Archive Home arrow Reviews: arrow Video Cards arrow ASUS ENGTX580 GeForce GTX 580 Video Card
ASUS ENGTX580 GeForce GTX 580 Video Card E-mail
Reviews - Featured Reviews: Video Cards
Written by Olin Coles   
Monday, 22 November 2010
Table of Contents: Page Index
ASUS ENGTX580 GeForce GTX 580 Video Card
GeForce GTX 580 Closer Look
GeForce GTX 580 Detailed
Features and Specifications
Video Card Testing Methodology
DX10: 3DMark Vantage
DX10: Crysis Warhead
DX11: Aliens vs Predator
DX11: Battlefield Bad Company 2
DX11: BattleForge
DX11: Lost Planet 2
DX9 SSAO: Mafia II
DX11: Metro 2033
DX11: Tom Clancy's HAWX2
DX11: Unigine Heaven 2.1
ASUS ENGTX580 Overclocking
GeForce GTX 580 Temperatures
VGA Power Consumption
ASUS GeForce GTX 580 Conclusion

NVIDIA Fermi Features

In today's complex graphics, tessellation offers the means to store massive amounts of coarse geometry, with expand-on-demand functionality. In the NVIDIA GF100-series GPU, tessellation also enables more complex animations. In terms of model scalability, dynamic Level of Detail (LOD) allows for quality and performance trade-offs whenever it can deliver better picture quality over performance without penalty. Comprised of three layers (original geometry, tessellation geometry, and displacement map), the final product is far more detailed in shade and data-expansion than if it were constructed with bump-map technology. In plain terms, tessellation gives the peaks and valleys with shadow detail in-between, while previous-generation technology (bump-mapping) would give the illusion of detail.

id-imp-tessellated-character.jpg

Using GPU-based tessellation, a game developer can send a compact geometric representation of an object or character and the tessellation unit can produce the correct geometric complexity for the specific scene. Consider the "Imp" character illustrated above. On the far left we see the initial quad mesh used to model the general outline of the figure; this representation is quite compact even when compared to typical game assets. The two middle images of the character are created by finely tessellating the description at the left. The result is a very smooth appearance, free of any of the faceting that resulted from limited geometry. Unfortunately this character, while smooth, is no more detailed than the coarse mesh. The final image on the right was created by applying a displacement map to the smoothly tessellated third character to the left.

Benchmark Reviews also more detail in our full-length NVIDIA GF100 GPU Fermi Graphics Architecture guide.

Tessellation in DirectX-11

Control hull shaders run DX11 pre-expansion routines, and operates explicitly in parallel across all points. Domain shaders run post-expansion operations on maps (u/v or x/y/z/w) and is also implicitly parallel. Fixed function tessellation is configured by Level of Detail (LOD) based on output from the control hull shader, and can also produce triangles and lines if requested. Tessellation is something that is new to NVIDIA GPUs, and was not part of GT200 because of geometry bandwidth bottlenecks from sequential rendering/execution semantics.

In regard to the GF110 graphics processor, NVIDIA has added a new PolyMorph and Raster engines to handle world-space processing (PolyMorph) and screen-space processing (Raster). There are sixteen PolyMorph engines and four Raster engines on the GF110, which depend on an improved L2 cache to keep buffered geometric data produced by the pipeline on-die.

GF100 Compute for Gaming

As developers continue to search for novel ways to improve their graphics engines, the GPU will need to excel at a diverse and growing set of graphics algorithms. Since these algorithms are executed via general compute APIs, a robust compute architecture is fundamental to a GPU's graphical capabilities. In essence, one can think of compute as the new programmable shader. GF110's compute architecture is designed to address a wider range of algorithms and to facilitate more pervasive use of the GPU for solving parallel problems. Many algorithms, such as ray tracing, physics, and AI, cannot exploit shared memory-program memory locality is only revealed at runtime. GF110's cache architecture was designed with these problems in mind. With up to 48 KB of L1 cache per Streaming Multiprocessor (SM) and a global L2 cache, threads that access the same memory locations at runtime automatically run faster, irrespective of the choice of algorithm.

