Guru3D.com
  • HOME
  • NEWS
    • Channels
    • Archive
  • DOWNLOADS
    • New Downloads
    • Categories
    • Archive
  • GAME REVIEWS
  • ARTICLES
    • Rig of the Month
    • Join ROTM
    • PC Buyers Guide
    • Guru3D VGA Charts
    • Editorials
    • Dated content
  • HARDWARE REVIEWS
    • Videocards
    • Processors
    • Audio
    • Motherboards
    • Memory and Flash
    • SSD Storage
    • Chassis
    • Media Players
    • Power Supply
    • Laptop and Mobile
    • Smartphone
    • Networking
    • Keyboard Mouse
    • Cooling
    • Search articles
    • Knowledgebase
    • More Categories
  • FORUMS
  • NEWSLETTER
  • CONTACT

New Reviews
be quiet Pure Loop 2 FX 280mm LCS review
HP FX900 1 TB NVMe Review
Scythe FUMA2 Rev.B CPU Cooler review
SK Hynix Platinum P41 2TB M.2 NVMe SSD Review
Corsair K70 RGB PRO Mini Wireless review
MSI MPG A1000G - 1000W PSU Review
Goodram IRDM PRO M.2 SSD 2 TB NVMe SSD Review
Samsung T7 Shield Portable 1TB USB SSD review
DeepCool LS720 (LCS) review
Fractal Design Pop Air RGB Black TG review

New Downloads
Display Driver Uninstaller Download version 18.0.5.4
FurMark Download v1.31
Intel HD graphics Driver Download Version: 31.0.101.3222
Intel ARC graphics Driver Download Version: 30.0.101.1743
AMD Radeon Software Adrenalin 22.7.1 driver download
GeForce 516.93 WHQL Studio driver download
Corsair Utility Engine Download (iCUE) Download v4.26.110
ReShade download v5.3.0
AIDA64 Download Version 6.75
7-Zip v22.01 Download


New Forum Topics
Latest Nvidia Drivers trigger multiple WMI Errors on PC startup and wake up! At Least One AIB Halts Production ARC Graphics cards - Is ARC Doomed? MSI Shows Custom Intel Arc A380 Graphics Card Intel Core i9-13900K with and without power management settings MSI G2422 24-inch full HD LCD for eSports, compatible with FreeSync Premium EK Releases New Delta2 TEC for LGA 1700 and Intel 12th Gen Core CPUs NVIDIA Studio Driver 516.93 WHQL 760 GTX Need Attention-Scrap Razer released gaming chairs inspired by Williams and Koenigsegg (at 1299 USD) Enabling ReBar without accessing BIOS?




Guru3D.com » Review » Inno3D GeForce GTX 760 iChill review » Page 2

Inno3D GeForce GTX 760 iChill review - The Technology and Specs

by Hilbert Hagedoorn on: 06/26/2013 08:17 AM [ 4] 4 comment(s)

Tweet

 

Technology and Specs

So then, it's time to talk business. The GeForce GTX 680 GeForce or GTX 670 GeForce GTX 760 being reviewed today is based on Kepler GPU architecture, which we all are familiar with by now. The GeForce GTX 760 is based on the 28nm GK104 GPU, the same as the GTX 670 and 680 uses, yet with a few shader clusters disabled. Still, the 10" long GeForce GTX 760 boasts a good 1152 CUDA (shader processors) cores. The product is obviously PCI-Express 3.0 ready and has a TDP of give or take a typical draw of 150~160W. But let me first show you the actual GK104 die:



NVIDIA GK104 Kepler architecture GPU used in the Geforce GTX 760 and 770
 

As far as the memory specs of the GK104 Kepler GPU are concerned, the boards will feature a 256-bit memory bus connected to 2 GB or alternatively 4 GB of GDDR5 video buffer memory. On the memory controller side of things you'll see very significant improvements as the reference memory clock is now set at 6 GHz / Gbps. This boils down to a memory bandwidth of 192GB/s on that 256-bit memory bus. Both the GPU core and the shader processor domain are clocked at 1:1, meaning both the core and shader domain clock in at a 1046 MHz base clock. With this release, NVIDIA now has the first series 700 cards on its way. The new graphics adapters are of course DirectX 11.1 ready. With Windows 8, 7 and Vista also being DX11.1 ready all we need are some games to take advantage of DirectCompute, multi-threading, hardware tessellation and the latest shader 5.0 extensions. For Kepler, NVIDIA kept their memory controllers GDDR5 compatible. Memory wise NVIDIA has nice large memory volumes due to their architecture, we pass 2 GB as standard these days.

