Graphics Card Technology Part 1: Shader Processing Units

Written by Paulo Proenca on Tue, Apr 16, 2013 9:42 AM

This article will provide our lovely members with a basic knowledge about Graphics Cards. As you may have seen, we have recently launched a Graphics Card Comparison Feature that is accessible around the site.

The following article will focus on the main aspects of Graphics Cards and will endow you with enough knowledge to help you make the best choice when buying a new graphics card. It will be divided into various parts, this first one will focus on Shader Processing Units...

Before getting into the juicy bits, let’s start by clearing one thing up. It’s a mistake we all make and that I myself have made: the difference between GPU and GFX.

GFX VS GPU

GFX stands for Graphics Card. A Graphics Card/Video Card is GeForce GTX 680, Radeon HD 7970.

GPU stands for Graphics Processing Unit. The GPU is the processor inside/embedded in the Graphics Card. GeForce GTX 680's GPU is Kepler GK104 and Radeon HD 7970's is Tahiti XT.

I am better off showing than describing and so, take a look for yourself:

 

 Now that's cleared up, let's move on to the most important aspects of a Graphics Card.

Shader Processing Unit

What is a Shader Processing Unit? Also known as Stream Processor (AMD), CUDA Core (NVIDIA), the Shader Processing Unit became the most important component on a Graphics Card,  upon the release of Shader Unified Architectures, back in 2006.

Before the release of Unified Shaders, GPUs had 2 types of Shaders: Pixel Shaders and Vertex Shaders, whereas Pixel Shaders were in charge of computing color and Vertex Shaders allowed the control of movement, lightning, position, and color in any scene involving 3D models.

Both are important and GPUs always had a major problem: an unbalance between Pixel and Vertex Shaders. Example: Radeon X1900 - 36 Pixel Shaders and 8 Vertex Shaders. The problem? Depending on the need of 3D application, often Pixel or Vertex would be idling. Pretty much like a computer waiting for its printer to finish a print order. Another term would be: bottleneck.

Well, unified Shaders will switch between the two depending upon what work needs to be done. In other words, you don't have one or the other, ideally, sitting idle waiting for the other group to finish their work and thus allows higher efficiency.

Now, what you want to know is how to compare Graphics Cards - especially from different manufacturers. AMD and NVIDIA have always had different Shader Processing Units (SPUs) designs which make it harder to compare Graphics Cards directly. AMD's cards have a higher count of SPUs but are less complex. On the other hand, NVIDIA has a lower count of SPUs but they are much more complex/powerful.

A good way to help comparing the SPUs of the 2 manufacturers, is to attribute them a "power".

Until the release of the Kepler Architecture, NVIDIA's Shaders have always been higher clocked, when compared to the central unit of the GPU.

The First Shader Unified Architecture, used on the GeForce 8000/9000 Series, had the Shaders Clocked up to 2.5 times (2.5X) as fast as the central unit. That's why you see that the central unit of GeForce 8800 GT is 600MHz but its Shader/Processor Clock is 1500MHz. This allowed tremendous power but caused enormous heat.

With the release of the Fermi architecture, this changed. NVIDIA developed a technique known as Hot Clocking: the Shaders were clocked twice as fast as the central unit - 2X. This meant less heating and less power consumption.

However, the big bang was with the Kepler architecture.

On the Kepler architecture, the Shaders were clocked at the same speed of the central unit and the concept "Shader/Processor clock" became obsolete. Adding to that, they were up to thrice as energy efficient as Fermi Shaders.

Now, you must be wondering: if the speed of the Shaders has been reduced over the years, how come graphics cards are more powerful now?

The thing is, each architecture has allowed a higher SPU count. Less powerful but more of them. Add to that the smaller technology (which I will go over in part 2), allows higher clock-frequencies.

Here is a simple way to calculate raw shading power:

GeForce 8800 GT:

128 SPUs, 2.5X Power, 600MHz Core-Clock. Raw Power: 128 X 2.5 X 600 = 192000

GeForce GTX 285:

240 SPUs, 2.3X Power, 648MHz Core-Clock. Raw Power: 240 x 2.3 x 648 = 357696

GeForce GTX 580:

512 SPUs, 2.0X Power, 772MHz Core-Clock. Raw Power: 580 x 2.0 x 772 = 895520

GeForce GTX 680:

1536 SPUs, 1.0X Power, 1006MHz Core-Clock. Raw Power: 1536 x 1.0 x 1006 = 1545216 + Boost Clock = up to 1625856

GeForce GTX Titan:

2688 SPUs, 1.0X Power, 836.5MHz Core-Clock. Raw Power: 2688 x 1.0 x 836.5 = 2248512 + Boost Clock = 2353344 + Max Clock (e.g.: 1GHz, as it depends on Temperature & Voltage) =  up to 2688000

Things are much more simple with AMD. "Shader Power" like I have called it before, has only changed with the Graphics Core Next (GCN) architecture.