NVIDIA Codename NEXUS brings CPU and GPU code development together in Microsoft Visual Studio 2008 for a shared process timeline. NEXUS also introduces the first hardware-based shader debugger. NVIDIA's GF100-series is the first GPU to ever offer full C++ support, the programming language of choice among game developers. To ease the transition to GPU programming, NVIDIA developed Nexus, a Microsoft Visual Studio programming environment for the GPU. Together with new hardware features that provide better debugging support, developers will be able enjoy CPU-class application development on the GPU. The end results is C++ and Visual Studio integration that brings HPC users into the same platform of development. NVIDIA offers several paths to deliver compute functionality on the GF110 GPU, such as CUDA C++ for video games.

Image processing, simulation, and hybrid rendering are three primary functions of GPU compute for gaming. Using NVIDIA's GF100-series GPU, interactive ray tracing becomes possible for the first time on a standard PC. Ray tracing performance on the NVIDIA GF100 is roughly 4x faster than it was on the GT200 GPU, according to NVIDIA tests. AI/path finding is a compute intensive process well suited for GPUs. The NVIDIA GF110 can handle AI obstacles approximately 3x better than on the GT200. Benefits from this improvement are faster collision avoidance and shortest path searches for higher-performance path finding.

GF110 Specifications

  • 512 CUDA Cores
  • 16 Geometry Units
  • 4 Raster Units
  • 64 Texture Units
  • 48 ROP Units
  • 384-bit GDDR5
  • DirectX-11 API Support

GeForce 400-Series Products

Graphics Card

GeForce GTS 450

GeForce GTX 460

GeForce GTX 465

GeForce GTX 470

GeForce GTX 480

GeForce GTX 580
GPU Transistors 1.17 Billion 1.95 Billion 3.2 Billion 3.2 Billion 3.2 Billion 3.0 Billion

Graphics Processing Clusters

1 2

4

4

4

4

Streaming Multiprocessors

4 7 11

14

15

16

CUDA Cores

192 336 352

448

480

512

Texture Units

32 56 44

56

60

64

ROP Units

16 768MB=24 / 1GB=32 32

40

48

48

Graphics Clock
(Fixed Function Units)

783 MHz

675 MHz

607 MHz

607 MHz

700 MHz

772 MHz

Processor Clock
(CUDA Cores)

1566 MHz

1350 MHz

1215 MHz

1215 MHz

1401 MHz

1544 MHz

Memory Clock
(Clock Rate/Data Rate)

902/3608 MHz

900/3600 MHz

837/3348 MHz

837/3348 MHz

924/3696 MHz

1002/4016 MHz

Total Video Memory

1024MB GDDR5

768MB / 1024MB GDDR5

1024MB GDDR5

1280MB GDDR5

1536MB GDDR5

1536MB GDDR5

Memory Interface

128-Bit 768MB=192 / 1GB=256-Bit

256-Bit

320-Bit

384-Bit

384-Bit

Total Memory Bandwidth

57.7 GB/s

86.4 / 115.2 GB/s

102.6 GB/s

133.9 GB/s

177.4 GB/s

192.4 GB/s

Texture Filtering Rate
(Bilinear)

25.1 GigaTexels/s

37.8 GigaTexels/s

26.7 GigaTexels/s

34.0 GigaTexels/s

42.0 GigaTexels/s

49.4 GigaTexels/s

GPU Fabrication Process

40 nm

40 nm

40 nm

40 nm

40 nm

40 nm

Output Connections

2x Dual-Link DVI-I
1x Mini HDMI

2x Dual-Link DVI-I
1x Mini HDMI

2x Dual-Link DVI-I
1x Mini HDMI

2x Dual-Link DVI-I
1x Mini HDMI

2x Dual-Link DVI-I
1x Mini HDMI

2x Dual-Link DVI-I
1x Mini HDMI

Form Factor

Dual-Slot

Dual-Slot

Dual-Slot

Dual-Slot

Dual-Slot

Dual-Slot

Power Input

6-Pin

2x 6-Pin

2x 6-Pin

2x 6-Pin

6-Pin + 8-Pin

6-Pin + 8-Pin

Thermal Design Power (TDP)

106 Watts 768MB=150W / 1GB=160W

200 Watts

215 Watts

250 Watts

244 Watts

Recommended PSU

400 Watts

450 Watts

550 Watts

550 Watts

600 Watts

600 Watts

GPU Thermal Threshold

95°C

104°C

105°C

105°C

105°C

97°C

GeForce Fermi Chart Courtesy of Benchmark Reviews



 

