclaimers about how many volts they want you to put over their cpus, and how much notice of that do enthusiasts take notice of ? In today's climate we need disclaimers on bridges that dangling your baby off the edge of it could result in harm and the bridge people won't be held responsible.
Quack Quack quack quack quack......
I cant confirm this, I cant confirm that, what a bunch of doubletalk. All these excuses just to get out of admitting your wrong. It's like cupojoe said arguing for the sake of argument.
Man up and admit your wrong
/warning shot off the bow
Aye calm it down now you lot please.
/warning shot of the bow
Roger that!
Apologies to all for my part, that includes you Rastalovich.
Roger that!
Apologies to all for my part, that includes you Rastalovich.
Way to step up oneseraph.
I share your sentiment.
Daddy liky
CSS Valid
If so then we're going to hell together. lol