This article provides insufficient context for those unfamiliar with the subject.(October 2020)
Graphics Core Next (GCN) is the codename for both a series of microarchitectures as well as for an instruction set architecture that was developed by AMD for their GPUs as the successor to their TeraScale microarchitecture/instruction set. The first product featuring GCN was launched on January 9, 2012.
GCN is a RISC SIMD (or rather SIMT) microarchitecture contrasting the VLIW SIMD architecture of TeraScale. GCN requires considerably more transistors than TeraScale, but offers advantages for GPGPU computation. It makes the compiler simpler and should also lead to better use.
GCN graphics chips are fabricated with CMOS at 28 nm, and with FinFET at 14 nm (by Samsung Electronics and GlobalFoundries) and 7 nm (by TSMC), available on selected models in the Radeon HD 7000, HD 8000, 200, 300, 400, 500 and Vega series of AMD Radeon graphics cards, including the separately released Radeon VII. GCN is also used in the graphics portion of AMD Accelerated Processing Units (APU), such as in the PlayStation 4 and Xbox One APUs.
The GCN instruction set is owned by AMD (that also owns the X86-64 instruction set). The GCN instruction set has been developed specifically for GPUs (and GPGPU) and, for example, has no micro-operation for division.
Documentation is available for:
The GNU Compiler Collection (GCC) supports GCN 3 (Fiji, Carrizo) and GCN 5 (Vega) since 2019 (GCC 9) for single-threaded, stand-alone programs and with GCC 10 also offloading via OpenMP and OpenACC.
MIAOW is an open-source RTL implementation of the AMD Southern Islands GPGPU instruction set (aka Graphics Core Next).
At the "Super Computing 15" AMD showed their Heterogeneous Compute Compiler (HCC), a headless Linux driver and HSA runtime infrastructure for cluster-class, High Performance Computing (HPC) and the Heterogeneous-compute Interface for Portability (HIP) tool for porting CUDA-based applications to a common C++ programming model.
As of July 2017, the family of microarchitectures implementing the identically called instruction set "Graphics Core Next" has seen five iterations. The differences in the instruction set are rather minimal and do not differentiate too much from one another. An exception is the fifth generation GCN architecture, which heavily modified the stream processors to improve performance and support the simultaneous processing of two lower precision numbers in place of a single higher precision number.
The "Graphics Command Processor" (GCP) is a functional unit of the GCN microarchitecture. Among other tasks, it is responsible for Asynchronous Shaders. The short video AMD Asynchronous Shaders visualizes the differences between "multi thread", "preemption" and "Asynchronous Shaders".
The Asynchronous Compute Engine (ACE) is a distinct functional block serving computing purposes. Its purpose is similar to that of the Graphics Command Processor.[ambiguous]
Since the third iteration of GCN, the hardware contains two schedulers: One to schedule wavefronts during shader execution (CU Scheduler, see below) and a new one to schedule execution of draw and compute queues. The latter helps performance by executing compute operations when the CUs are underutilized because of graphics commands limited by fixed function pipeline speed or bandwidth limited. This functionality is known as Async Compute.
For a given shader, the GPU drivers also need to select a good instruction order, in order to minimize latency. This is done on cpu, and is sometimes referred as "Scheduling".
The geometry processor contains the Geometry Assembler, the Tesselator and the Vertex Assembler.
One compute unit combines 64 shader processors with 4 TMUs. The compute unit is separate from, but feed into, the render output units (ROPs). Each Compute Unit consists of a CU Scheduler, a Branch & Message Unit, 4 SIMD Vector Units (each 16-lane wide), 4 64KiB VGPR files, 1 scalar unit, a 4 KiB GPR file, a local data share of 64 KiB, 4 Texture Filter Units, 16 Texture Fetch Load/Store Units and a 16 KiB L1 Cache. Four Compute units are wired to share a 16KiB L1 Instruction Cache and a 32KiB L1 data cache, both of which are read-only. A SIMD-VU operates on 16 elements at a time (per cycle), while a SU can operate on one a time (one/cycle). In addition the SU handles some other operations like branching.