Comments 

 
# RE: ASUS ENGTX580 GeForce GTX 580 Video CardHarry 2010-11-22 20:42
What a waste of money to be honest to spend $550 bucks on a card which will be beaten within a few months- this thing is nearly as fast as 6850CF
Report Comment
 
 
# RE: RE: ASUS ENGTX580 GeForce GTX 580 Video CardOlin Coles 2010-11-22 22:05
Harry, not everyone wants to buy two 6850's and have no upgrade path a year or two later. So you might reconsider who's wasting money.
Report Comment
 
 
# RE: RE: RE: ASUS ENGTX580 GeForce GTX 580 Video CardHarry 2010-11-23 03:14
Olin, in two years time you'll be wanting to change your motherboard as well because the new PCIE interface is already in works. Also there aren't many heavy games around to justify an expensive upgrade. I mean a 5870 can easily handle games for up to two more years. The real upgrade mania comes when there is a game or a multimedia application to justify an upgrade, all these chip manufacturers are going through serious financial issues that's why they bring out newer models even when there isn't any demand.
Report Comment
 
 
# RE: RE: RE: RE: ASUS ENGTX580 GeForce GTX 580 Video CardOlin Coles 2010-11-23 14:26
See my reply to Gordon, below.

Using your own example, if someone is to purchase a new motherboard simply for the benefit of a new PCI-E 3.0 specification, which entails a new processor and possibly memory, why wouldn't they also buy the best graphics card now? You can't have it both ways, Harry.
Report Comment
 
 
# RE: RE: RE: RE: ASUS ENGTX580 GeForce GTX 580 Video CardTom 2010-11-24 04:56
Harry, even if the GTX580 is a bit slower than 6850CF and costs a little bit more, there's still the issue of Multi-card configurations CF/SLI being substantially more complicated than single card.
Just take a look inside any ATI/nvidia release notes and you'll see that the vast majority of bugs are in multi-card configurations.
I'm willing to sacrifices 10% performance and pay 10% more for the more simple/compatible option.
Report Comment
 
 
# justifyclaydough 2011-01-31 22:21
As an artist dreaming for a next gen 2.0 that can someday match what an "army" of creative types can achieve...
( not only in resolution but in pure/massive imaginary scope )
This attitude is very de-moralizing.
Why even read a graphic card benchmark review if you live for the "status quo"? Just git yerself a console and button mash to that.
The graphic card arms race assures me a future with tesselation, vsm soft shadows, and
global illumination:
##youtube.com/watch?v=Dd8yMPZzWfE&feature=related
none of which is or will be ready for prime time if we don't fan the flames.
The product is coming.
Report Comment
 
 
# RE: ASUS ENGTX580 GeForce GTX 580 Video CardDavid Ramsey 2010-11-22 20:59
Um, be beaten in a few months by what, exactly? And how do you know that whatever upcoming wonder card might beat it will be priced anywhere near it? As a hardware geek who must have the latest and greatest, I welcome whatever ATI can come up with in response. Right now, the GTX580's combination of power, noise, and low temps is unbeatable.
Report Comment
 
 
# superTrajan Long 2010-11-22 21:15
Crushes AMD's single GPU like roadkill. SLI is unbeatable. Great image quality and CUDA. IN many games 580 minimums are better than 5970. Smoother gameplay all around.
Report Comment
 
 
# SuccellusFranck 2010-11-23 04:12
Great article cover the single card contenders and the SLI that could and can match it.
Its clear its not worth yet buying it unless you dont thrust SLI and you have power consuption/heat issues.
BUT, since SLI is only relevant in few games its a factor to be considered.
Many already rushed to the market to buy the vanilla 580, i suppose many more will go to the premium that are coming.
Report Comment
 
 
# Worth it to doRealNeil 2010-11-23 09:06
It's worth the money if you have it, and I probably wouldn't over clock it either. It run well enough at stock, and if I want more performance then I'll buy another one of them. I had decided earlier on that I was going to but two GTX460's and put them into SLI for my new PC, but this card is smarter for me, so it's what I'll get.
Report Comment
 
