Maxwell (microarchitecture)

Last updated
Nvidia Maxwell
GTX980tiFE.jpg
NVIDIA GeForce GTX 980ti of the GeForce 900-line of graphics-cards, was the final major iteration featuring the Maxwell microarchitecture (GM200-310).
Release dateFebruary 2014
Fabrication process TSMC 28 nm, 20 nm, 16 nm
History
Predecessor Kepler
Successor Pascal
Support status
Supported
Photo of James Clerk Maxwell, eponym of architecture James Clerk Maxwell.png
Photo of James Clerk Maxwell, eponym of architecture

Maxwell is the codename for a GPU microarchitecture developed by Nvidia as the successor to the Kepler microarchitecture. The Maxwell architecture was introduced in later models of the GeForce 700 series and is also used in the GeForce 800M series, GeForce 900 series, and Quadro Mxxx series, as well as some Jetson products.

Contents

The first Maxwell-based products were the GeForce GTX 745 (OEM), GeForce GTX 750, and the GeForce GTX 750 Ti. Both were released on February 18, 2014, both with the chip code number GM107. Earlier GeForce 700 series GPUs had used Kepler chips with the code numbers GK1xx. First-generation Maxwell GPUs (code numbers GM10x) are also used in the GeForce 800M series and the Quadro Kxxx series. A second generation of Maxwell-based products was introduced on September 18, 2014 with the GeForce GTX 970 and GeForce GTX 980, followed by the GeForce GTX 960 on January 22, 2015, the GeForce GTX Titan X on March 17, 2015, and the GeForce GTX 980 Ti on June 1, 2015. The final and lowest spec Maxwell 2.0 card was the GTX950 released on Aug 20th, 2015. These GPUs have GM20x chip code numbers.

Maxwell introduced an improved Streaming Multiprocessor (SM) design that increased power efficiency, [1] the sixth and seventh generation PureVideo HD, and CUDA Compute Capability 5.2.

The architecture is named after James Clerk Maxwell, the founder of the theory of electromagnetic radiation.

The Maxwell architecture is used in the system on a chip (SOC), mobile application processor, Tegra X1.

First generation Maxwell (GM10x)

Maxwell 107 chip on GTX 750 Ti graphics card with heatsink removed Nvidia maxwell-chip.jpg
Maxwell 107 chip on GTX 750 Ti graphics card with heatsink removed

First generation Maxwell GPUs (GM107/GM108) were released as GeForce GTX 745, GTX 750/750 Ti, GTX 850M/860M (GM107) and GeForce 830M/840M (GM108). These new chips introduced few consumer-facing additional features, as Nvidia instead focused more on increasing GPU power efficiency. The L2 cache was increased from 256 KiB on Kepler to 2 MiB on Maxwell, reducing the need for more memory bandwidth. Accordingly, the memory bus was reduced from 192 bit on Kepler (GK106) to 128 bit, reducing die area, cost, and power draw. [2]

The "SMX" streaming multiprocessor design from Kepler was also retooled and partitioned, being renamed "SMM" for Maxwell. The structure of the warp scheduler was inherited from Kepler, with the texture units and FP64 CUDA cores still shared, but the layout of most execution units were partitioned so that each warp schedulers in an SMM controls one set of 32 FP32 CUDA cores, one set of 8 load/store units and one set of 8 special function units. This is in contrast to Kepler, where each SMX had 4 schedulers that scheduled to a shared pool of execution units. [3] The latter necessitated an SMX-wide crossbar that used unnecessary power to allow all execution units to be shared. [3] Conversely, Maxwell's more modular design allows for a finer-grained and more efficient allocation of resources, 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 while efficiency increases by a factor of 2. [2] Also, each Graphics Processing Cluster, or GPC, contains up to 4 SMX units in Kepler, and up to 5 SMM units in first generation Maxwell. [2]

GM107 also 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 also 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.

