GeForce GTX 780 review - Kepler GK110-400 graphics architecture

Name: GeForce GTX 780 review
Brand: GeForce GTX 780

by Hilbert Hagedoorn on: 05/23/2013 02:58 PM [

] 157 comment(s)

Kepler GK110-400 graphics architecture

As you can understand, the massive memory partitions, bus-width and combination of GDDR5 memory (quad data rate) allow the GPU to work with a very high framebuffer bandwidth (effective). Let's again put most of the data in a chart to get an idea and better overview of changes:

Graphics card	GeForce GTX 480	GeForce GTX 580	GeForce GTX 680	GeForce GTX 780	GeForce GTX Titan
Fabrication node	40nm	40nm	28nm	28nm	28nm
Shader processors	480	512	1536	2304	2688
Streaming Multiprocessors (SMX)	15	16	8	12	14
Texture Units	60	64	128	192	224
ROP units	48	48	32	48	48
Graphics Clock (Core)	700 MHz	772 MHz	1006/1058 MHz	863/900 MHz	836/876 MHz
Shader Processor Clock	1401 MHz	1544 MHz	1006/1058 MHz	863/900 MHz	836/876 MHz
Memory Clock / Data rate	924 MHz / 3696 MHz	1000 MHz / 4000 MHz	1502 MHz / 6008 MHz	1502 MHz / 6008 MHz	1502 MHz / 6008 MHz
Graphics memory	1536 MB	1536 MB	2048 MB	3072 MB	6144 MB
Memory interface	384-bit	384-bit	256-bit	384-bit	384-bit
Memory bandwidth	177 GB/s	192 GB/s	192 GB/s	288 GB/s	288 GB/s
Power connectors	1x6-pin PEG, 1x8-pin PEG	1x6-pin PEG, 1x8-pin PEG	2x6-pin PEG	1x6-pin PEG, 1x8-pin PEG	1x6-pin PEG, 1x8-pin PEG
Max board power (TDP)	250 Watts	244 Watts	170 Watts	250 Watts	250 Watts
Recommended Power supply	600 Watts	600 Watts	550 Watts	600 Watts	600 Watts
GPU Thermal Threshold	105 degrees C	97 degrees C	98 degrees C	95 degrees C	95 degrees C

So we talked about the core clocks, specifications and memory partitions. Obviously there's a lot more to talk through. We feel that to be able to understand a graphics processor, you simply need to break it down into small pieces to better understand it. Let's first look at the raw data that most of you can understand and grasp. This bit will be about the Kepler GK110 architecture, if you're not interested in geek talk, by all means please browse to the next page.

Right so have a close look at the GK110 die as shown above. You'll notice the five green clusters. These are the polymorph GPC engines, each containing 3 SMX clusters, 5 x 3 = 15 SMX clusters in total. You'll spot six 64-bit memory interfaces, bringing in a 384-bit path towards the graphics memory. That's instant extra memory bandwith by the way, combined with a 6 Gbps clock, the cards can reach 288 GB/sec.

So above we see the GK110 block diagram that entails Kepler architecture. Let's break it down into bits and pieces. The GK110-400 will have:

2688 (Titan) or 2304 (GTX 780) CUDA processors (Shader cores)
192 CUDA cores per cluster (SMX).

The more important thing to focus on are the SM (block of shader processors) clusters (or SMX as Nvidia likes to call it for the GTX 600 series, which has 192 Shader processors.

SMX: 192 single‐precision CUDA cores, 64 double‐precision units, 32 special function units (SFU), and 32 load/store units.

When we zoom in ever further at one SMX cluster (192 shader processors) we see a change change from the GK104 (GTX 680) as there are 64 double-precision math units.

See the GeForce GTX 680 SMX had 192 single-precision (SP) floating point CUDA Cores, and 8 double-precision (DP) CUDA cores. As a result, DP operations per clock ran at effectively 1/24 the SP rate. For GTX Titan, it includes a full 64 DP CUDA Cores per SMX (compared to 192 SP CUDA Cores), or 1/3rd the number of DP cores to SP for substantially more double-precision horsepower. So based on a full 15 SMX 2880 shader cores chip the GK110 has 960 DP units linked to its total of 2,880 CUDA cores, that would be 896 DP units on tested GTX 780 with 12 activated SMXes. Double precision wise, to unlock full performance, you must open the Nvidia Control Panel, navigate to “Manage 3D Settings”. In the Global Settings box you will find an option titled “CUDA – Double Precision” which needs to be enabled, but... GeForce GTX Titan and GTX 780 will run at reduced clock speeds when full double-precision is enabled. Still a great option if you are working on CUDA applications.

The SMX has quite a bit more bite in terms of shader, texture and geometry processing. 192 CUDA cores, that's six times the number of cores per SM opposed to Fermi. In the pipeline we run into the ROP (Raster Operation) engine and the GK110 has 48 engines for features like pixel blending and AA. The GK110 has 64KB of L1 cache for each SMX plus a special 48KB texture unit memory that can be utilized as a read-only cache. L2 cache wise things remain the same across the SMX units compared to the GK104, 1.5MB. The GPU’s Texture units are a valuable resource for compute programs with a need to sample or filter image data. The texture throughput in Kepler is significantly increased compared to Fermi – each SMX unit contains 16 texture filtering units.

GeForce GTX 580 has 16 SMX x 4 Texture units = 64
GeForce GTX 680 has 8 SMX x 16 Texture units = 128
GeForce GTX 780 has 12 SMX x 16 Texture units = 192
GeForce GTX Titan has 14 SMX x 16 Texture units = 224

So there's a total 15 SMX x16 TU = 240 texture filtering units available for the GK110 silicon itself (if all SMXes where enabled). Still with me?

29 pages « 2 3 4 5 next »

Palit GeForce GTX 1630 4GB Dual review
NVIDIA has released a budget series graphics card, don't expect flying framerates, but a fun little card for entry-level gaming. Meet the GeForce GTX 1630 4GB from Palit, in a DUAL fan version....

ASUS GeForce RTX 3080 Noctua OC review
Enjoy the silence, since who doesn't remember that tune from the 1980s? Join us as we analyze the all new GeForce RTX 3080 Noctua OC model. You can dispute its appearance and style, but the card perf...

Gigabyte GeForce RTX 3090 Ti Gaming OC review
Gigabyte has released their GeForce RTX 3090 'Ti' Gaming OC. The new flagship was fitted with faster memory, a boost frequency of 1905 MHz, more shaders, and a TGP passing 450 Watts. This review ben...

ASUS GeForce RTX 3090 Ti TUF Gaming review
It's been boiling for a while, a GeForce RTX 3090 'Ti'. The 3090 flagship series now has quicker memory, more shaders, and a TGP of 450-500 Watts. In this review, we benchmark the GeForce RTX 309...