 
# RE: ASUS ENGTX580 GeForce GTX 580 Video CardGordon 2010-11-23 14:00
Harry is making a very simple point that I think you are looking past in defense of being a techy like we all most likely are. The card may be great but the cost is too high and the performance over earlier cards isnt enough to justify spending another 3-5 hundred dollars a couple times a year to stay up to date. Technology moves forward at a quick rate but the improvements are very incremental, not large leaps. For each incremental improvement they pump out a new card and charge top dollar for it and add a higher number to the end of the cards type IE GTX - 260/270/280/290/blah blah blah 460/470/480 blah blah 580. I still own dual 260's and have no problems playing all the new games at an appealing framerate. When DX 11 becomes the absolute standard for games and games demand it then I will upgrade. But why stay ahead of the technology train when the tracks that guide it havent been built yet?
Report Comment
 
 
# RE: RE: ASUS ENGTX580 GeForce GTX 580 Video CardOlin Coles 2010-11-23 14:23
This isn't my first high-end video card review. I've been testing top level graphics cards for over four years, and every single time it's the same argument and complaint. It's a stale topic that I've written about at length, which I doubt you or Harry has ever read, so I'll summarize for you: They don't continue to make Ferrari's and Lamborghini's because they're fuel efficient or economic, the make them because people will buy them. To this end, high-end video cards will always have a market. It may not be the majority market which is filled with integrated graphics and casual computer users, and it may not be the mainstream graphics market that spawns from these high-end successes, but it?s a big enough market to continue pushing the envelope.

Additionally, you claim that software doesn't exist to warrant the need for products like the GTX 580, and I believe this to be untrue. Have you ever tried playing Crysis with all the graphical glory it was designed with? How about Metro 2033, Lost Planet 2, or Mafia II? Sure, you can play them with PhysX disabled and AA/AF turned down low, but that's like watching movies in standard definition. The logic of this argument also means that DVD and Blu-ray are unnecessary technologies, and VHS was 'good enough'. Not everybody wants to stand still... some of us enjoy the slow march forward.
Report Comment
 
 
# *THOROUGH*Zack 2010-11-24 07:06
The last two reviews (the general 580 and this one) were incredibly thorough and useful. The tools used, the descriptions, the overclocked charts, price per FPS per game, etc. were all GREATLY appreciated by this humbled little nerd. Even the comments were knowledgable and delivered in a useful way.

THANK YOU VERY MUCH for this review. I hope you get a raise.
Report Comment
 
 
# Tricky Spot...BeerGutBarry 2010-11-25 01:54
I love the new GTX 580, but can get the new Gigabyte 480 SOC for a bit cheaper...torn.

Curious to see how far partners can push this card though. The improvements made to the 480 from release to now have been pretty impressive, so wondering if similar gains can be made with the 580, or if this is already pushing the limit (without causing a massive increase in power consumption and removing throttling).
Report Comment
 
 
# RE: Tricky Spot...Olin Coles 2010-11-25 04:57
If it's a choice between the GTX 580 and GTX 480 SOC, I'd jump on the GTX 580 in an instant.
Report Comment
 
 
# how can i spend 770$ ??LASER 2010-11-25 03:26
Right Olin!! yea i dont wanna feel that my card is struggling to play the best game i liked..looking at tesselation video that was just to explain the idea.... coming game might share something which wont feel like u playing a game...so tesselating hardware i decided...

GTX 580 in India is for 770$ !!!
GTX 570 will be at 520$ for me....
i saw ferrari ships for 20$-84$ if u order some merchandise...
So is there any chances i can get that GTX 580 in 550$ here???
Report Comment
 
 
# Great ReviewWarren 2010-11-25 12:54
Great review. I bought two of these, thus proving that there are people out there who will buy the very best regardless of cost.

As you rightly mention, Ferrari do not make cars that are the most fuel efficient at 60Mph, they know that there are buyers out there who want the best at any price.
Report Comment
 
 
# To Ferrari or Not FerrariRickard 2010-12-06 02:14
Weird review, mixing crossfire/sli scores with normal setups. Would have been more easily understood if you had had a separate table for cf/sli setups.

Also it would be interesting to see the scores mapped to a dollar/performance index to see where the best gain is to be had.

