- Add news of Hailo’s Hailo-8 chip.
- Add MLIR in AI Compiler section and news of Tesla's FSD chip..
- Add startup Eta Compute.
- Add Compiler ONNC.
- Add Chinese startup EEasy Tech.
- Add startup Optaylsys.
- Add a new section of AI Chip Compilers.
- Add Chinese startup Enflame.
- Add news of Google Edge TPU.
- Add a new section of AI Chip Benchmarks.
- Add news of Horizon Robotics.
- Add startup Anaflash.
- Add Chinese startup Canaan.
- Add startup Cornami.
- Add startup Preferred Network.
- Add news of Mediatek's Helio P90.
- Add "White Paper on AI Chip Technologies 2018" in Reference.
- Add news of MLPerf in Reference.
- Add startup Flex Logix.
- Add news of Qualcomm's Snapdragon 855 Mobile Platform.
- Add startup aiCTX.
- Add news of AWS's inference chip: Inferentia.
- Add news of Samsung's NPU.
- Add Marvell’s AI SSD controller proof-of-concept architecture solution.
- Add news of Tesla's AI Chip.
- Add news of Huawei's AI Chip.
- Add two more AI benchmarks, AI Benchmark and AI Matrix in Reference.
- Add news of Xilinx's AI inference card.
- Add news of Alibaba's AI Chip.
- Add startup Habana.
- Add Apple's A12, Huawei's Kirin980 and Nvidia's T4.
- Add Chinese startup NextVPU.
<style> table, th, td { border: 1px solid black; } </style>
<style> table, th, td { border: 1px solid black; } </style>
| Both | Datacenter | Edge |
|---|---|---|
| Intel, Nvidia, IBM, Xilinx, HiSilicon, Google, Baidu, Alibaba Group, Cambricon, Bitmain, Wave Computing,Tachyum,AlphaICs, Marvell | AMD, Microsoft, Apple, Tencent Cloud,Aliyun, Baidu Cloud, HUAWEI Cloud, Fujitsu, Nokia, Facebook, HPE, Thinkforce, Cerebras, Graphcore, Groq, SambaNova Systems, Adapteva, PEZY, Habana, Enflame | Qualcomm, Samsung, STMicroelectronics, NXP, MediaTek, Tesla, Rockchip, Amazon_AWS, ARM, Synopsys, Imagination, CEVA, Cadence, VeriSilicon, Videantis, Horizon Robotics, Chipintelli, Unisound, AISpeech, Rokid, Tenstorrent, ThinCI, Koniku, Knowm, Mythic, Kalray, BrainChip, AImotive, DeepScale, Leepmind, Krtkl, NovuMind, REM, TERADEEP, DEEP VISION, KAIST DNPU, Kneron, Esperanto Technologies, Gyrfalcon Technology, GreenWaves Technology, Lightelligence, Lightmatter, ThinkSilicon, Innogrit, Kortiq, Hailo,Syntiant, NextVPU, aiCTX, Cornami, Anaflash, Eesay Tech, Optaylsys, Eta Compute |
Intel® Nervana™ Neural Network Processors (NNP) Redefine AI Silicon > As our Intel CEO Brian Krzanich discussed earlier today at Wall Street Journal’s D.Live event, Intel will soon be shipping the world’s first family of processors designed from the ground up for artificial intelligence (AI): the Intel® Nervana™ Neural Network Processor family (formerly known as “Lake Crest”). This family of processors is over 3 years in the making, and on behalf of the team building it, I’d like to share a bit more insight on the motivation and design behind the world’s first neural network processor. > Mobileye is currently developing its fifth generation SoC, the EyeQ®5, to act as the vision central computer performing sensor fusion for Fully Autonomous Driving (Level 5) vehicles that will hit the road in 2020. To meet power consumption and performance targets, EyeQ® SoCs are designed in most advanced VLSI process technology nodes – down to 7nm FinFET in the 5th generation. MYRIAD 2 IS A MULTICORE, ALWAYS-ON SYSTEM ON CHIP THAT SUPPORTS COMPUTATIONAL IMAGING AND VISUAL AWARENESS FOR MOBILE, WEARABLE, AND EMBEDDED APPLICATIONS. THE VISION PROCESSING UNIT INCORPORATES PARALLELISM, INSTRUCTION SET ARCHITECTURE, AND MICROARCHITECTURAL FEATURES TO PROVIDE HIGHLY SUSTAINABLE PERFORMANCE EFFICIENCY ACROSS A RANGE OF COMPUTATIONAL IMAGING AND COMPUTER VISION APPLICATIONS, INCLUDING THOSE WITH LOW LATENCY REQUIREMENTS ON THE ORDER OF MILLISECONDS.
