Engineering:AMD RX Vega series

From HandWiki
AMD RX Vega
AMD Radeon graphics logo 2016
Release date2017
Codename
  • Vega
ArchitectureGCN 5th gen
Cards
High-endRX Vega 56
RX Vega 64
EnthusiastRX Vega 64 Liquid
API support
Direct3D
  • Direct3D 12.0 (feature level 12 1)[1]
  • Shader Model 6.0
OpenCLOpenCL 2.0[1]
OpenGLOpenGL 4.6[1][2][3]
Vulkan
History
PredecessorRadeon 500 Series
SuccessorNavi Architecture

The AMD RX Vega series is a series of consumer graphics processors produced by AMD. These cards use the Vega architecture, feature HBM2 high-speed memory[7] and are based on the 14nm FinFET architecture.[8] Vega is the first to use the new Graphics Core Next 5th generation and was slated to be succeeded by the Navi architecture[9] in 2018, but in the new roadmap, changed to 2019. The lineup was released on the 14th of August 2017. It included the RX Vega 56 and the RX Vega 64, priced at $399 and $499 respectively.[10]

History

The Vega microarchitecture is AMD's high-end graphics cards line,[11] and is the successor to the R9 300 series enthusiast Fury products. Partial specifications of the architecture and Vega 10 GPU were announced with the Radeon Instinct MI25 in December 2016.[12] AMD later teased details of the Vega architecture.

Announcement

Vega was originally announced at AMD's CES 2017 presentation on January 5, 2017,[13] alongside the Zen line of CPUs.[14]

New features

Vega targets increased instructions per clock, higher clock speeds, and support for HBM2.[15][16][17]

AMD's Vega features new memory hierarchy with high-bandwidth cache and its controller

Support for HBM2 featuring double the bandwidth-per-pin over previous generation HBM. HBM2 allows for higher capacities with less than half the footprint of GDDR5 memory. Vega architecture is optimized for streaming very large datasets and can work with a variety of memory types with up to 512TB of virtual address space.

Primitive shader for improved geometry processing. Replaces vertex and geometry shaders in geometry processing pipelines with a more programmable single stage. The primitive shader stage is more efficient, introduces intelligent load balancing technologies and higher throughput.[18]

NCU: Next Compute Unit a Next-generation compute engine. The Vega GPU introduces the Next-Gen Compute Unit. Versatile architecture featuring flexible compute units that can natively process 8-bit, 16-bit, 32-bit or 64-bit operations in each clock cycle. And run at higher frequencies. Vega brings support for Rapid Packed Math, processing two half-precision (16-bit) in the same time as a single 32-bit floating-point operation. Up to 128 32-bit, 256 16-bit or 512 8-bit ops per clock are possible with the Vega architecture.[19]

Draw Stream Binning Rasterizer designed for higher performance and power efficiency. It allows for “fetch once, shade once” of pixels through the use of a smart on-chip bin cache and early culling of pixels invisible in a final scene.

Vega bumps Direct3D feature level support from 12_0 to 12_1.

Vega's rasteriser brings hardware-acceleration support for Rasterizer Ordered Views and Conservative Rasterisation Tier 3.[20]

Chipset table

Desktop

Model
(codename)
Release Date
& Price
Architecture
Fab
Transistors
& Die Size
Core Fillrate[lower-alpha 1][lower-alpha 2][lower-alpha 3] Processing power[lower-alpha 1][lower-alpha 4]
(GFLOPS)
Memory TBP Bus
interface
Config[lower-alpha 5] Clock[lower-alpha 1] (MHz) Texture
(GT/s)
Pixel
(GP/s)
Half Single Double Bus type
& width
Size
(GiB)
Clock
(MT/s)
Bandwidth
(GB/s)
Radeon RX Vega 56
(Vega10 XL)[21][22][23]
2017
$399 USD
GCN 5th gen
Samsung/GloFo
14LPP[24][lower-alpha 6]
12.5×109
486 mm2
3584:224:64
56 CU
1156
1471
258.9
329.5
74.0
94.1
16572
21088
8286
10544
518
659
HBM2
2048-bit
8 1600 410 210 W PCIe 3.0
×16
Radeon RX Vega 64
(Vega10 XT)[21][23][26]
2017
$499 USD
4096:256:64
64 CU
1247
1546
319.2
395.8
79.8
98.9
20431
25330
10215
12665
638
792
1890 483.8 295 W
Radeon RX Vega 64 Liquid
(Vega10 XT)[21][23][26]
2017
$699 USD
1406
1677
359.9
429.3
90.0
107.3
23036
27476
11518
13738
720
859
345 W