Radeon HD 2900 PRO:

320 SPUs, 0.65X Power, 600MHz Core-Clock. Raw Power: 320 X 0.65 X 600 = 124800

Radeon HD 4890:

800 SPUs, 0.65X Power, 850MHz Core-Clock. Raw Power: 800 X 0.65 X 850 = 442000

Radeon HD 6970:

1536 SPUs, 0.65X Power, 880MHz Core-Clock. Raw Power: 1536 X 0.65 X 880 = 878592

Radeon HD 7970:

2048 SPUs, 1X Power, 925MHz Core-Clock. Raw Power: 2048 X 1 X 925 = 1894400

Radeon HD 7970 GHz Edition:

2048 SPUs, 1X Power, 1000MHz Core-Clock. Raw Power: 2048 X 1 X 1000 = 2048000 + Boost Clock = up to 2150400

 

 

For both cases, Open GL and Shader optimizations provided by certified drivers for each architecture can unlock GFX potential even further.

Which company does it better?

In my opinion, until the release of the Kepler architecture, AMD had been allocating more SPUs with fewer transistor counts compared to NVIDIA's Graphics Cards (e.g GTX 580 Vs 6970). Which means it was more efficient in allocating SPUs per transistors.

Having a lower count of SPUs and having them higher clocked also produced unecessary heat and increased power consumption.

Therefore, I vote that AMD did it better. However, with the release of Kepler Graphics Cards, things have changed. NVIDIA's Graphics Cards now heat as much as AMD's GFX and consume far less power, in a way a dual-graphics GeForce GTX 690 consumes as much as AMD's Radeon HD 7970 GHz Edition.

In the end, how can you compare a NVIDIA SPU to an AMD SPU? You can't.

Only by making simple tech comparisons, which are estimated and not counting driver support. So this formula is pretty accurate between NVIDIA-NVIDIA and AMD-AMD Graphics Cards but might not yield accurate results for NVIDIA-AMD cards.

Also, its important to note that these raw values don't mean a whole lot by themselves, which is why we have a new GFX comparison area that helps you compare one GPU to another across a broad range of criteria. A card has other components that are important as well, which I will go over in the following articles.

In Graphics Card Technology Part 2 I will be focusing on the manufacturing technology and how it has changed over the years. Plus Texture Mapping Units and Render Output Units and how they have and still do affect a Graphics Card's performance.

In Graphics Card Technology Part 3, I will be taking a look at the memory bandwidth and memory size - one of today's most overrated components - especially in low-end cards.

 

Based on the SPU GFX tech who performs better now?

Login or Register to join the debate

Rep
48
Offline
21:07 Oct-21-2015

if the SPU are clocked at 2.5 times the speed of the core clock, isn't the shader clock bottlenecked by the core clock and operate at core clock speed?

0
Rep
6
Offline
13:31 Apr-25-2013

nicely explained!!

1
Rep
30
Offline
admin approved badge
10:21 Apr-22-2013

well-written, and a juicy introductory to us whom left outdated with the technology. Great insight, Pip! :)

1
Rep
14
Offline
08:13 Apr-22-2013

Excellent article and very insightful.

1
Rep
0
Offline
11:47 Apr-21-2013

this is a good and helpful blog ^^

1
Rep
64
Offline
16:41 Apr-20-2013

Though I knew some parts of these before, like the difference between GPU and GFX. But still a great and helpful blog...

0
Rep
-9
Offline
15:20 Apr-20-2013

Just so all you (fan)boys know, Nvidia cards are best out of the box, whereas AMD ones have better OC potential.

0
Rep
8
Offline
14:20 Apr-20-2013

This article helped me a lot , good job :D

0
Rep
133
Offline
admin approved badge
19:39 Apr-19-2013

Pip looks pretty Awesome in that picture,who knows whats going through that head,Delving in a dark mysterious Adventure?speculating how many Transistors the malta will have?i dare not imagine.