Nvidia's video encoder, NVENC, was upgraded to be 1.5 to 2 times faster than on Kepler-based GPUs, meaning it can encode video at six to eight times playback speed. [2] Nvidia also claims an eight to ten times performance increase in PureVideo Feature Set E video decoding due to the video decoder cache, paired with increases in memory efficiency. However, H.265 is not supported for full hardware decoding in first generation Maxwell GPUs, relying on a mix of hardware decoding and software decoding (CPU decoding). [2] When decoding video, a new low power state "GC5" is used on Maxwell GPUs to conserve power. [2]

Maxwell GPUs were thought to use tile-based rendering, [4] but they actually use tiled caching. [5]

Since first generation Maxwell, UEFI Graphics Output Protocol is fully supported on NVIDIA GPUs.

Chips

Second generation Maxwell (GM20x)

Die shot of the GM200 GPU found inside GeForce GTX 980 Ti cards Nvidia@28nm@Maxwell@GM200@GeForce GTX 980 Ti@A TAIWAN 1542A1 TFH800.COW GM200-310-A1 DSCx3 polysilicon@5x.jpg
Die shot of the GM200 GPU found inside GeForce GTX 980 Ti cards

Second generation Maxwell GPUs introduced several new technologies: Dynamic Super Resolution, [6] Third Generation Delta Color Compression, [7] Multi-Pixel Programming Sampling, [8] Nvidia VXGI (Real-Time-Voxel-Global Illumination), [9] VR Direct, [9] [10] [11] Multi-Projection Acceleration, [7] Multi-Frame Sampled Anti-Aliasing(MFAA) [12] (however, support for Coverage-Sampling Anti-Aliasing(CSAA) was removed), [13] and Direct3D12 API at Feature Level 12_1. HDMI 2.0 support was also added. [14] [15]

The ROP to memory controller ratio was changed from 8:1 to 16:1. [16] However, some of the ROPs are generally idle in the GTX 970 because there are not enough enabled SMMs to give them work to do, reducing its maximum fill rate. [17]

The Polymorph Engine responsible for tessellation was upgraded to version 3.0 in second generation Maxwell GPUs, resulting in improved tessellation performance per unit/clock.

Second generation Maxwell also has up to 4 SMM units per GPC, compared to 5 SMM units per GPC. [16]

GM204 supports CUDA Compute Capability 5.2 (compared to 5.0 on GM107/GM108 GPUs, 3.5 on GK110/GK208 GPUs and 3.0 on GK10x GPUs). [7] [16] [18]

GM20x GPUs have an upgraded NVENC which supports HEVC encoding and adds support for H.264 encoding resolutions at 1440p/60FPS & 4K/60FPS (compared to NVENC on Maxwell first generation GM10x GPUs which only supported H.264 1080p/60FPS encoding). [11]

After consumer complaints, [19] Nvidia revealed that it is able to disable individual units, each containing 256KB of L2 cache and 8 ROPs, without disabling whole memory controllers. [20] This comes at the cost of dividing the memory bus into high speed and low speed segments that cannot be accessed at the same time for reads, because the L2/ROP unit managing both of the GDDR5 controllers shares the read return channel and the write data bus between the GDDR5 controllers. This makes simultaneous reading from both GDDR5 controllers or simultaneous writing to both GDDR5 controllers impossible. [20] This is used in the GeForce GTX 970, which therefore can be described as having 3.5 GB in a high-speed segment on a 224-bit bus and 512 MB in a low-speed segment on a 32-bit bus. [20] The peak speed of such a GPU can still be attained, but the peak speed figure is only reachable if one segment is executing a read operation while the other segment is executing a write operation. [20]

Chips

Performance

The theoretical single-precision processing power of a Maxwell GPU in FLOPS is computed as 2 (operations per FMA instruction per CUDA core per cycle) × number of CUDA cores × core clock speed (in Hz).

