AMD Unveils Instinct MI200 'Aldebaran' GPU, First 6nm MCM Product With 58 Billion Transistors, Over 14,000 Cores & 128 GB HBM2e Memory

AMD has officially announced its next-generation MI200 HPC GPU codenamed Aldebaran that uses a 6nm CDNA two compages to deliver insane compute performance.

AMD Unveils Instinct MI200, Powering The Next-Gen Compute Powerhouse With First 6nm MCM GPU Technology & Over 95 TFLOPs FP32 Performance

AMD is officially the first to MCM technology and they are doing so with a g product which is their Instinct MI200 codenamed Aldebaran. The AMD Aldebaran GPU will come in diverse forms & sizes simply it'due south all based on the brand new CDNA 2 architecture which is the most refined variation of Vega. Some of the main features before we go into detail are listed below:

AMD CDNA 2 architecture – 2d Gen Matrix Cores accelerating FP64 and FP32 matrix operations, delivering up to 4X the peak theoretical FP64 operation vs. AMD previous-gen GPUs.
Leadership Packaging Applied science – Industry-kickoff multi-dice GPU design with 2.5D Elevated Fanout Span (EFB) technology delivers 1.8X more cores and 2.7X college retention bandwidth vs. AMD previous-gen GPUs, offer the industry's all-time amass peak theoretical memory bandwidth at 3.2 terabytes per 2nd.
3rd Gen AMD Infinity Fabric technology – Up to 8 Infinity Fabric links connect the AMD Instinct MI200 with thirdGen EPYC CPUs and other GPUs in the node to enable unified CPU/GPU memory coherency and maximize system throughput, allowing for an easier on-ramp for CPU codes to tap the power of accelerators.

AMD Instinct MI200 GPU Dice Shot:

Inside the AMD Instinct MI200 is an Aldebaran GPU featuring ii dies, a secondary and a primary. Information technology has ii dies with each consisting of 8 shader engines for a total of 16 SE's. Each Shader Engine packs 16 CUs with full-rate FP64, packed FP32 & a 2d Generation Matrix Engine for FP16 & BF16 operations.

Each die, as such, is composed of 128 compute units or 8192 stream processors. This rounds up to a total of 220 compute units or 14,080 stream processors for the entire chip. The Aldebaran GPU is also powered past a new XGMI interconnect. Each chiplet features a VCN two.half dozen engine and the main IO controller.

Built on AMD CDNA two architecture, AMD Instinct MI200 series accelerators evangelize leading awarding performance for a broad set of HPC workloads. The AMD Instinct MI250X accelerator provides up to 4.9X ameliorate performance than competitive accelerators for double precision (FP64) HPC applications and surpasses 380 teraflops of peak theoretical half-precision (FP16) for AI workloads to enable disruptive approaches in further accelerating information-driven research.

In terms of performance, AMD is touting diverse record wins in the HPC segment over NVIDIA's A100 solution with up to 3x performance improvements in AMG.

As for DRAM, AMD has gone with an 8-channel interface consisting of 1024-flake interfaces for an 8192-bit wide bus interface. Each interface can support 2GB HBM2e DRAM modules. This should requite usa up to xvi GB of HBM2e memory capacity per stack and since there are eight stacks in total, the total corporeality of chapters would be a whopping 128 GB. That's 48 GB more than than the A100 which houses 80 GB HBM2e retentiveness. The memory will clock in at an insane speed of 3.2 Gbps for a full-on bandwidth of 3.2 TB/south. This is a whole ane.two TB/s more bandwidth than the A100 lxxx GB which has two TB/s.

The AMD Instinct MI200 will exist powering three top-tier supercomputers which include the United States' exascale Frontier system; the European Matrimony's pre-exascale LUMI system; and Australia's petascale Setonix system. The competition includes the A100 eighty GB which offers nineteen.5 TFLOPs of FP64, 156 TFLOPs of FP32 and 312 TFLOPs of FP16 compute ability. But nosotros are probable to hear about NVIDIA'due south own Hopper MCM GPU side by side year so there's going to exist a heated competition betwixt the two GPU juggernauts in 2022.