v · d · e

  1. 1.0 1.1 1.2 Boost values (if available) are stated below the base value in italic.
  2. Texture fillrate is calculated as the number of Texture Mapping Units multiplied by the base (or boost) core clock speed.
  3. Pixel fillrate is calculated as the number of Render Output Units multiplied by the base (or boost) core clock speed.
  4. Precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.
  5. Unified Shaders : Texture Mapping Units : Render Output Units and Compute Units (CU)
  6. GlobalFoundries' 14 nm 14LPP FinFET process is second-sourced from Samsung Electronics.[25]

Radeon features

See also

References

  1. 1.0 1.1 1.2 "Radeon Software Crimson ReLive Edition 17.8.1 Release Notes". AMD. https://support.amd.com/en-us/kb-articles/Pages/Radeon-Software-Crimson-ReLive-Edition-17.8.1-Release-Notes.aspx. Retrieved 2018-04-20. 
  2. "AMDGPU-PRO Driver for Linux Release Notes". 2017. Archived from the original on 2017-08-15. https://web.archive.org/web/20170815061617/https://support.amd.com/en-us/kb-articles/Pages/AMDGPU-PRO-Driver-for-Linux-Release-Notes.aspx. Retrieved 2018-04-20. 
  3. "Mesamatrix". mesamatrix.net. https://mesamatrix.net/. Retrieved 2018-04-22. 
  4. "Radeon Software Adrenalin Edition 18.3.4 Release Notes". AMD. https://support.amd.com/en-us/kb-articles/Pages/Radeon-Software-Adrenalin-Edition-18.3.4-Release-Notes.aspx. Retrieved 2018-04-20. 
  5. "Radeon™ Software for Linux® with Vulkan® 1.1 support". AMD. https://support.amd.com/en-us/kb-articles/Pages/Radeon-Software-for-Linux-with-Vulkan-1.1-support.aspx. Retrieved 2018-04-21. 
  6. "AMD Open Source Driver for Vulkan". GPUOpen. https://gpuopen.com/gaming-product/amd-open-source-driver-for-vulkan/. Retrieved 2018-04-20. 
  7. "AMD Radeon Vega Frontier Edition Unveiled: 13 TFLOPs, 16GB HBM2, Ships In June" (in en-us). http://hothardware.com/news/amd-radeon-rx-vega-unveiled-13-tflops-16gb-hbm2-4k60. 
  8. "Radeon Vega Frontier Edition: First AMD Vega GPU is for the pros" (in en-us). Ars Technica. Archived from the original on 2017-05-17. https://web.archive.org/web/20170517131524/https://arstechnica.com/gadgets/2017/05/radeon-vega-gpu-details-price/. 
  9. Smith, Ryan. "AMD Updates GPU Architecture Roadmap: After Navi Comes “Next Gen”". Archived from the original on 2017-05-17. https://web.archive.org/web/20170517102933/http://www.anandtech.com/show/11404/amd-updates-gpu-architecture-roadmap-after-navi-comes-next-gen. 
  10. "AMD's Radeon RX Vega 64 will cost $499 and battle the GeForce GTX 1080" (in en). PCWorld. Archived from the original on 2017-08-01. https://web.archive.org/web/20170801011827/http://www.pcworld.com/article/3212206/computers/radeon-rx-vega-64-pricing.html. 
  11. Smith, Ryan (5 January 2017). "The AMD Vega GPU Architecture Teaser". Anandtech.com. http://www.anandtech.com/show/11002/the-amd-vega-gpu-architecture-teaser. Retrieved 30 June 2017. 
