GeForce 800M series

From Wikipedia, the free encyclopedia
  (Redirected from GeForce 800 series)
Jump to: navigation, search
GeForce 800M Series
Release date March 12, 2014[1]
Codename Kepler (GK) or Maxwell (GM)
Models GeForce Series
  • GeForce GTX Series
Fabrication process and transistors 28 nm
Cards
Entry-level GeForce 820M, GeForce 830M, GeForce 840M
Mid-range GeForce GTX 850M, GeForce GTX 860M
High-end GeForce GTX 870M
Enthusiast GeForce GTX 880M
Rendering support
Direct3D full Direct3D 11.0 and partial Direct3D 12.0 support[2][3]
OpenCL OpenCL 1.1
OpenGL OpenGL 4.5[4]
History
Variant GeForce 700 series
Successor GeForce 900 series

The GeForce 800M Series is a family of graphics processing units by Nvidia for laptop PCs.[2] It consists of rebrands of mobile versions of the GeForce 700 series[2] and some newer chips that are lower end compared to the rebrands.

The GeForce 800 series name was originally planned to be used for both desktop and mobile chips based on the Maxwell microarchitecture (GM-codenamed chips), named after the Scottish theoretical physicist James Clerk Maxwell, which was previously introduced into the GeForce 700 series in the GTX 750 and GTX 750 Ti, released on February 18, 2014.[5] However, due to the fact that mobile GPUs under the GeForce 800M series have already been released using the Kepler architecture, Nvidia decided to rename its GeForce 800 series desktop GPUs as the GeForce 900 series.[2]

The Maxwell microarchitecture, the successor to Kepler microarchitecture, will for the first time feature an integrated ARM CPU of its own.[6] This will make Maxwell GPUs more independent from the main CPU according to Nvidia's CEO Jen-Hsun Huang.[7] Nvidia expects three major things from the Maxwell architecture: improved graphics capabilities, simplified programming as well as better energy-efficiency compared to the GeForce 700 Series and GeForce 600 Series [8]

Maxwell was announced in September 2010.[9] The first GeForce consumer-class products based on the Maxwell architecture were released in early 2014.[10] Nvidia is expected to release the Maxwell-powered Tesla accelerator cards as well as Quadro professional graphics cards based on this architecture in late 2014. Eventually, Maxwell architecture will be used for mobile application processors that belong to the Erista family of Tegra SoCs.[8]

Architecture[edit]

First generation Maxwell (GM10x)[edit]

Main article: Nvidia NVENC

First generation Maxwell GM107/GM108 provides few consumer-facing additional features; Nvidia instead focused on power efficiency. Nvidia's video encoder, NVENC, is 1.5 to 2 times faster than on Kepler-based GPUs meaning it can encode video at 6 to 8 times playback speed.[5] Nvidia also claims an 8 to 10 times performance increase in PureVideo Feature Set E video decoding due to the video decoder cache paired with increases in memory efficiency. However, HEVC is not supported for full hardware decoding, relying on a mix of hardware and software decoding.[5] When decoding video, a new low power state "GC5" is used on Maxwell GPUs to conserve power.[5]

Nvidia increased the amount of L2 cache on GM107 to 2 MB, up from 256 KB on GK107, reducing the memory bandwidth needed. Accordingly, Nvidia cut the memory bus to 128 bit on GM107 from 192 bit on GK106, further saving power.[5] Nvidia also changed the streaming multiprocessor design from that of Kepler (SMX), naming it SMM. The layout of SMM units is partitioned so that each of the four warp schedulers controls isolated FP32 CUDA cores, load/store units and special function units, unlike Kepler, where the warp schedulers share the resources. Texture units and FP64 CUDA cores are still shared.[5] SMM allows for a finer-grain allocation of resources than SMX, saving power when the workload isn't optimal for shared resources. Nvidia claims a 128 CUDA core SMM has 90% of the performance of a 192 CUDA core SMX.[5]

GM107/GM108 supports CUDA Compute Capability 5.0 compared to 3.5 on GK110/GK208 GPUs and 3.0 on GK10x GPUs. Dynamic Parallelism and HyperQ, two features in GK110/GK208 GPUs, are also supported across the entire Maxwell product line.

Maxwell provides native shared memory atomic operations for 32-bit integers and native shared memory 32-bit and 64-bit compare-and-swap (CAS), which can be used to implement other atomic functions.

Maxwell supports DirectX 12.[11]

Products[edit]

Products formerly placed into the GeForce 800 (8xx) Series[edit]

Main article: GeForce 900 series

Nvidia has announced that the company will be skipping the GeForce 800 series for desktop graphics cards, most likely due to the fact that the GTX 800M series consists of high-end Kepler and low-end Maxwell based components. Instead, Nvidia has announced that the newly renamed GeForce GTX 980 and GTX 970 will be introduced formally on the 19th of September 2014.[2]

GeForce 800M (8xxM) Series[edit]

The GeForce 800M Series is designed for notebooks. The processing power is obtained by multiplying shader clock speed, the number of cores and how many instructions the cores are capable of performing per cycle. Note that all GK104 based GPUs are using the older Kepler Architecture and the 820M uses GF117 cores based on the Fermi Architecture.

