Radeon Instinct

AMD Radeon Instinct is AMD's brand of deep learning oriented GPUs.[1][2] It replaced AMD's FirePro S brand in 2016. Compared to the Radeon brand of mainstream consumer/gamer products, the Radeon Instinct branded products are intended to accelerate deep learning, artificial neural network, and high-performance computing/GPGPU applications.

AMD Radeon Instinct

Design firm	Advanced Micro Devices
Introduced	2016
Type	Servers

The Radeon Instinct product line directly competes with Nvidia's Tesla and Intel Xeon Phi lines of deep learning and GPGPU cards.

Since MI100 introduction in November 2020, the Radeon Instinct family is known as AMD Instinct, dropping Radeon brand from its name.

Products

The three initial Radeon Instinct products were announced in December 2016, with each based on a different architecture.

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.[3]

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.[4]

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.[5]

Software

MxGPU

The MI6, MI8, and MI25 products all support AMD's MxGPU virtualization technology, enabling sharing of GPU resources across multiple users.[1][6]

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.[6] 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, The 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 (codename)	Release date	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 Instinct MI6 (Polaris 10) [1][7][8][9][10]	December 2016	GCN 4^th gen 14 nm	5.7×10^9 232 mm^2	2304:144:32 36 CU	1120 1233	177.6	39.46	5800	5800	358	GDDR5 256-bit	16	7000	224	150 W	PCIe 3.0 x16
Radeon Instinct MI8 (Fiji XT) [1][7][8][11][12]		GCN 3^rd gen 28 nm	8.9×10^9 596 mm^2	4096:256:64 64 CU	1000	256.0	64.0	8200	8200	512	HBM 4096-bit	4	1000	512	175 W
Radeon Instinct MI25 (Vega 10 XT)[1][7][8][13][14][15]		GCN 5^th gen 14 nm	12.5×10^9 510 mm^2	4096:256:64 64 CU	1400 1500	384	96.0	24600	12300	768	HBM2 2048-bit	16	1704	436.2	300 W
Radeon Instinct MI25 mxgpu (Prototype, June 2017)[16]	Unreleased		2× 12.5×10^9 510 mm^2	2× 4096:256:64 64 CU	1400 1500	2× 384	2× 96.0	2× 24600	2× 12300	2× 768		2× 16		2× 436.2
Radeon Instinct MI50 (Vega 20 GL)[17][18][19][20]	November 2018	GCN 5^th gen 7 nm	13.2×10^9 331 mm^2	3840:240:- 60 CU	1450 1725	348 414	-	26500	13300	6600	HBM2 4096-bit	16 or 32[21]	2000	1024		PCIe 4.0 x16
Radeon Instinct MI60 (Vega 20 GL)[17][22][23]				4096:256:- 64 CU	1500 1800	384 460.8	-	29450	14725	7362.5		32		1024
AMD Instinct MI100 (MI100 XL)[24]	November 12, 2020	CDNA 1.0 7 nm	? 750 mm^2	7680:480:- 120 CU	1000 1502	480 720	-	184600	23100	11500			2400	1228.8

1. 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)

See also

• 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

1. Smith, Ryan (12 December 2016). "AMD Announces Radeon Instinct: GPU Accelerators for Deep Learning, Coming in 2017". Anandtech. Retrieved 12 December 2016.
2. Shrout, Ryan (12 December 2016). "Radeon Instinct Machine Learning GPUs include Vega, Preview Performance". PC Per. Retrieved 12 December 2016.
3. "Radeon Instinct MI6". Radeon Instinct. AMD. Retrieved 22 June 2017.
4. "Radeon Instinct MI8". Radeon Instinct. AMD. Retrieved 22 June 2017.
5. "Radeon Instinct MI25". Radeon Instinct. AMD. Retrieved 22 June 2017.
6. Kampman, Jeff (12 December 2016). "AMD opens up machine learning with Radeon Instinct". TechReport. Retrieved 12 December 2016.
7. Shrout, Ryan (12 December 2016). "Radeon Instinct Machine Learning GPUs include Vega, Preview Performance". PC Per. Retrieved 12 December 2016.
8. Kampman, Jeff (12 December 2016). "AMD opens up machine learning with Radeon Instinct". TechReport. Retrieved 12 December 2016.
9. "Radeon Instinct MI6". Radeon Instinct. AMD. Retrieved 22 June 2017.
10. "AMD Radeon Instinct MI6 Specs". TechPowerUp.
11. "Radeon Instinct MI8". Radeon Instinct. AMD. Retrieved 22 June 2017.
12. "AMD Radeon Instinct MI8 Specs". TechPowerUp.
13. Smith, Ryan (5 January 2017). "The AMD Vega Architecture Teaser: Higher IPC, Tiling, & More, coming in H1'2017". Anandtech.com. Retrieved 10 January 2017.
14. "Radeon Instinct MI25". Radeon Instinct. AMD. Retrieved 22 June 2017.
15. "AMD Radeon Instinct MI25 Specs". TechPowerUp.
16. "AMD Radeon Instinct MI25 MxGPU Specs". TechPowerUp.
17. "Next Horizon – David Wang Presentation" (PDF). AMD. AMD.
18. "Radeon Instinct MI50". AMD. AMD.
19. "Radeon Instinct MI50 Datasheet" (PDF). AMD. AMD.
20. "Hands on with the AMD Radeon VII". PC Gamer. Jarred Walton.
21. "AMD Radeon Instinct™ MI50 Accelerator (32GB)".
22. "Radeon Instinct MI60". AMD. AMD.
23. "Radeon Instinct MI60 Datasheet" (PDF). AMD. AMD.
24. "Radeon Instinct MI100". AMD. AMD.