0
Rep
87
Offline
13:59 Apr-19-2013

AMD FTW!

1
Rep
59
Offline
13:46 Apr-19-2013

i wonder why intel doesn't make GPU's. well dedicated i mean.

0
Rep
52
Offline
16:42 Apr-19-2013

and i dont see why they should make them althought it would be cool to have a little more competition

0
Rep
140
Offline
admin approved badge
18:50 Apr-19-2013

they don't need to, their processors are so popular now anyway, there's no need to join the GFX part of hardware just yet, they may need to when AMD get popular since their processors are being used in consoles

0
Rep
1,041
Offline
senior admin badge
22:43 Apr-19-2013

they focus on powerful igpus,
I can say from my own experience that if you don't use recent Intel igpus for gaming, they handle everything very well :Đ

0
Rep
65
Offline
13:04 Apr-18-2013

Nvidia it's meant to be played lol ;)

0
Rep
75
Offline
13:41 Apr-18-2013

Thats right 100%

0
Rep
2,003
Offline
admin approved badge
11:27 Apr-19-2013

Gaming Evolved ;)

0
Rep
75
Offline
12:32 Apr-19-2013

to slow...hehe ....need more technology to catch last release from Nvidia it's meant to be played!

0
Rep
2,003
Offline
admin approved badge
16:27 Apr-19-2013

I'm running everything just fine with my HD7970 ;)

0
Rep
21
Offline
16:32 Apr-19-2013

Except for Crysis 3 :)

0
Rep
2,003
Offline
admin approved badge
18:08 Apr-19-2013

Nope, that too ;)

0
Rep
133
Offline
admin approved badge
19:00 Apr-19-2013

tero's rig is better no one buys GTX 680 anymore! after the 7970 was it became the best GFX for high end users and 7770 best mid/low end users.

0
Rep
366
Offline
admin approved badge
19:08 Apr-19-2013

You are a tad mistaken, HD7970GHz was the fastest, better than GTX680, but not the regular HD7970. Anyway, Titan is here now ;)

0
Rep
133
Offline
admin approved badge
19:13 Apr-19-2013

yes but ATI offered better prices and i meant 7970 Ghz (the normal one isnt really that good and ithink is a bit stronger than GTX 670 but still pretty Awesome),and who the Fu** can afford a titan,and btw THERE SHALL BE MALTA!

0
Rep
366
Offline
admin approved badge
19:20 Apr-19-2013

So you can't afford Titan but you can afford a Malta GFX? ;D

0
Rep
133
Offline
admin approved badge
19:28 Apr-19-2013

No ;) both Malta and Titan Are just bullSh** GFX and theironly purpose is become the worlds fastest GPU!who cares about what the fastest GPU is when they cant buy it!are we supposed to be jealous?coz im not.im jealous from a guy who has a FX 4300 and 7870,not some guy who has 3-Way SLI titan with i7 3970X and cazillion GB of RAM!a person has to be reasonable in life,you cant always get what you want,You Get What You Get!did you know that with the money of 4 Titans i could buy a brand new Car?

0
Rep
366
Offline
admin approved badge
19:34 Apr-19-2013

↑ Tell me about it

0
Rep
133
Offline
admin approved badge
19:36 Apr-19-2013

i just did ;) bet you saw that one coming :P

0
Rep
91
Offline
admin approved badge
00:23 Apr-20-2013

Sorry to say this, but something isn't bullsh it just because you can not buy it. There purpose is to move the gpu world forward. People have bought the titan and not just for benchmarking. multi-monitor, 3d, or 4k resolution needs a very powerful card. People that are buying these cards have money to spend on them. Just because you can't get it doesn't mean it shouldn't be made or others shouldn't get it. I don't get why people are saying this card is dumb. This card is exactly what this industry needs to keep moving forward instead of being held back by consoles for another 10 years.

0
Rep
133
Offline
admin approved badge
11:00 Apr-20-2013

what this industry needs is GFX's for average user,which as you may have guessed isnt Extremely rich people.so what this industry needs is GFX that is affordable and gives Good performance compared to the cost.What this industry needs is AMD!