Every SIMD-VU has some private memory where it stores its registers. There are two types of registers: scalar registers (s0, s1, etc.), which hold 4 bytes number each, and vector registers (v0, v1, etc.), which represent a set of 64 4 bytes numbers each. When you operate on the vector registers, every operation is done in parallel on the 64 numbers. Every time you do some work with them, you actually work with 64 inputs. For example, you work on 64 different pixels at a time (for each of them the inputs are slightly different, and thus you get slightly different color at the end).
Every SIMD-VU has room for 512 scalar registers and 256 vector registers.
The CU scheduler is the hardware functional block choosing for the SIMD-VU which wavefronts to execute. It picks one SIMD-VU per cycle for scheduling. This is not to be confused with other schedulers, in hardware or software.
In all GCN-GPUs, a "wavefront" consists of 64 threads, and in all Nvidia GPUs a "warp" consists of 32 threads.
AMD's solution is to attribute multiple wavefronts to each SIMD-VU. The hardware distributes the registers to the different wavefronts, and when one wavefront is waiting on some result, which lies in memory, the CU Scheduler decides to make the SIMD-VU work on another wavefront. Wavefronts are attributed per SIMD-VU. SIMD-VUs do not exchange wavefronts. At max 10 wavefronts can be attributed per SIMD-VU (thus 40 per CU).
AMD CodeXL shows tables with the relationship between number of SGPRs and VGPRs to the number of wavefronts, but basically for SGPRS it is min(104, 512/numwavefronts) and VGPRS 256/numwavefronts.
Note that in conjunction with the SSE instructions this concept of most basic level of parallelism is often called a "vector width". The vector width is characterized by the total number of bits in it.
Each SIMD Vector Unit has:
Each SIMD-VU has 10 wavefront instruction buffer, and it takes 4 cycles to execute one wavefront.
Some of the specific HSA features implemented in the hardware need support from the operating system's kernel (its subsystems) and/or from specific device drivers. For example, in July 2014 AMD published a set of 83 patches to be merged into Linux kernel mainline 3.17 for supporting their Graphics Core Next-based Radeon graphics cards. The special driver titled "HSA kernel driver" resides in the directory /drivers/gpu/hsa while the DRM-graphics device drivers reside in /drivers/gpu/drm and augments the already existent DRM driver for Radeon cards. This very first implementation focuses on a single "Kaveri" APU and works alongside the existing Radeon kernel graphics driver (kgd).
This section needs expansion. You can help by adding to it. (August 2018)
They are used to perform scheduling and offload the assignment of compute queues to the ACEs from the driver to hardware by buffering these queues until there is at least one empty queue in at least one ACE, causing the HWS to immediately assign buffered queues to the ACEs until all queues are full or there are no more queues to safely assign. Part of the scheduling work performed includes prioritized queues which allow critical tasks to run at a higher priority than other tasks without requiring the lower priority tasks to be preempted to run the high priority task, therefore allowing the tasks to run concurrently with the high priority tasks scheduled to hog the GPU as much as possible while letting other tasks use the resources that the high priority tasks are not using. These are essentially Asynchronous Compute Engines that lack dispatch controllers. They were first introduced in the fourth generation GCN microarchitecture, but were present in the third generation GCN microarchitecture for internal testing purposes. A driver update has enabled the hardware schedulers in third generation GCN parts for production use.
This unit discards degenerate triangles before they enter the vertex shader and triangles that do not cover any fragments before they enter the fragment shader. This unit was introduced with the fourth generation GCN microarchitecture.
|Release date||January 2012][|
|Successor||Graphics Core Next 2|
The GCN 1 microarchitecture was used in several Radeon HD 7000 series graphics cards.
discrete GPUs (Southern Islands family):
|Release date||September 2013][|
|Predecessor||Graphics Core Next 1|
|Successor||Graphics Core Next 3|
GCN 2nd generation was introduced with Radeon HD 7790 and is also found in Radeon HD 8770, R7 260/260X, R9 290/290X, R9 295X2, R7 360, R9 390/390X, as well as Steamroller-based Desktop Kaveri APUs and Mobile Kaveri APUs and in the Puma-based "Beema" and "Mullins" APUs. It has multiple advantages over the original GCN, including FreeSync support, AMD TrueAudio and a revised version of AMD PowerTune technology.