Further its interesting that the cf/sli scores are slightly different with other reviews where the 580 consistently (well almost) had scored lower than the 5970.

I know many think hey its a dual gpu card but hey it costs less and you can still cf/sli it so I dont really see the point they are trying to make.

Imo Nvidia is on the right track and the 460 and 580 is really good cards but atm they are basically catching up and I am waiting to see what the response from AMD is before I commit to an upgrade.
Report Comment
 
 
# fast new cards?Laszlo 2010-12-10 08:21
For me, a top card is one which beat the older top dog. But none of the new cards doing this (eg. none of them beat the GTX295's raw power: most of them slower except in 1-2 titles),even if they equipped much faster ram, clocked higher,... I would like to see a GTX295 manufactured with 40nm process, GDDR5 mems, 512bit wide mem bus, clocked as high as the new cards. What would touch it in DX10? So we get new feats (which maybe could be SW rendered too..) but not speed. Like 470 -> 570: most of the 25-27% speed advantage of the 570 gained from higher clock speed and just fragments from any optimization and the one less disabled CUDA or other unit. From another point of view: it's not a coincident that first came the slower 480 (vs. 580), then 470 (crippled) then 580 with more MHz, then 570.. maybe we will see a 680 too if the 40nm process allows more than 1 GHz - as a new, very advanced chip of course (which is internally the same). We getting the same, only the price/crippling ratio changes. We are very far away from ray traced rendering, real physics, real materials, ballistics, etc.
Report Comment
 
 
# RE: fast new cards?Olin Coles 2010-12-10 08:29
I might be misunderstanding your message, so allow me to seek clarification. Are you saying a GeForce GTX 295 outperforms the GTX 580? Just so that you know, it really doesn't even come close. Here's exhibit A: benchmarkreviews.com/index.php?option=com_content&task=view&id=403

If the Radeon HD 5970 can clear it by 35-40% in Crysis and Heaven, then look at where the GTX 580 sits against the Radeon 5970 on those tests in this article.
Report Comment
 
 
# RE: ASUS ENGTX580 GeForce GTX 580 Video CardLaszlo 2010-12-10 10:38
Yes I was misunderstable. Clock to clock (same GPU/Mem) the 580 would be slower. Or: the 295 would be much faster, manufactured @ 40nm process and set to same clock speeds with same memory modules. = We don't get what nv can make, we get crippled cards - but with new feats. If the 580 would get the GTX285's 512 bit wide memory bus (the 295 also crippled to 448 bit, and from 32 to 28 ROPs per GPU), and 2048 MB memory, 64 ROPs, etc (as it's double number of transistors dictate), then it would be a real high end (technology wise) card. Now it's a very fast card, that milk the consumers like a cow, then nv needs to 'invent' some of their inventions again and we never get a new generation of cards that 100 or more percent faster than their precedessors. When they made their first(?), the Riva128(?) then there was no other just their best chip, and it was blown away by their next, faster chip. Not by one more feat, but by raw power - in ALL games.
Sadly they make only high end chips these days too, then crippling them down to a price category.
So these new cards are fast but far less than they should be - and they should be 50-100% faster in everything, especially in 'future' games that really use their new feats.
Report Comment
 
 
# Nice reviewThomas 2011-01-15 14:42
Very nice review, well done. Thanks very much! That'll be about it...greetings
Report Comment
 
 
# To Olin ColesGordon 2011-01-31 22:59
Olin,

Hey, so to clarify I was only trying to make what Harry was saying a little more clear as I assumed there was some misunderstanding. I agree with you in every way as to what you replied, and I should say that your comparison to Ferrari and Lamroghini was spot on. I am definately in the casual group but not by choice, lol. I would, if i could afford, like to stay up with all the newest hardware as I am a firm believer that your rig is only as strong as its weakest component. My gtx 260's: Do they play even Crysis on appealing settings at a decent enough frame rate? Yes I can push them to relatively high settings and play at about 30 or more FPS. BUUUUUT I want a GTX 580 so I can crank that up just as much as the next guy HAHA. Either way I liked your review and found it very informative.
Report Comment
 

Comments have been disabled by the administrator.

Search Benchmark Reviews
QNAP Network Storage Servers

Follow Benchmark Reviews on FacebookReceive Tweets from Benchmark Reviews on Twitter