The theoretical double-precision processing power of a Maxwell GPU is 1/32 of the single precision performance (which has been noted as being very low compared to the previous generation Kepler). [21]

Successor

The successor to Maxwell is codenamed Pascal. [22] The Pascal architecture features higher bandwidth unified memory and NVLink. [22]

See also

Related Research Articles

<span class="mw-page-title-main">GeForce</span> Brand of GPUs by Nvidia

GeForce is a brand of graphics processing units (GPUs) designed by Nvidia and marketed for the performance market. As of the GeForce 40 series, there have been eighteen iterations of the design. The first GeForce products were discrete GPUs designed for add-on graphics boards, intended for the high-margin PC gaming market, and later diversification of the product line covered all tiers of the PC graphics market, ranging from cost-sensitive GPUs integrated on motherboards, to mainstream add-in retail boards. Most recently, GeForce technology has been introduced into Nvidia's line of embedded application processors, designed for electronic handhelds and mobile handsets.

<span class="mw-page-title-main">Quadro</span> Brand of Nvidia graphics cards used in workstations

Quadro was Nvidia's brand for graphics cards intended for use in workstations running professional computer-aided design (CAD), computer-generated imagery (CGI), digital content creation (DCC) applications, scientific calculations and machine learning from 2000 to 2020.

<span class="mw-page-title-main">GeForce 8 series</span> Series of GPUs by Nvidia

The GeForce 8 series is the eighth generation of Nvidia's GeForce line of graphics processing units. The third major GPU architecture developed by Nvidia, Tesla represents the company's first unified shader architecture.

<span class="mw-page-title-main">Tesla (microarchitecture)</span> GPU microarchitecture by Nvidia

Tesla is the codename for a GPU microarchitecture developed by Nvidia, and released in 2006, as the successor to Curie microarchitecture. It was named after the pioneering electrical engineer Nikola Tesla. As Nvidia's first microarchitecture to implement unified shaders, it was used with GeForce 8 series, GeForce 9 series, GeForce 100 series, GeForce 200 series, and GeForce 300 series of GPUs, collectively manufactured in 90 nm, 80 nm, 65 nm, 55 nm, and 40 nm. It was also in the GeForce 405 and in the Quadro FX, Quadro x000, Quadro NVS series, and Nvidia Tesla computing modules.

PureVideo is Nvidia's hardware SIP core that performs video decoding. PureVideo is integrated into some of the Nvidia GPUs, and it supports hardware decoding of multiple video codec standards: MPEG-2, VC-1, H.264, HEVC, and AV1. PureVideo occupies a considerable amount of a GPU's die area and should not be confused with Nvidia NVENC. In addition to video decoding on chip, PureVideo offers features such as edge enhancement, noise reduction, deinterlacing, dynamic contrast enhancement and color enhancement.

The Evergreen series is a family of GPUs developed by Advanced Micro Devices for its Radeon line under the ATI brand name. It was employed in Radeon HD 5000 graphics card series and competed directly with Nvidia's GeForce 400 series.

<span class="mw-page-title-main">GeForce 400 series</span> Series of GPUs by Nvidia

The GeForce 400 series is a series of graphics processing units developed by Nvidia, serving as the introduction of the Fermi microarchitecture. Its release was originally slated in November 2009, however, after delays, it was released on March 26, 2010, with availability following in April 2010.

<span class="mw-page-title-main">GeForce 500 series</span> Series of GPUs by Nvidia

The GeForce 500 series is a series of graphics processing units developed by Nvidia, as a refresh of the Fermi based GeForce 400 series. It was first released on November 9, 2010 with the GeForce GTX 580.

<span class="mw-page-title-main">GeForce 600 series</span> Series of GPUs by Nvidia

The GeForce 600 series is a series of graphics processing units developed by Nvidia, first released in 2012. It served as the introduction of the Kepler architecture. It is succeeded by the GeForce 700 series.

