AMD Instinct
This article needs to be updated.(January 2022) |
Release date | June 20, 2017 |
---|---|
Designed by | AMD |
Marketed by | AMD |
Architecture | |
Models | MI Series |
Transistors |
|
Fabrication process | |
History | |
Predecessor |
AMD Instinct is AMD's brand of professional GPUs.[1][2] It replaced AMD's FirePro S brand in 2016. Compared to the Radeon brand of mainstream consumer/gamer products, the Instinct product line is intended to accelerate deep learning, artificial neural network, and high-performance computing/GPGPU applications.
The Radeon Instinct product line directly competes with Nvidia's Ampere and Intel Xeon Phi and incoming Intel Xe lines of machine learning and GPGPU cards.
Before MI100 introduction in November 2020, the Instinct family was known as AMD Radeon Instinct, AMD dropped the Radeon brand from its name.
Supercomputers based on (AMD CPUs and) AMD Instinct GPUs now take the lead on the Green500 supercomputer list with over 50% lead over any other, and top the first 4 spots, including the second, which is the current fastest in the world on the TOP500 list, Frontier.
Products
This section needs to be updated.(January 2022) |
The three initial Radeon Instinct products were announced on December 12, 2016, and released on June 20, 2017, with each based on a different architecture.[3][4]
MI6
The MI6 is a passively cooled, Polaris 10 based card with 16 GB of GDDR5 memory and with a <150 W TDP.[1][2] At 5.7 TFLOPS (FP16 and FP32), the MI6 is expected to be used primarily for inference, rather than neural network training. The MI6 has a peak double precision (FP64) compute performance of 358 GFLOPS.[5]
MI8
The MI8 is a Fiji based card, analogous to the R9 Nano, and expected to have a <175W TDP.[1] The MI8 has 4 GB of High Bandwidth Memory. At 8.2 TFLOPS (FP16 and FP32), the MI8 is marked toward inference. The MI8 has a peak (FP64) double precision compute performance 512 GFLOPS.[6]
MI25
The MI25 is a Vega based card, utilizing HBM2 memory. The MI25 performance is expected to be 12.3 TFLOPS using FP32 numbers. In contrast to the MI6 and MI8, the MI25 is able to increase performance when using lower precision numbers, and accordingly is expected to reach 24.6 TFLOPS when using FP16 numbers. The MI25 is rated at <300W TDP with passive cooling. The MI25 also provides 768 GFLOPS peak double precision (FP64) at 1/16th rate.[7]
Accelerator | Architecture | Lithography | Compute Units | Memory | Memory Type | PCIe Support | Form Factor | FP16 Performance | BF16 Performance | FP32 Performance | FP32 Matrix Performance | FP64 Performance | FP64 Matrix Performance | INT8 Performance | INT4 Performance | TBP Peak |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
MI6 | GCN 4 | 14 nm | 36 | 16 GB | GDDR5 | 3.0 | PCIe | 5.7 TFLOPS | N/A | 5.7 TFLOPS | N/A | 358 GFLOPS | N/A | N/A | N/A | 150 W |
MI8 | GCN 3 | 28 nm | 64 | 4 GB | HBM | 8.2 TFLOPS | 8.2 TFLOPS | 512 GFLOPS | 175 W | |||||||
MI25 | GCN 5 | 14 nm | 64 | 16 GB | HBM2 | 26.4 TFLOPS | 12.3 TFLOPS | 768 GFLOPS | 300 W | |||||||
MI50 | GCN 5 | 7 nm | 60 | 4.0 | 26.5 TFLOPS | 13.3 TFLOPS | 6.6 TFLOPS | 53 TOPS | 300 W | |||||||
MI60 | GCN 5 | 64 | 32 GB | 29.5 TFLOPS | 14.7 TFLOPS | 7.4 TFLOPS | 59 TOPS | 300 W | ||||||||
MI100 | CDNA | 120 | 184.6 TFLOPS | 92.3 TFLOPS | 23.1 TFLOPS | 46.1 TFLOPS | 11.5 TFLOPS | 184.6 TOPS | 300 W | |||||||
MI210 | CDNA 2 | 6 nm | 104 | 64 GB | HBM2e | 181 TFLOPS | 22.6 TFLOPS | 45.3 TFLOPS | 22.6 TFLOPS | 45.3 TFLOPS | 181 TOPS | 300 W | ||||
MI250 | CDNA 2 | 208 | 128 GB | OAM | 362.1 TFLOPS | 45.3 TFLOPS | 90.5 TFLOPS | 45.3 TFLOPS | 90.5 TFLOPS | 362.1 TOPS | 560 W | |||||
MI250X | CDNA 2 | 220 | 383 TFLOPS | 47.92 TFLOPS | 95.7 TFLOPS | 47.9 TFLOPS | 95.7 TFLOPS | 383 TOPS | 560 W |
Software
ROCm
Following software is, as of 2022, regrouped under the Radeon Open Compute meta-project.