Myriad™ X is the first VPU to feature the Neural Compute Engine - a dedicated hardware accelerator for running on-device deep neural network applications. Interfacing directly with other key components via the intelligent memory fabric, the Neural Compute Engine is able to deliver industry leading performance per Watt without encountering common data flow bottlenecks encountered by other architectures.
Intel FPGA OpenCL and Solutions. Intel's Loihi test chip is the First-of-Its-Kind Self-Learning Chip. > The Loihi research test chip includes digital circuits that mimic the brain’s basic mechanics, making machine learning faster and more efficient while requiring lower compute power. Neuromorphic chip models draw inspiration from how neurons communicate and learn, using spikes and plastic synapses that can be modulated based on timing. This could help computers self-organize and make decisions based on patterns and associations.
Snapdragon 855 Mobile Platform
Our 4th generation on-device AI engine is the ultimate personal assistant for camera, voice, XR and gaming – delivering smarter, faster and more secure experiences. Utilizing all cores, it packs 3 times the power of its predecessor for stellar on-device AI capabilities. Greater than 7 trillion operations per second (TOPS)
NVIDIA TESLA T4 TENSOR CORE GPU
Powering the TensorRT Hyperscale Inference Platform.
NVIDIA Reveals Next-Gen Turing GPU Architecture: NVIDIA Doubles-Down on Ray Tracing, GDDR6, & More
at NVIDIA’s SIGGRAPH 2018 keynote presentation, company CEO Jensen Huang formally unveiled the company’s much awaited (and much rumored) Turing GPU architecture. The next generation of NVIDIA’s GPU designs, Turing will be incorporating a number of new features and is rolling out this year.
Nvidia’s DGX-2 System Packs An AI Performance Punch
Building Bigger, Faster GPU Clusters Using NVSwitches
Nvidia launched its second-generation DGX system in March. In order to build the 2 petaflops half-precision DGX-2, Nvidia had to first design and build a new NVLink 2.0 switch chip, named NVSwitch. While Nvidia is only shipping NVSwitch as an integral component of its DGX-2 systems today, Nvidia has not precluded selling NVSwitch chips to data center equipment manufacturers.
Nvidia's latest GPU can do 15 TFlops of SP or 120 TFlops with its new Tensor core architecture which is a FP16 multiply and FP32 accumulate or add to suit ML.
Nvidia is packing up 8 boards into their DGX-1for 960 Tensor TFlops.
Nvidia Volta - 架构看点 gives some insights of Volta architecture.
Now the open source DLA is available on Github and more information can be found here. > The NVIDIA Deep Learning Accelerator (NVDLA) is a free and open architecture that promotes a standard way to design deep learning inference accelerators. With its modular architecture, NVDLA is scalable, highly configurable, and designed to simplify integration and portability. The hardware supports a wide range of IoT devices. Delivered as an open source project under the NVIDIA Open NVDLA License, all of the software, hardware, and documentation will be available on GitHub. Contributions are welcome.
Samsung resently unveiled “The new Exynos 9810 brings premium features with a 2.9GHz custom CPU, an industry-first 6CA LTE modem and deep learning processing capabilities”.
Tesla is reportedly developing its own processor for artificial intelligence, intended for use with its self-driving systems, in partnership with AMD. Tesla has an existing relationship with Nvidia, whose GPUs power its Autopilot system, but this new in-house chip reported by CNBC could potentially reduce its reliance on third-party AI processing hardware.