0
Rep
91
Offline
admin approved badge
11:19 Apr-20-2013

Nvidia offers many mid range cards. You can not make cards cheaper if you never make better cards. 3 years from now the average user $250 card will be a titan maybe better and that is because newer better cards have been developed and made.

0
Rep
133
Offline
admin approved badge
11:29 Apr-20-2013

yeah right,prices dont drop that much dude,GTX 285 is 330$,and its only DX11,im just saying that the titan isnt a normal guys run-of-the mill GFX,it is intended for ONLY EXTREMELY High-end people with HIGH-End seemingly endless pockets,or spoiled rich kids who want to show off,atleast for me it is.

0
Rep
91
Offline
admin approved badge
11:41 Apr-20-2013

To buy a GTX 285 for that much would be the dumbest thing you could do. my point is that you can get a way more powerful current card for the same price. Of course it isn't a run of the mill card. But once the 900 series comes out the titan is going to be irrelevant and that same power is going to be in a $250 mid range 900 card. if not better.Should we not advance cards and just keep everything the same. Because without enthusiast cards like the 690 and the titan that won't happen.

0
Rep
133
Offline
admin approved badge
11:52 Apr-20-2013

yes,but NVidia is implying that You Must buy Titan or you just cant run games well!they have already probably ruined the way the people were supposed to run Metro by making it Nvidia optimized and require a Titan/690 to max (yet to be seen),they dont try to be fair,theyre out to make money ,but thats what i hate the most about nvidia,AMD isnt like them they dont make stuff and say ONLY AMD users can use them,just like TressFX,and AMD has never Tried To sabotage Nvidia's work unlike Nvidia.And i stand by what i said Titan is Bullsh and the Limit for NOT bullsh GFX ends at 7970 Ghz and 680.the only purpose of titan IS TO BE THE FASTEST GFX!

0
Rep
91
Offline
admin approved badge
11:58 Apr-20-2013

I don't think they are implying anything and just stating something. Just because you can't max the physx and play in 3d at 60fps doesn't mean you can't run it well. Of course the game is going to be Nvidia optimized because Nvidia has put there own money and resources into the game such as AMD did with Tomb Raider, I guess you could say that with tomb raider they MADE you have to get an AMD card to run the game well. Nvidia isn't sabotaging anything its called a business and having to have a niche over your opponent so that you can stand out. Thats the dumbest thing i've heard that the 7970 and 680 are where graphics should end. Would you have said the same thing about the 480 and the 580. Should we not make cards faster than the titan?

0
Rep
91
Offline
admin approved badge
12:01 Apr-20-2013

Who cares if the Titan is only there to be the fastest? IT IS. Good for Nvidia. AMD is trying to make faster and faster cards all the time. They just made a 7990. Is that bull**** to? Or no because its AMD so it is fine. AMD does the same thing as Nvida in sponsoring games and trying to make the game run better on there graphics card. So don't be calling out company's and praising others when the both do the exact same thing.

0
Rep
133
Offline
admin approved badge
12:03 Apr-20-2013

i meant in this Generation.and AMD didnt Say You MUST have AMD GFX to run TR well,Nvidia card run it the same way,even TressFX it only needed a patch.

0
Rep
91
Offline
admin approved badge
12:07 Apr-20-2013

Well than don't say Nvidia is implying that you must have a Nvidia card than because that's not what they are saying. They are just offering a bonus to there customers as who buy there cards because we helped pay for the technology. You can still run PhysX just on your CPU. And why just for this generation. When does the development for one generation stop and another start? If titan came out for next gen would that have been ok?

0
Rep
133
Offline
admin approved badge
12:11 Apr-20-2013

No,the whole Reason of the titan is what Baffles me,no purpose,I honestly dont know why Nvidia Made it,i mean its not really that much of a replacement for GTX690 since its the same Performance and about the same price,the only reason i can think of is . . . .

0
Rep
75
Offline
13:28 Apr-20-2013

if you want the best you have to pay GTX 680 better then any single card from AMD
And Titan is supreme single ....but is kinda overprice!

0
Rep
133
Offline
admin approved badge
13:36 Apr-20-2013

7970 Ghz (with drivers) is the best (beside titan),and im not a fanboy,im just stating the facts.

0
Rep
75
Offline
13:39 Apr-20-2013

with 314.22 nvidia whql i really don't think so
FPS don't lie but yes 7970ghz perform better in some games and GTX 680 in others !!

0
Rep
133
Offline
admin approved badge
13:41 Apr-20-2013

it depend on the optimization.