  12. Shrout, Ryan (12 December 2016). "Radeon Instinct Machine Learning GPUs include Vega, Preview Performance". PC Per. https://www.pcper.com/reviews/Graphics-Cards/Radeon-Instinct-Machine-Learning-GPUs-include-Vega-Preview-Performance. Retrieved 12 December 2016. 
  13. "Vega: AMD's New Graphics Architecture for Virtually Unlimited Workloads" (in en-US). Archived from the original on 2017-06-25. https://web.archive.org/web/20170625194359/http://www.amd.com/en-us/press-releases/Pages/vega-amds-new-2017jan05.aspx. 
  14. "AMD Ryzen CPUs: 7 all-new details revealed at CES 2017" (in en). PCWorld. http://www.pcworld.com/article/3155496/computers/amd-ryzen-cpus-7-all-new-details-revealed-at-ces-2017.html. 
  15. Kampman, Jeff (5 January 2017). "The curtain comes up on AMD's Vega architecture". TechReport.com. http://techreport.com/review/31224/the-curtain-comes-up-on-amd-vega-architecture. Retrieved 10 January 2017. 
  16. Shrout, Ryan (5 January 2017). "AMD Vega GPU Architecture Preview: Redesigned Memory Architecture". PC Perspective. https://www.pcper.com/reviews/Graphics-Cards/AMD-Vega-GPU-Architecture-Preview-Redesigned-Memory-Architecture. Retrieved 10 January 2017. 
  17. Smith, Ryan (5 January 2017). "The AMD Vega Architecture Teaser: Higher IPC, Tiling, & More, coming in H1'2017". Anandtech.com. Archived from the original on 10 January 2017. https://web.archive.org/web/20170110160731/http://www.anandtech.com/show/11002/the-amd-vega-gpu-architecture-teaser. Retrieved 10 January 2017. 
  18. "Vega Architecture" (in en-US). Radeon Gaming. https://gaming.radeon.com/en/vega-architecture/. 
  19. "Vega Architecture" (in en-US). Radeon Gaming. https://gaming.radeon.com/en/vega-architecture/. 
  20. "AMD Vega Microarchitecture Technical Overview" (in en). TechPowerUp. https://www.techpowerup.com/reviews/AMD/Vega_Microarchitecture_Technical_Overview/4.html. 
  21. 21.0 21.1 21.2 Smith, Ryan (30 July 2017). "Radeon RX Vega Unveiled". anand. http://www.anandtech.com/show/11680/radeon-rx-vega-unveiled-amd-announecs-499-rx-vega-64-399-rx-vega-56-launching-in-august. Retrieved 31 July 2017. 
  22. "Radeon RX Vega 56". AMD. https://gaming.radeon.com/en-us/product/vega/radeon-rx-vega-56/. Retrieved 31 July 2017. 
  23. 23.0 23.1 23.2 Angelini, Chris (30 July 2017). "AMD Radeon RX Vega 64: Bundles, Specs, And Aug. 14 Availability". Tom's Hardware. http://www.tomshardware.com/news/amd-radeon-rx-vega-64-specs-availability,35112.html. Retrieved 31 July 2017. 
  24. "14LPP 14nm FinFET Technology". GLOBALFOUNDRIES. https://www.globalfoundries.com/technology-solutions/cmos/performance/14lpp. 
  25. Schor, David (2018-07-22). "VLSI 2018: GlobalFoundries 12nm Leading-Performance, 12LP" (in en-US). https://fuse.wikichip.org/news/1497/vlsi-2018-globalfoundries-12nm-leading-performance-12lp/. 
  26. 26.0 26.1 "Radeon RX Vega 64". AMD. https://gaming.radeon.com/en-us/product/vega/radeon-rx-vega-64/. Retrieved 31 July 2017. 

External Links