Xilinx Launches the World's Fastest Data Center and AI Accelerator Cards
Xilinx launched Alveo, a portfolio of powerful accelerator cards designed to dramatically increase performance in industry-standard servers across cloud and on-premise data centers.
Xilinx provide "Machine Learning Inference Solutions from Edge to Cloud" and naturally claim their FPGA's are best for INT8 with one of their white papers.
Whilst performance per Watt is impressive for FPGAs, the vendors' larger chips have long had earth shatteringly high chip prices for the larger chips. Finding a balance between price and capability is the main challenge with the FPGAs.
TrueNorth is IBM's Neuromorphic CMOS ASIC developed in conjunction with the DARPA SyNAPSE program.
It is a manycore processor network on a chip design, with 4096 cores, each one simulating 256 programmable silicon "neurons" for a total of just over a million neurons. In turn, each neuron has 256 programmable "synapses" that convey the signals between them. Hence, the total number of programmable synapses is just over 268 million (228). In terms of basic building blocks, its transistor count is 5.4 billion. Since memory, computation, and communication are handled in each of the 4096 neurosynaptic cores, TrueNorth circumvents the von-Neumann-architecture bottlenecks and is very energy-efficient, consuming 70 milliwatts, about 1/10,000th the power density of conventional microprocessors. Wikipedia
With IBM POWER9, we’re all riding the AI wave
"With POWER9, we’re moving to a new off-chip era, with advanced accelerators like GPUs and FPGAs driving modern workloads, including AI...POWER9 will be the first commercial platform loaded with on-chip support for NVIDIA’s next-generation NVLink, OpenCAPI 3.0 and PCI-Express 4.0. These technologies provide a giant hose to transfer data."
ST preps second neural network IC
STMicroelectronics is designing a second iteration of the neural networking technology that the company reported on at the International Solid-State Circuits Conference (ISSCC) in February 2017.
ISSCC2017 Deep-Learning Processors文章学习 （一） is a reference.
S32 AUTOMOTIVE PLATFORM
S32 AUTOMOTIVE PLATFORM
The NXP S32 automotive platform is the world’s first scalable automotive computing architecture. It offers a unified hardware platform and an identical software environment across application domains to bring rich in-vehicle experiences and automated driving functions to market faster.
The S32V234 is our 2nd generation vision processor family designed to support computation intensive applications for image processing and offers an ISP, powerful 3D GPU, dual APEX-2 vision accelerators, security and supports SafeAssure™. S32V234 is suited for ADAS, NCAP front camera, object detection and recognition, surround view, machine learning and sensor fusion applications. S32V234 is engineered for automotive-grade reliability, functional safety and security measures to support vehicle and industrial automation.
Marvell Demonstrates Artificial Intelligence SSD Controller Architecture Solution
Marvell will demonstrate today at the Flash Memory Summit how it will provide artificial intelligence capabilities to a broad range of industries by incorporating NVIDIA’s Deep Learning Accelerator (NVDLA) technology in its family of data center and client SSD controllers.
MediaTek announced Helio P90, highlighting AI processing.
This article, "MediaTek Announces New Premium Helio P90 SoC", from AnandTech has more in-deepth analysis.
Kirin for Smart Phone
Kirin 980, the World's First 7nm Process Mobile AI Chipset
Introducing the Kirin 980, the world's first 7nm process mobile phone SoC chipset, the world’s first cortex-A76 architecture chipset, the world’s first dual NPU design, and the world’s first chipset to support LTE Cat.21. The Kirin 980 combines multiple technological innovations and leads the AI trend to provide users with impressive mobile performance and to create a more convenient and intelligent life.
HiSilicon Kirin 970 Processor annouced fearturing with dedicated Neural-network Processing Unit.
In this article,we can find more details about NPU in Kirin970.