0
Rep
5
Offline
18:31 Apr-17-2013

really helpful ........ thanks.......

0
Rep
5
Offline
18:31 Apr-17-2013

really helpfull ........ thanks.......

0
Rep
94
Offline
16:41 Apr-17-2013

Yeah,give us more of these!!! :D Great one!

0
Rep
1,041
Offline
senior admin badge
12:05 Apr-17-2013

btw I voted "what about Intel" :Đ

0
Rep
2
Offline
11:50 Apr-17-2013

wow, that amazing :D

0
Rep
59
Offline
08:52 Apr-17-2013

thx man always waanted to know what are SPU's ,ROP's etc etc meant and what they really did. maybe you can even do an article on CPU components.

0
Rep
20
Offline
admin approved badge
07:19 Apr-17-2013

Amazing article but now that more will be published, I'll be waiting for memory bandwidth, texel/pixel rate and those "connections" :)

0
Rep
54
Offline
05:51 Apr-17-2013

WOW! This is exactly why i joined here. lol More knowledge.
Thanks a million pip. I am really glad you could put up this article, cause i am going crazy trying to decipher all the graphics cards terminologies. lol But now, thanks to you, i have learnt some important facts about GPU. :)

0

Can They Run... |

| 60FPS, Medium, 1080p
Core i7-870 Quad 2.93GHz GeForce GTX 560 Ti 448 Edition 12GB
| 60FPS, Ultra, 1080p
Core i9-10900X 10-Core 3.7GHz GeForce RTX 2070 Super Gigabyte Gaming OC 3X 8GB 32GB
| 60FPS, Ultra, 1080p
Core i7-7700K 4-Core 4.2GHz Intel HD Graphics 630 Mobile 32GB
0% No [2 votes]
| 60FPS, Ultra, 1080p
Core i7-4770K 4-Core 3.5GHz GeForce RTX 2080 EVGA XC Gaming 8GB 16GB
| 60FPS, Medium, 1080p
Ryzen 7 3700X 8-Core 3.6GHz GeForce GTX 1660 Ti Zotac Gaming 6GB 16GB
100% Yes [1 votes]
| 60FPS, High, 1080p
Core i5-11400H 6-Core 2.20GHz GeForce RTX 3050 Mobile 8GB
| 30FPS, 720p
Core i7-5500U 2-Core 2.4GHz GeForce 940M 2GB 8GB
| 60FPS, High, 1080p
Core i7-3770 4-Core 3.4GHz GeForce GTX 1070 Gigabyte G1 Gaming 8GB Edition 16GB
| 60FPS, High, 1080p
Core i7-10750H 6-Core 2.60GHz GeForce RTX 3060 Mobile 16GB
100% Yes [1 votes]
| 60FPS, Low, 1080p
Core i5-6400 2.7GHz GeForce GTX 1050 8GB
| 60FPS, Ultra, 1440p
Core i5-9600K 6-Core 3.7GHz GeForce RTX 2060 Asus ROG Strix Gaming OC 6GB 32GB
| 60FPS, High, 1080p
FX-6300 GeForce GTX 760 24GB
0% No [1 votes]
| 30FPS, Ultra, 1080p
Core i5-11400H 6-Core 2.20GHz GeForce RTX 3050 Mobile 8GB
| 30FPS, High, 1080p
Core i5-11400H 6-Core 2.20GHz GeForce RTX 3050 Mobile 8GB
100% Yes [2 votes]
| 60FPS, High, 1080p
Core i5-11400H 6-Core 2.20GHz GeForce RTX 3050 Mobile 8GB
| 60FPS, High, 1080p
Core i5-11400H 6-Core 2.20GHz GeForce RTX 3050 Mobile 8GB
100% Yes [3 votes]
| 60FPS, High, 720p
Athlon II X4 640 Radeon RX 470 Asus ROG Strix Gaming OC 4GB 8GB
100% Yes [2 votes]
| 30FPS, Low, 720p
Core i7-4710MQ 4-Core 2.5GHz GeForce GTX 960M 2GB 16GB
| 30FPS, Low, 720p
Pentium G2030 3.0GHz Radeon HD 4650 1GB 4GB
| 60FPS, High, 1080p
Core i5-6402P 2.8GHz GeForce GTX 1050 Ti MSI Gaming X 4GB 16GB