AMD Radeon Instinct Accelerators 2020

Accelerator Name	AMD Instinct MI300	AMD Instinct MI250X	AMD Instinct MI250	AMD Instinct MI210	AMD Instinct MI100	AMD Radeon Instinct MI60	AMD Radeon Instinct MI50	AMD Radeon Instinct MI25	AMD Radeon Instinct MI8	AMD Radeon Instinct MI6
GPU Architecture	TBA (CDNA iii)	Aldebaran (CDNA 2)	Aldebaran (CDNA 2)	Aldebaran (CDNA 2)	Arcturus (CDNA 1)	Vega 20	Vega 20	Vega 10	Fiji XT	Polaris 10
GPU Process Node	Advanced Process Node	6nm	6nm	6nm	7nm FinFET	7nm FinFET	7nm FinFET	14nm FinFET	28nm	14nm FinFET
GPU Dies	4 (MCM)?	2 (MCM)	two (MCM)	i (MCM)	one (Monolithic)	1 (Monolithic)	ane (Monolithic)	ane (Monolithic)	ane (Monolithic)	1 (Monolithic)
GPU Cores	28,160?	14,080	13,312	6656	7680	4096	3840	4096	4096	2304
GPU Clock Speed	TBA	1700 MHz	1700 MHz	~1700 MHz?	~1500 MHz	1800 MHz	1725 MHz	1500 MHz	1000 MHz	1237 MHz
FP16 Compute	TBA	383 TOPs	362 TOPs	~176 TOPs	185 TFLOPs	29.5 TFLOPs	26.5 TFLOPs	24.vi TFLOPs	8.2 TFLOPs	5.7 TFLOPs
FP32 Compute	TBA	95.vii TFLOPs	xc.5 TFLOPs	~44 TFLOPs	23.1 TFLOPs	fourteen.7 TFLOPs	xiii.3 TFLOPs	12.3 TFLOPs	8.2 TFLOPs	v.7 TFLOPs
FP64 Compute	TBA	47.9 TFLOPs	45.iii TFLOPs	~22 TFLOPs	11.5 TFLOPs	vii.iv TFLOPs	6.6 TFLOPs	768 GFLOPs	512 GFLOPs	384 GFLOPs
VRAM	TBA	128 GB HBM2e	128 GB HBM2e	64 GB HBM2e	32 GB HBM2	32 GB HBM2	16 GB HBM2	16 GB HBM2	4 GB HBM1	sixteen GB GDDR5
Memory Clock	TBA	3.two Gbps	3.two Gbps	iii.two Gbps?	1200 MHz	1000 MHz	g MHz	945 MHz	500 MHz	1750 MHz
Memory Double-decker	TBA	8192-bit	8192-bit	4096-flake	4096-bit passenger vehicle	4096-scrap bus	4096-flake passenger vehicle	2048-bit jitney	4096-bit coach	256-bit bus
Retentiveness Bandwidth	TBA	iii.2 TB/southward	3.ii TB/s	1.vi TB/s	i.23 TB/due south	1 TB/s	one TB/due south	484 GB/s	512 GB/south	224 GB/s
Form Factor	TBA	OAM	OAM	Dual Slot Card	Dual Slot, Total Length	Dual Slot, Full Length	Dual Slot, Total Length	Dual Slot, Full Length	Dual Slot, Half Length	Single Slot, Full Length
Cooling	TBA	Passive Cooling	Passive Cooling	Passive Cooling	Passive Cooling	Passive Cooling	Passive Cooling	Passive Cooling	Passive Cooling	Passive Cooling
TDP	TBA	560W	500W?	300W?	300W	300W	300W	300W	175W	150W

The Aldebaran MI200 GPU will come up in three configurations, the OAM only MI250 and MI250X & the dual-slot PCIe MI210. AMD has but shared total specifications and functioning figures for its MI250 class HPC GPUs. The MI250X features the full 14,080 configurations and delivers 47.9, 95.7, 383 TFLOPs of FP64/FP32/FP16 while the MI250 features 13,312 cores with 45.3,90.5,362.1 TFLOPs of FP64/FP32/FP16 performance. The retention configuration remains the same between the two GPU configurations.

AMD Instinct MI200 GPU Package:

Source: https://wccftech.com/amd-unveils-instinct-mi200-aldebaran-gpu-first-6nm-mcm-product-with-58-billion-transistors-over-14000-cores-128-gb-hbm2e-memory/

Posted by: healeyhispossiond.blogspot.com