MxGPU
The MI6, MI8, and MI25 products all support AMD's MxGPU virtualization technology, enabling sharing of GPU resources across multiple users.[1][8]
MIOpen
MIOpen is AMD's deep learning library to enable GPU acceleration of deep learning.[1] Much of this extends the GPUOpen's Boltzmann Initiative software.[8] This is intended to compete with the deep learning portions of Nvidia's CUDA library. It supports the deep learning frameworks: Theano, Caffe, TensorFlow, MXNet, Microsoft Cognitive Toolkit, Torch, and Chainer. Programming is supported in OpenCL and Python, in addition to supporting the compilation of CUDA through AMD's Heterogeneous-compute Interface for Portability and Heterogeneous Compute Compiler.
Chipset table
Model (Code name) |
Launch | Architecture & fab |
Transistors & die size |
Core | Fillrate[a][b][c] | Processing power[a][d] (TFLOPS) |
Memory | TBP | Bus interface | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Config[e] | Clock[a] (MHz) |
Texture (GT/s) |
Pixel (GP/s) |
Half | Single | Double | Size (GB) |
Bus type & width |
Bandwidth (GB/s) |
Clock (MT/s) | ||||||
Radeon Instinct MI6 (Polaris 10)[9][10][11][12][13][14] |
Jun 20, 2017 | GCN 4 GloFo 14LP |
5.7×109 232 mm2 |
2304:144:32 36 CU |
1120 1233 |
161.3 177.6 |
35.84 39.46 |
5.161 5.682 |
5.161 5.682 |
0.323 0.355 |
16 | GDDR5 256-bit |
224 | 7000 | 150 W | PCIe 3.0 ×16 |
Radeon Instinct MI8 (Fiji)[9][10][11][15][16][17] |
GCN 3 TSMC 28 nm |
8.9×109 596 mm2 |
4096:256:64 64 CU |
1000 | 256.0 | 64.00 | 8.192 | 8.192 | 0.512 | 4 | HBM 4096-bit |
512 | 1000 | 175 W | ||
Radeon Instinct MI25 (Vega 10)[9][10][11][18][19][20][21] |
GCN 5 GloFo 14LP |
12.5×109 510 mm2 |
1400 1500 |
358.4 384.0 |
89.60 96.00 |
22.94 24.58 |
11.47 12.29 |
0.717 0.768 |
16 | HBM2 2048-bit |
484 | 1890 | 300 W | |||
Radeon Instinct MI50 (Vega 20)[22][23][24][25][26][27] |
Nov 18, 2018 | GCN 5 TSMC N7 |
13.2×109 331 mm2 |
3840:240:64 60 CU |
1450 1725 |
348.0 414.0 |
92.80 110.4 |
22.27 26.50 |
11.14 13.25 |
5.568 6.624 |
16 32 |
HBM2 4096-bit |
1024 | 2000 | PCIe 4.0 ×16 | |
Radeon Instinct MI60 (Vega 20)[23][28][29][30] |
4096:256:64 64 CU |
1500 1800 |
384.0 460.8 |
96.00 115.2 |
24.58 29.49 |
12.29 14.75 |
6.144 7.373 |
32 | ||||||||
AMD Instinct MI100 (Arcturus)[31][32][33] |
Nov 16, 2020 | CDNA TSMC N7 |
25.6×109 750 mm2 |
7680:480:- 120 CU |
1000 1502 |
480.0 721.0 |
— | 122.9 184.6 |
15.36 23.07 |
7.680 11.54 |
1228.8 | 2400 | ||||
AMD Instinct MI210 (Aldebaran)[34][35][36] |
Mar 22, 2022 | CDNA 2 TSMC N6 |
28 x 109 ~770 mm2 |
6656:416:- 104 CU (1 × GCD)[f] |
1000 1700 |
416.0 707.2 |
106.5 181.0 |
13.31 22.63 |
13.31 22.63 |
64 | HBM2E 4096-bit |
1638.4 | 3200 | |||
AMD Instinct MI250 (Aldebaran)[37][38][39] |
Nov 8, 2021 | 58 x 109 1540 mm2 |
13312:832:- 208 CU (2 × GCD) |
832.0 1414 |
213.0 362.1 |
26.62 45.26 |
26.62 45.26 |
2 × 64 | HBM2E 2 × 4096-bit[g] |
2 × 1638.4 | 500 W 560 W (Peak) | |||||
AMD Instinct MI250X (Aldebaran)[40][38][41] |
