(→Process Technology) |
(→Bibliography) |
||
(3 intermediate revisions by the same user not shown) | |||
Line 19: | Line 19: | ||
==Architecture == | ==Architecture == | ||
=== Block Diagram === | === Block Diagram === | ||
− | + | :[[File:habana goya block diagram.svg|400px]] | |
+ | |||
== Overview == | == Overview == | ||
− | {{ | + | Goya is designed as a microarchitecture for the [[acceleration]] of inference. Since the target market is the data center, the [[thermal design point]] for those chips was relatively high - at around 200 W. Goya relies on [[PCIe]] 4.0 to interface to a host processor. Habana's software compiles the models and associated instructions into independent recipes which can then be sent to the accelerator for execution. The design itself uses a heterogenous approach comprising of a large General Matrix Multiply (GMM) engine, Tensor Processor Cores (TPCs), and a large shared memory pool. |
− | == | + | |
− | {{ | + | === Tensor Processor Cores (TPC) === |
+ | [[File:habana hl-100.jpg|right|thumb|{{habana|HL|HL-100/102}} PCIe Card]] | ||
+ | There are eight TPCs. Each TPC also incorporates its own local memory but omits caches. The on-die caches and memory can be either hardware-managed or fully software-managed, allowing the compiler to optimize the residency of data and reducing [[data movement|movement]]. Each of the individual TPCs is a [[VLIW]] DSP design that has been optimized for AI applications. This includes [[AI]]-specific [[instructions]] and operations. The TPCs are designed for flexibility and can be programmed in plain [[C]]. The TPC supports mixed-prevision operations including 8-bit, 16-bit, and 32-bit SIMD vector operations for both [[integer]] and [[floating-point]]. This was done in order to allow accuracy loss tolerance to be controlled on a per-model design by the programmer. Goya offers both coarse-grained precision control and fine-grained down to the tensor level. | ||
+ | |||
+ | == Bibliography == | ||
+ | * {{bib|hc|31|Habana}} | ||
+ | * Habana, AI Hardware Summit 2019 | ||
+ | * Habana, Linley Fall Processor Conference 2019 | ||
+ | |||
== See also == | == See also == | ||
* {{\\|Gaudi}} | * {{\\|Gaudi}} | ||
* {{habana|HL}} series | * {{habana|HL}} series |
Latest revision as of 15:29, 28 December 2019
Edit Values | |
Goya µarch | |
General Info | |
Arch Type | NPU |
Designer | Habana |
Manufacturer | TSMC |
Introduction | 2018 |
Process | 16 nm |
PE Configs | 8 |
Contemporary | |
Gaudi |
Goya is a 16-nanometer microarchitecture for inference neural processors designed by Habana Labs.
Contents
Process Technology[edit]
Goya-based processors are fabricated on TSMC 16-nanometer process.
Architecture[edit]
Block Diagram[edit]
Overview[edit]
Goya is designed as a microarchitecture for the acceleration of inference. Since the target market is the data center, the thermal design point for those chips was relatively high - at around 200 W. Goya relies on PCIe 4.0 to interface to a host processor. Habana's software compiles the models and associated instructions into independent recipes which can then be sent to the accelerator for execution. The design itself uses a heterogenous approach comprising of a large General Matrix Multiply (GMM) engine, Tensor Processor Cores (TPCs), and a large shared memory pool.
Tensor Processor Cores (TPC)[edit]
There are eight TPCs. Each TPC also incorporates its own local memory but omits caches. The on-die caches and memory can be either hardware-managed or fully software-managed, allowing the compiler to optimize the residency of data and reducing movement. Each of the individual TPCs is a VLIW DSP design that has been optimized for AI applications. This includes AI-specific instructions and operations. The TPCs are designed for flexibility and can be programmed in plain C. The TPC supports mixed-prevision operations including 8-bit, 16-bit, and 32-bit SIMD vector operations for both integer and floating-point. This was done in order to allow accuracy loss tolerance to be controlled on a per-model design by the programmer. Goya offers both coarse-grained precision control and fine-grained down to the tensor level.
Bibliography[edit]
- Habana, IEEE Hot Chips 31 Symposium (HCS) 2019.
- Habana, AI Hardware Summit 2019
- Habana, Linley Fall Processor Conference 2019
See also[edit]
codename | Goya + |
designer | Habana + |
first launched | 2018 + |
full page name | habana/microarchitectures/goya + |
instance of | microarchitecture + |
manufacturer | TSMC + |
name | Goya + |
process | 16 nm (0.016 μm, 1.6e-5 mm) + |
processing element count | 8 + |