Mobile Camera SoC
According to a Brief Data Sheet of Hi3559A V100ESultra-HD Mobile Camera SoC, it has:
Dual-core CNN@700 MHz neural network acceleration engine
Rockchip Released Its First AI Processor RK3399Pro -- NPU Performance up to 2.4TOPs
RK3399Pro adopted exclusive AI hardware design. Its NPU computing performance reaches 2.4TOPs, and indexes of both high performance and low consumption keep ahead: the performance is 150% higher than other same type NPU processor; the power consumption is less than 1%, comparing with other solutions adopting GPU as AI computing unit.
Google begins selling the $150 Coral Dev Board, a hardware kit for accelerated AI edge computing
If you’re a software dev looking to get a head start on AI development at the edge, why not try on Google’s new hardware for size? The search company today made available the Coral Dev Board, a $150 computer featuring a removable system-on-module with one of its custom tensor processing unit (TPU) AI chips.
Google's original TPU had a big lead over GPUs and helped power DeepMind's AlphaGo victory over Lee Sedol in a Go tournament. The original 700MHz TPU is described as having 95 TFlops for 8-bit calculations or 23 TFlops for 16-bit whilst drawing only 40W. This was much faster than GPUs on release but is now slower than Nvidia's V100, but not on a per W basis. The new TPU2 is referred to as a TPU device with four chips and can do around 180 TFlops. Each chip's performance has been doubled to 45 TFlops for 16-bits. You can see the gap to Nvidia's V100 is closing. You can't buy a TPU or TPU2.
Lately, Google is making Cloud TPUs available for use in Google Cloud Platform (GCP). Here you can find the latest banchmark result of Google TPU2.
Pixel Visual Core is Google’s first custom-designed co-processor for consumer products. It’s built into every Pixel 2, and in the coming months, we’ll turn it on through a software update to enable more applications to use Pixel 2’s camera for taking HDR+ quality pictures.
Tearing Apart Google’s TPU 3.0 AI Coprocessor
Google did its best to impress this week at its annual IO conference. While Google rolled out a bunch of benchmarks that were run on its current Cloud TPU instances, based on TPUv2 chips, the company divulged a few skimpy details about its next generation TPU chip and its systems architecture. The company changed from version notation (TPUv2) to revision notation (TPU 3.0) with the update, but ironically the detail we have assembled shows that the step from TPUv2 to what we will call TPUv3 probably isn’t that big; it should probably be called TPU v2r5 or something like that.
AI is pervasive today, from consumer to enterprise applications. With the explosive growth of connected devices, combined with a demand for privacy/confidentiality, low latency and bandwidth constraints, AI models trained in the cloud increasingly need to be run at the edge. Edge TPU is Google’s purpose-built ASIC designed to run AI at the edge. It delivers high performance in a small physical and power footprint, enabling the deployment of high-accuracy AI at the edge.
Other references are:
Google TPU3 看点
Google TPU 揭密
Google的神经网络处理器专利
脉动阵列 - 因Google TPU获得新生
Should We All Embrace Systolic Arrays?
Amazon may be developing AI chips for Alexa
The Information has a report this morning that Amazon is working on building AI chips for the Echo, which would allow Alexa to more quickly parse information and get those answers.
AWS Inferentia. High performance machine learning inference chip, custom designed by AWS.
AWS Inferentia provides high throughput, low latency inference performance at an extremely low cost. Each chip provides hundreds of TOPS (tera operations per second) of inference throughput to allow complex models to make fast predictions. For even more performance, multiple AWS Inferentia chips can be used together to drive thousands of TOPS of throughput. AWS Inferentia will be available for use with Amazon SageMaker, Amazon EC2, and Amazon Elastic Inference.
AWS FPGA instance
Amazon EC2 F1 is a compute instance with field programmable gate arrays (FPGAs) that you can program to create custom hardware accelerations for your application. F1 instances are easy to program and come with everything you need to develop, simulate, debug, and compile your hardware acceleration code, including an FPGA Developer AMI and Hardware Developer Kit (HDK). Once your FPGA design is complete, you can register it as an Amazon FPGA Image (AFI), and deploy it to your F1 instance in just a few clicks. You can reuse your AFIs as many times, and across as many F1 instances as you like.
Inside the Microsoft FPGA-based configurable cloud is also a good reference if want to know Microsoft's vision on FPGA in cloud.