14080:880:- 220 CU (2 × GCD) |
880.0 1496 |
225.3 383.0 |
28.16 47.87 |
28.16 47.87 | |||||||||||
AMD Instinct MI300A (Antares)[42][43][44][45] |
Dec 6, 2023 | CDNA 3 TSMC N5 & N6 |
146 x 109 1017 mm2 |
14592:912:- 228 CU (6 × XCD) (24 AMD Zen 4 x86 CPU cores) |
2100 | 912.0 1550.4 |
980.6 1961.2 (With Sparsity) |
122.6 |
61.3 122.6 (FP64 Matrix) |
128 | HBM3 8192-bit |
5300 | 5200 | 550 W 760 W (Liquid Cooling) |
PCIe 5.0 ×16 | |
AMD Instinct MI300X (Aqua Vanjaram)[46][47][48][49] |
153 x 109 1017 mm2 |
19456:1216:- 304 CU (8 × XCD) |
1216.0 2062.1 |
1307.4 2614.9 (With Sparsity) |
163.4 |
81.7 163.4 (FP64 Matrix) |
192 | 750 W |
- ^ a b c Boost values (if available) are stated below the base value in italic.
- ^ Texture fillrate is calculated as the number of texture mapping units multiplied by the base (or boost) core clock speed.
- ^ Pixel fillrate is calculated as the number of render output units multiplied by the base (or boost) core clock speed.
- ^ Precision performance is calculated from the base (or boost) core clock speed based on a FMA operation.
- ^ Unified shaders : Texture mapping units : Render output units and Compute units (CU)
- ^ GCD Refers to a Graphics Compute Die. Each GCD is a different piece of silicon.
- ^ CDNA 2.0 Based cards adopt a design using two dies on the same package.They are linked with 400GB/s Bidirectional Infinity Fabric link, The dies are addressed as individual GPUs by the host system.
See also
- ROCm - AMD's open compute software stack
- AMD FirePro - AMD's predecessor to Radeon Instinct
- AMD Radeon Pro - AMD's workstation graphics and GPGPU solution
- Nvidia Quadro - Nvidia's competing workstation graphics solution
- Nvidia Tesla - Nvidia's competing GPGPU solution
- Xeon Phi - Intel's competing massively-parallel multicore processor line
- List of AMD graphics processing units
References
- ^ a b c d e Smith, Ryan (December 12, 2016). "AMD Announces Radeon Instinct: GPU Accelerators for Deep Learning, Coming in 2017". Anandtech. Retrieved December 12, 2016.
- ^ a b Shrout, Ryan (December 12, 2016). "Radeon Instinct Machine Learning GPUs include Vega, Preview Performance". PC Per. Retrieved December 12, 2016.
- ^ WhyCry (December 12, 2016). "AMD announces first VEGA accelerator:RADEON INSTINCT MI25 for deep-learning". VideoCardz. Retrieved June 6, 2022.
- ^ Mujtaba, Hassan (June 21, 2017). "AMD Radeon Instinct MI25 Accelerator With 16 GB HBM2 Specifications Detailed – Launches Today Along With Instinct MI8 and Instinct MI6". Wccftech. Retrieved June 6, 2022.
- ^ "Radeon Instinct MI6". Radeon Instinct. AMD. Retrieved June 22, 2017.[permanent dead link]
- ^ "Radeon Instinct MI8". Radeon Instinct. AMD. Retrieved June 22, 2017.[permanent dead link]
- ^ "Radeon Instinct MI25". Radeon Instinct. AMD. Retrieved June 22, 2017.[permanent dead link]
- ^ a b Kampman, Jeff (December 12, 2016). "AMD opens up machine learning with Radeon Instinct". TechReport. Retrieved December 12, 2016.