Model Launch Code name Fab (nm) Bus interface Memory (MiB) Core config1 Clock speed Fillrate Memory API support (version) Processing Power2
(GFLOPS)
TDP (watts) Notes
Core (MHz) Shader (MHz) Memory (MT/s) Pixel (GP/s) Texture (GT/s) Bandwidth (GB/s) Bus type Bus width (bit) DirectX OpenGL
GeForce 820M [12] February 2014 GF117 40 PCIe 2.0 x16 2048 96:16:4 719-954 1438-1908 2000 2.9-3.8 11.5-15.3 16 DDR3 64 11.0 4.5 276.1-366.3
GeForce 830M [13] March 12, 2014 GM108 28 PCIe 3.0 x16 2048 256:16:8
(2 SMM)
2000 DDR3 64 11.2 4.5
GeForce 840M [14] March 12, 2014 GM108 28 PCIe 3.0 x16 2048 384:24:8
(3 SMM)
1029 1029 2000 8.2 24.7 16 DDR3 64 11.2 4.5 790.3 30
GeForce GTX 850M [15] March 12, 2014 GM107 28 PCIe 3.0 x16 2048 640:40:16
(5 SMM)
936+Boost 936+Boost 5000 14.0 35.0 32 to 80 DDR3 or GDDR5 128 11.2 4.5 1198.1 40
GeForce GTX 860M [16] March 12, 2014 GM107
GK104
28 PCIe 3.0 x16 2048 640:40:16
(5 SMM)
1152:96:16
(6 SMX)
1029+Boost
797+Boost
1029+Boost
797+Boost
5000 16.5
12.8
41.2
76.5
80 GDDR5 128 11.2
11.0
4.5 1317.1
1836.3
40-45
75
GeForce GTX 870M [17] March 12, 2014 GK104 28 PCIe 3.0 x16 3072
6144
1344:112:24
(7 SMX)
941+Boost 941+Boost 5000 22.6 105.4 120 GDDR5 192 11.0 4.5 2529.4 100
GeForce GTX 880M [18] March 12, 2014 GK104 28 PCIe 3.0 x16 4096
8192
1536:128:32
(8 SMX)
954+Boost 954+Boost 5000 30.5 122.1 160 GDDR5 256 11.0 4.5 2930.7 105

Successor Architecture[edit]

After Maxwell, the next architecture will be codenamed Pascal.[19][20] Nvidia has announced that the Pascal GPU will feature stacked DRAM, Unified Memory, and NVLink.[19]

Chipset table[edit]

See also[edit]

References[edit]

  1. ^ Burnes, Andrew (March 12, 2014). "Introducing GTX 800M Notebook GPUs". GeForce.com. Nvidia. Retrieved August 31, 2014. 
  2. ^ a b c d e Shilov, Anton (August 29, 2014). "Nvidia to skip GeForce GTX 800 series, to introduce GeForce GTX 970, GTX 980 in mid-September". KitGuru. 
  3. ^ Kowaliski, Cyril (March 21, 2014). "DirectX 12 will also add new features for next-gen GPUs". The Tech Report. Retrieved April 1, 2014. 
  4. ^ http://www.techpowerup.com/gpudb/2534/geforce-gtx-880m.html
  5. ^ a b c d e f g Smith, Ryan; T S, Ganesh (18 February 2014). "The NVIDIA GeForce GTX 750 Ti and GTX 750 Review: Maxwell Makes Its Move". AnandTech. Archived from the original on 18 February 2014. Retrieved 18 February 2014. 
  6. ^ Nvidia Maxwell to be first GPU with ARM CPU in 2013, Guru3d.com
  7. ^ Nvidia Maxwell Graphics Processors to Have Integrated ARM General-Purpose Cores., xbitlabs.com
  8. ^ a b Nvidia: Next-Generation Maxwell Architecture Will Break New Grounds.., xbitlabs.com
  9. ^ NVIDIA reveals upcoming GPU architectures., hexus.net
  10. ^ http://videocardz.com/45403/nvidia-to-launch-more-cards-this-year-maxwell-in-q1-2014
  11. ^ http://blogs.nvidia.com/blog/2014/03/20/directx-12/
  12. ^ GeForce 820M | Specifications | GeForce
  13. ^ GeForce 830M | Specifications | GeForce
  14. ^ GeForce 840M | Specifications | GeForce
  15. ^ GeForce GTX 850M | Specifications | GeForce
  16. ^ GeForce GTX 860M | Specifications | GeForce
  17. ^ GeForce GTX 870M | Specifications | GeForce
  18. ^ GeForce GTX 880M | Specifications | GeForce
  19. ^ a b http://blogs.nvidia.com/blog/2014/03/25/gpu-roadmap-pascal/
  20. ^ Smith, Ryan (26 March 2014). "NVIDIA Updates GPU Roadmap; Unveils Pascal Architecture For 2016". AnandTech. Retrieved 26 March 2014. 
  21. ^ http://www.xbitlabs.com/news/cpu/display/20110119204601_Nvidia_Maxwell_Graphics_Processors_to_Have_Integrated_ARM_General_Purpose_Cores.html

External links[edit]