This article "智慧云中的FPGA" gives and overview about FPGA used in AI aceleration in the cloud.
Drilling Into Microsoft’s BrainWave Soft Deep Learning Chip shows more details based on Microsoft's presentation on Hot Chips 2017.
Real-time AI: Microsoft announces preview of Project Brainwave
At Microsoft’s Build developers conference in Seattle this week, the company is announcing a preview of Project Brainwave integrated with Azure Machine Learning, which the company says will make Azure the most efficient cloud computing platform for AI.
Microsoft is hiring engineers to work on A.I. chip design for its cloud
Microsoft is following Google's lead in designing a computer processor for artificial intelligence, according to recent job postings.
A12 Bionic The smartest, most powerful chip in a smartphone.
A whole new level of intelligence. The A12 Bionic, with our next-generation Neural Engine, delivers incredible performance. It uses real-time machine learning to transform the way you experience photos, gaming, augmented reality, and more.
Apple unveiled the new processor powering the new iPhone 8 and iPhone X - the A11 Bionic. The A11 also includes dedicated neural network hardware that Apple calls a "neural engine", which can perform up to 600 billion operations per second.
Core ML is Apple's current sulotion for machine learning application.
Alibaba to launch own AI chip next year
Chinese internet giant sets up a semiconductor company and unveils plans to release its own artificial intelligence processor, as it looks to boost support for its cloud and Internet of Things businesses.
Alibaba is making its own neural network chip
Alibaba is developing its own neural network chip, the Ali-NPU, which will be used in AI applications, such as image video analysis, machine learning, and other scenarios.
FPGA Cloud server (Beta) is an computing instance of a field-programmable gate array (FPGA) that allows users to easily create FPGA design in minutes and create custom, dedicated hardware accelerators based on the Alibaba Cloud Elastic Computing Framework.
深度学习（Deep Learning）是一种多层计算模型，可以对复杂输入进行建模，在图像分类、语音识别、自然语言处理中取得了成果，FPGA 实例由于其细粒度并行的硬件特性，非常适合小批量数据的深度学习预测过程，以低功耗、低延迟、高性能著称，以 AlexNet 模型为例，使用 FPGA计算实例进行图片类别预测，速度比仅用CPU的普通实例快 8~15 倍。
Tencent cloud introduces FPGA instance(Beta), with three different specifications based on Xilinx Kintex UltraScale KU115 FPGA. They will provide more choices equiped with Inter FPGA in the future.
AN EARLY LOOK AT BAIDU’S CUSTOM AI AND ANALYTICS PROCESSOR
We’ve written much over the last few years about the company’s emphasis on streamlining deep learning processing, most notably with GPUs, but Baidu has a new processor up its sleeve called the XPU. For now, the device has just been demonstrated in FPGA, but if it continues to prove useful for AI, analytics, cloud, and autonomous driving the search giant could push it into a full-bore ASIC.
Baidu creates Kunlun silicon for AI
A pair of chips from the Chinese search giant are aimed at cloud and edge use cases. The company said it started developing a field-programmable gate array AI accelerator in 2011, and that Kunlun is almost 30 times faster. The chips are made with Samsung's 14nm process, have 512GBps memory bandwidth, and are capable of 260 tera operations per second at 100 watts.
FPGA Cloud Compute is open for beta test.
在百度内部，FPGA从2013年开始就应用在许多典型的深度学习模型中，如DNN，RNN，CNN，LSTM等，涵盖了语音识别，自然语言处理，推荐算法，图像识别等广泛的应用领域。百度FPGA云服务器中开放了基于FPGA的深度卷积神经网络加速服务，单卡提供3Tops的定点计算能力，支持典型深度卷积网络算子，如卷积、逆卷积、池化、拼接、切割等，有效加速典型网络结构如VggNet、GoogLeNet、ResNet等。我们基于FPGA的深度学习硬件，深度定制优化了主流深度学习平台如caffe等，用户可以直接将深度学习业务切换到FPGA平台，而无需考虑底层硬件细节。
Chinese tech giant Huawei unveils A.I. chips, taking aim at giants like Qualcomm and Nvidia
Huawei unveils two new artificial intelligence (AI) chips called the Ascend 910 and Ascend 310. The two chips are aimed at uses in data centers and internet-connected consumer devices, Rotating Chairman Eric Xu says at the Huawei Connect conference in Shanghai. The move pits the Chinese tech giant against major chipmakers including Qualcomm and Nvidia.