 

The graphics architecture that is Kepler

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 680
GeForce
GTX 760
GeForce
GTX 770
GeForce
GTX 780
GeForce
GTX Titan
Fabrication node 28nm 28nm 28nm 28nm 28nm
Shader processors 1536 1152 1536 2304 2688
Streaming Multiprocessors (SMX) 8 6 8 12 14
Texture Units 128 96 128 192 224
ROP units 32 32 32 48 48
Graphics Clock (Core) 1006 MHz 980 MHz 1046 863 836
Boost Processor Clock 1058 MHz 1033 MHz 1085 MHz 900 MHz 876 MHz
Memory Clock / Data rate  MHz / 6008 MHz 1502 MHz / 6008 MHz 1750 MHz / 7000 MHz 1502 MHz / 6008 MHz 1502 MHz / 6008 MHz
Graphics memory 2048 MB 2048 MB 2048 MB 3072 MB 6144 MB
Memory interface 256-bit 256-bit 256-bit 384-bit 384-bit
Memory bandwidth 192 GB/s 192 GB/s 224 GB/s 288 GB/s 288 GB/s
Power connectors 2x6-pin PEG  2x6-pin PEG 1x6-pin PEG, 1x8-pin PEG 1x6-pin PEG, 1x8-pin PEG 1x6-pin PEG, 1x8-pin PEG
Max board power (TDP) 170 Watts 170 Watts 230 Watts 250 Watts 250 Watts
Recommended Power supply 550 Watts 550 Watts 600 Watts 600 Watts 600 Watts
GPU Thermal Threshold 98 degrees C 98 degrees C 95 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. To understand a graphics processor you simply need to break it down into smaller 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 architecture, if you're not interested in g33k talk by all means please browse to the next page.
 

GeForce GTX 680

 

So above we see the GK104 block diagram that entails the Kepler architecture. Let's break it down into bits and pieces. The reference GK104 will have:

  • 1536 CUDA processors (Shader cores)
  • 192 CUDA core clusters (SM/SMX).
  • 8 geometry units
  • 4 raster Units
  • 128 Texture Units
  • 32 ROP engines
  • 256-bit GDDR5 memory bus
  • DirectX 11.1

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 680/770, which  has 192 Shader processors. That's radically different from Fermi, the GeForce GTX 580 for example had 32 shader processors per SM cluster. 1536 : 192 = 8 Shader clusters (SMs). Let's blow up one such cluster:

 GeForce GTX 680

Above the block diagram for a single Shader processor cluster, aka SM or SMX as NVIDIA now calls it. The new 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. Now, at the end of the pipeline we run into the ROP (Raster Operation) engine and the GTX 680 again has 32 engines for features like pixel blending and AA. There's a total of 128 texture filtering units available for the GeForce GTX 680. The math is simple here, each SM has 16 texture units tied to it.

  • GeForce GTX 580 has 16 SMs X 4 Texture units = 64
  • GeForce GTX 680 & 770 have 8 SMs X 16 Texture units = 128
  • GeForce GTX 760 have 6 SMs X 16 Texture units = 96

Above the GK104 host interface - The Gigathread engine, four GPCs, four memory controllers, the ROP partitions, a 768 KB L2 cache. Each GPC has eight polymorph engines - ROP partitions are nearby to the L2 cache, each shader cluster is then tied to L1 and a shared L2 cache. Shading performance is going be increased quite bit, geometry performance will get a nice boost as well. NVIDIA is using 64KB Shared Memory/L1 per SMX – please note that they have a 16/48 – 48/16 ratio here for graphics/compute, as before with Fermi. For L2, 128KB per 64-bit memory controller. So that adds up to 512KB L2. In regards to architectural changes, on top of the pipeline NVIDIA has now added new Polymorph 2.0 (world space processing) engines and raster (screen space processing) engines; they act like a mini CPU really.




25 pages 1 2 3 4 next »



Related Articles
Inno3D GeForce GTX 770 iChill HerculeZ X3 Ultra review
In this article we review the beast, the GTX 770 HerculeZ X3 Ultra from Inno3D is locked and loaded with the mighty HerculeZ X3 cooler, that's three slots wide but keeping the GPU under 55 Degrees C and completely silent. Pure awesomeness. That's not all though, it is factory overclocked for you at a very high level but even so can be tweaked quite a bit further.

Inno3D GeForce GTX 760 iChill review
In this article we review the Inno3D GeForce GTX 760 iChill edition, this particular model comes with a three-slot HerleZ cooing solution.

Inno3D GeForce GTX 660 Ti iChill review
In this review we'll look at the GeForce GTX 660 Ti from Inno3D, it's their all new GeForce GTX 660 Ti iCHILL version and to date is one of the most impressive graphics cards in the 660 Ti range we have tested.

Inno3D GeForce GTX 580 OC review
We review the Inno3D GeForce GTX 580OC. Despite a very high price tag the product seems to become a nice success. As such directly at launch several models based of this SKU where already announced, e.g. the regular clocked models, factory higher clocked models, liquid cooled models. This OC edition, in particular this is a reference GeForce GTX 580 that has been clocked faster to 820 MHz on the core where it also welcomes a nice bump on the overall memory frequency.

© 2022