<span class="mw-page-title-main">GeForce 700 series</span> Series of GPUs by Nvidia

The GeForce 700 series is a series of graphics processing units developed by Nvidia. While mainly a refresh of the Kepler microarchitecture, some cards use Fermi (GF) and later cards use Maxwell (GM). GeForce 700 series cards were first released in 2013, starting with the release of the GeForce GTX Titan on February 19, 2013, followed by the GeForce GTX 780 on May 23, 2013. The first mobile GeForce 700 series chips were released in April 2013.

<span class="mw-page-title-main">Fermi (microarchitecture)</span> GPU microarchitecture by Nvidia

Fermi is the codename for a graphics processing unit (GPU) microarchitecture developed by Nvidia, first released to retail in April 2010, as the successor to the Tesla microarchitecture. It was the primary microarchitecture used in the GeForce 400 series and GeForce 500 series. All desktop Fermi GPUs were manufactured in 40nm, mobile Fermi GPUs in 40nm and 28nm. Fermi is the oldest microarchitecture from NVIDIA that received support for Microsoft's rendering API Direct3D 12 feature_level 11.

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

<span class="mw-page-title-main">GeForce 900 series</span> Series of GPUs by Nvidia

The GeForce 900 series is a family of graphics processing units developed by Nvidia, succeeding the GeForce 700 series and serving as the high-end introduction to the Maxwell microarchitecture, named after James Clerk Maxwell. They are produced with TSMC's 28 nm process.

<span class="mw-page-title-main">GeForce 10 series</span> Series of GPUs by Nvidia

The GeForce 10 series is a series of graphics processing units developed by Nvidia, initially based on the Pascal microarchitecture announced in March 2014. This design series succeeded the GeForce 900 series, and is succeeded by the GeForce 16 series and GeForce 20 series using the Turing microarchitecture.

<span class="mw-page-title-main">Kepler (microarchitecture)</span> GPU microarchitecture by Nvidia

Kepler is the codename for a GPU microarchitecture developed by Nvidia, first introduced at retail in April 2012, as the successor to the Fermi microarchitecture. Kepler was Nvidia's first microarchitecture to focus on energy efficiency. Most GeForce 600 series, most GeForce 700 series, and some GeForce 800M series GPUs were based on Kepler, all manufactured in 28 nm. Kepler found use in the GK20A, the GPU component of the Tegra K1 SoC, and in the Quadro Kxxx series, the Quadro NVS 510, and Nvidia Tesla computing modules.

<span class="mw-page-title-main">Pascal (microarchitecture)</span> GPU microarchitecture by Nvidia

Pascal is the codename for a GPU microarchitecture developed by Nvidia, as the successor to the Maxwell architecture. The architecture was first introduced in April 2016 with the release of the Tesla P100 (GP100) on April 5, 2016, and is primarily used in the GeForce 10 series, starting with the GeForce GTX 1080 and GTX 1070, which were released on May 17, 2016, and June 10, 2016, respectively. Pascal was manufactured using TSMC's 16 nm FinFET process, and later Samsung's 14 nm FinFET process.

Nvidia NVENC is a feature in Nvidia graphics cards that performs video encoding, offloading this compute-intensive task from the CPU to a dedicated part of the GPU. It was introduced with the Kepler-based GeForce 600 series in March 2012.

<span class="mw-page-title-main">Turing (microarchitecture)</span> GPU microarchitecture by Nvidia

Turing is the codename for a graphics processing unit (GPU) microarchitecture developed by Nvidia. It is named after the prominent mathematician and computer scientist Alan Turing. The architecture was first introduced in August 2018 at SIGGRAPH 2018 in the workstation-oriented Quadro RTX cards, and one week later at Gamescom in consumer GeForce 20 series graphics cards. Building on the preliminary work of Volta, its HPC-exclusive predecessor, the Turing architecture introduces the first consumer products capable of real-time ray tracing, a longstanding goal of the computer graphics industry. Key elements include dedicated artificial intelligence processors and dedicated ray tracing processors. Turing leverages DXR, OptiX, and Vulkan for access to ray tracing. In February 2019, Nvidia released the GeForce 16 series GPUs, which utilizes the new Turing design but lacks the RT and Tensor cores.