- ^ a b c Smith, Ryan (December 12, 2016). "AMD Announces Radeon Instinct: GPU Accelerators for Deep Learning, Coming in 2017". AnandTech. Retrieved December 12, 2016.
- ^ a b c Shrout, Ryan (December 12, 2016). "Radeon Instinct Machine Learning GPUs include Vega, Preview Performance". PCPerspective. Retrieved December 12, 2016.
- ^ a b c Kampman, Jeff (December 12, 2016). "AMD opens up machine learning with Radeon Instinct". Tech Report. Retrieved December 12, 2016.
- ^ "Radeon Instinct MI6". AMD. Archived from the original on August 1, 2017. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI6 Datasheet" (PDF). usermanual.wiki. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI6 Specs". TechPowerUp. Retrieved May 27, 2022.
- ^ "Radeon Instinct MI8". AMD. Archived from the original on August 1, 2017. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI8 Datasheet" (PDF). usermanual.wiki. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI8 Specs". TechPowerUp. Retrieved May 27, 2022.
- ^ Smith, Ryan (January 5, 2017). "The AMD Vega Architecture Teaser: Higher IPC, Tiling, & More, coming in H1'2017". AnandTech. Retrieved January 10, 2017.
- ^ "Radeon Instinct MI25". AMD. Archived from the original on August 1, 2017. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI25 Datasheet" (PDF). AMD. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI25 Specs". TechPowerUp. Retrieved May 27, 2022.
- ^ Walton, Jarred (January 10, 2019). "Hands on with the AMD Radeon VII". PC Gamer.
- ^ a b "Next Horizon – David Wang Presentation" (PDF). AMD.
- ^ "AMD Radeon Instinct MI50 Accelerator (16GB)". AMD. Retrieved December 24, 2022.
- ^ "AMD Radeon Instinct MI50 Accelerator (32GB)". AMD. Retrieved December 24, 2022.
- ^ "AMD Radeon Instinct MI50 Datasheet" (PDF). AMD. Retrieved December 24, 2022.
- ^ "AMD Radeon Instinct MI50 Specs". TechPowerUp. Retrieved May 27, 2022.
- ^ "Radeon Instinct MI60". AMD. Archived from the original on November 22, 2018. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI60 Datasheet" (PDF). AMD. Retrieved December 24, 2022.
- ^ "AMD Radeon Instinct MI60 Specs". TechPowerUp. Retrieved May 27, 2022.
- ^ "AMD Instinct MI100 Accelerator". AMD. Retrieved May 27, 2022.
- ^ "AMD Instinct MI100 Accelerator Brochure" (PDF). AMD. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI100 Specs". TechPowerUp. Retrieved May 26, 2022.
- ^ "AMD Instinct MI210 Accelerator". AMD. Retrieved May 27, 2022.
- ^ "AMD Instinct MI210 Accelerator Brochure" (PDF). AMD. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI210 Specs". TechPowerUp. Retrieved May 27, 2022.
- ^ "AMD Instinct MI250 Accelerator". AMD. Retrieved May 27, 2022.
- ^ a b "AMD Instinct MI200 Series Accelerator Datasheet" (PDF). AMD. Retrieved December 24, 2022.
- ^ "AMD Radeon Instinct MI250 Specs". TechPowerUp. Retrieved May 26, 2022.
- ^ "AMD Instinct MI250X Accelerator". AMD. Retrieved May 27, 2022.
- ^ "AMD Radeon Instinct MI250X Specs". TechPowerUp. Retrieved May 26, 2022.
- ^ "AMD Instinct MI300A APU". AMD. Retrieved December 12, 2023.
- ^ "AMD Instinct MI300A Series Accelerator Datasheet" (PDF). AMD. Retrieved December 12, 2023.
- ^ "AMD Radeon Instinct MI300 Specs". TechPowerUp. Retrieved December 12, 2023.
- ^ "AMD-CDNA3-white-paper" (PDF). AMD. Retrieved December 12, 2023.
- ^ "AMD Instinct MI300X GPU". AMD. Retrieved December 12, 2023.
- ^ "AMD Instinct MI300X Series Accelerator Datasheet" (PDF). AMD. Retrieved December 12, 2023.
- ^ "AMD Radeon Instinct MI300 Specs". TechPowerUp. Retrieved December 12, 2023.
- ^ "AMD-CDNA3-white-paper" (PDF). AMD. Retrieved December 12, 2023.