GCN 2nd generation introduced an entity called "Shader Engine" (SE). A Shader Engine comprises one geometry processor, up to 44 CUs (Hawaii chip), rasterizers, ROPs, and L1 cache. Not part of a Shader Engine is the Graphics Command Processor, the 8 ACEs, the L2 cache and memory controllers as well as the audio and video accelerators, the display controllers, the 2 DMA controllers and the PCIe interface.
discrete GPUs (Sea Islands family):
integrated into APUs:
|Release date||June 2015][|
|Predecessor||Graphics Core Next 2|
|Successor||Graphics Core Next 4|
GCN 3rd generation was introduced in 2014 with the Radeon R9 285 and R9 M295X, which have the "Tonga" GPU. It features improved tessellation performance, lossless delta color compression to reduce memory bandwidth usage, an updated and more efficient instruction set, a new high quality scaler for video, and a new multimedia engine (video encoder/decoder). Delta color compression is supported in Mesa. However, its double precision performance is worse compared to previous generation.
integrated into APUs:
|Release date||June 2016][|
|Predecessor||Graphics Core Next 3|
|Successor||Graphics Core Next 5|
GPUs of the Arctic Islands-family were introduced in Q2 of 2016 with the AMD Radeon 400 series. The 3D-engine (i.e. GCA (Graphics and Compute array) or GFX) is identical to that found in the Tonga-chips. But Polaris feature a newer Display Controller engine, UVD version 6.3, etc.
All Polaris-based chips other than the Polaris 30 are produced on the 14 nm FinFET process, developed by Samsung Electronics and licensed to GlobalFoundries. The slightly newer refreshed Polaris 30 is built on the 12 nm LP FinFET process node, developed by Samsung and GlobalFoundries. The fourth generation GCN instruction set architecture is compatible with the third generation. It is an optimization for 14 nm FinFET process enabling higher GPU clock speeds than with the 3rd GCN generation. Architectural improvements include new hardware schedulers, a new primitive discard accelerator, a new display controller, and an updated UVD that can decode HEVC at 4K resolutions at 60 frames per second with 10 bits per color channel.
In addition to dedicated GPUs, Polaris is utilized in the APUs of the PlayStation 4 Pro and Xbox One X, titled "Neo" and "Scorpio", respectively.
FP64 performance of all GCN 4th generation GPUs is 1/16 of FP32 performance.
|Release date||June 2017][|
|Predecessor||Graphics Core Next 4|
AMD began releasing details of their next generation of GCN Architecture, termed the 'Next-Generation Compute Unit', in January 2017. The new design was expected to increase instructions per clock, higher clock speeds, support for HBM2, a larger memory address space. The discrete graphics chipsets also include "HBCC (High Bandwidth Cache Controller)", but not when integrated into APUs. Additionally, the new chips were expected to include improvements in the Rasterisation and Render output units. The stream processors are heavily modified from the previous generations to support packed math Rapid Pack Math technology for 8-bit, 16-bit, and 32-bit numbers. With this there is a significant performance advantage when lower precision is acceptable (for example: processing two half-precision numbers at the same rate as a single single precision number).
Nvidia introduced tile-based rasterization and binning with Maxwell, and this was a big reason for Maxwell's efficiency increase. In January, AnandTech assumed that Vega would finally catch up with Nvidia regarding energy efficiency optimizations due to the new "DSBR (Draw Stream Binning Rasterizer)" to be introduced with Vega.
It also added support for a new shader stage - Primitive Shaders. Primitive shaders provide more flexible geometry processing and replace the vertex and geometry shaders in a rendering pipeline. As of December 2018, the Primitive shaders can't be used because required API changes are yet to be done.
integrated into APUs:
Double-precision floating-point (FP64) performance of all GCN 5th generation GPUs, except for Vega 20, is 1/16 of FP32 performance. For Vega 20 this is 1/2 of FP32 performance. All GCN 5th generation GPUs support half-precision floating-point (FP16) calculations which is double of FP32 performance.
Asynchronous Compute Engine (ACE)
AMD's new Asynchronous Compute Engines serve as the command processors for compute operations on GCN. The principal purpose of ACEs will be to accept work and to dispatch it off to the CUs for processing.