Ada Lovelace, also referred to simply as Lovelace, is a graphics processing unit (GPU) microarchitecture developed by Nvidia as the successor to the Ampere architecture, officially announced on September 20, 2022. It is named after the English mathematician Ada Lovelace, one of the first computer programmers. Nvidia announced the architecture along with the GeForce RTX 40 series consumer GPUs and the RTX 6000 Ada Generation workstation graphics card. The Lovelace architecture is fabricated on TSMC's custom 4N process which offers increased efficiency over the previous Samsung 8 nm and TSMC N7 processes used by Nvidia for its previous-generation Ampere architecture.

References

  1. ↑ "5 Things You Should Know About the New Maxwell GPU Architecture". 2014-02-21.
  2. 1 2 3 4 5 6 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.
  3. 1 2 Ryan Smith, Ganesh T S. "Maxwell: Designed For Energy Efficiency - The NVIDIA GeForce GTX 750 Ti and GTX 750 Review: Maxwell Makes Its Move".
  4. ↑ Kanter, David (August 1, 2016). "Tile-based Rasterization in Nvidia GPUs". Real World Technologies. Retrieved April 1, 2016.
  5. ↑ "On NVIDIA's Tile-Based Rendering". Tech Power Up. March 1, 2017. Retrieved May 9, 2020.
  6. ↑ "Dynamic Super Resolution Improves Your Games With 4K-Quality Graphics On HD Monitors". www.nvidia.com.
  7. 1 2 3 "Archived copy" (PDF). Archived from the original (PDF) on 2017-07-21. Retrieved 2014-09-19.{{cite web}}: CS1 maint: archived copy as title (link)
  8. ↑ "NVIDIA BatteryBoost: Ditch the Brick". NVIDIA.
  9. 1 2 "GeForce Articles, Guides, Gaming News, Featured Stories". www.nvidia.com.
  10. ↑ "How Maxwell's VR Direct Brings Virtual Reality Gaming Closer to Reality". The Official NVIDIA Blog.
  11. 1 2 Ryan Smith. "Display Matters: HDMI 2.0, HEVC, & VR Direct - The NVIDIA GeForce GTX 980 Review: Maxwell Mark 2".
  12. ↑ "Multi-Frame Sampled Anti-Aliasing Delivers Better Performance To Maxwell Gamers". www.nvidia.com.
  13. ↑ "New nVidia Maxwell chips do not support fast CSAA".
  14. ↑ "GeForce RTX 20 Series Graphics Cards and Laptops". NVIDIA.
  15. ↑ Ryan Smith. "The NVIDIA GeForce GTX 980 Review: Maxwell Mark 2".
  16. 1 2 3 Ryan Smith. "Maxwell 2 Architecture: Introducing GM204 - The NVIDIA GeForce GTX 980 Review: Maxwell Mark 2".
  17. ↑ "Here's another reason the GeForce GTX 970 is slower than the GTX 980". The Tech Report.
  18. ↑ "Maxwell: The Most Advanced CUDA GPU Ever Made". Parallel Forall. 19 September 2014.
  19. ↑ Geoffrey Tim (3 December 2015). "Nvidia's GTX970 has a rather serious memory allocation bug". Lazygamer.net.
  20. 1 2 3 4 Ryan Smith. "Diving Deeper: The Maxwell 2 Memory Crossbar & ROP Partitions - GeForce GTX 970: Correcting The Specs & Exploring Memory Allocation".
  21. ↑ Smith, Ryan (17 March 2015). "The NVIDIA GeForce GTX Titan X Review". AnandTech . p. 2. Retrieved 6 December 2015. ...puny native FP64 rate of just 1/32
  22. 1 2 "NVIDIA Updates GPU Roadmap; Announces Pascal". The Official NVIDIA Blog.
This page is based on this Wikipedia article
Text is available under the CC BY-SA 4.0 license; additional terms may apply.
Images, videos and audio are available under their respective licenses.