FPGA Accelerated Cloud Server, high performance FPGA instance is open for beta test.
FPGA云服务器提供CPU和FPGA直接的高达100Gbps PCIe互连通道，每节点提供8片Xilinx VU9P FPGA，同时提供FPGA之间高达200Gbps的Mesh光互连专用通道，让您的应用加速需求不再受到硬件限制。
This DLU that Fujitsu is creating is done from scratch, and it is not based on either the Sparc or ARM instruction set and, in fact, it has its own instruction set and a new data format specifically for deep learning, which were created from scratch. Japanese computing giant Fujitsu. Which knows a thing or two about making a very efficient and highly scalable system for HPC workloads, as evidenced by the K supercomputer, does not believe that the HPC and AI architectures will converge. Rather, the company is banking on the fact that these architectures will diverge and will require very specialized functions.
Nokia has developed the ReefShark chipsets for its 5G network solutions. AI is implemented in the ReefShark design for radio and embedded in the baseband to use augmented deep learning to trigger smart, rapid actions by the autonomous, cognitive network, enhancing network optimization and increasing business opportunities.
Facebook Is Forming a Team to Design Its Own Chips
Facebook Inc. is building a team to design its own semiconductors, adding to a trend among technology companies to supply themselves and lower their dependence on chipmakers such as Intel Corp. and Qualcomm Inc., according to job listings and people familiar with the matter.
HPE DEVELOPING ITS OWN LOW POWER “NEURAL NETWORK” CHIPS
In the context of a broader discussion about the company’s Extreme Edge program focused on space-bound systems, HPE’s Dr. Tom Bradicich, VP and GM of Servers, Converged Edge, and IoT systems, described a future chip that would be ideally suited for high performance computing under intense power and physical space limitations characteristic of space missions. To be more clear, he told us as much as he could—very little is known about the architecture, but there was some key elements he described.
Tesla’s new self-driving chip is here, and this is your best look yet
...And today, at Tesla’s Autonomy Investor Day in Palo Alto, California, the company gave the world its first, detailed glimpse at what Musk is now calling “the best chip in the world” — a 260 square millimeter piece of silicon, with 6 billion transistors, that the company claims offers 21 times the performance of the Nvidia chips it was using before.
Tesla’s new AI chip isn’t a silver bullet for self-driving cars
Processing power is important, but building chips could be an expensive distraction for Tesla
DynamIQ is embedded IP giant's answer to AI age. It may not be a revolutionary design but is important for sure.
ARM also provide a open source Compute Library contains a comprehensive collection of software functions implemented for the Arm Cortex-A family of CPU processors and the Arm Mali family of GPUs.
Arm Machine Learning Processor
Specifically designed for inference at the edge, the ML processor gives an industry-leading performance of 4.6 TOPs, with a stunning efficiency of 3 TOPs/W for mobile devices and smart IP cameras.
ARM Details "Project Trillium" Machine Learning Processor Architecture
Arm details more of the architecture of what Arm now seems to more consistently call their “machine learning processor” or MLP from here on now. The MLP IP started off a blank sheet in terms of architecture implementation and the team consists of engineers pulled off from the CPU and GPU teams.
DesignWare EV6x Embedded Vision Processors
处理器IP厂商的机器学习方案 - Synopsys
PowerVR Series2NX Neural Network Accelerator
Imagination Announces First PowerVR Series2NX Neural Network Accelerator Cores: AX2185 and AX2145
the company is announcing the first products in the 2NX NNA family: the higher-performance AX2185 and lower-cost AX2145.
CEVA-XM6 Fifth-generation computer vision and deep learning embedded platform
处理器IP厂商的机器学习方案 - CEVA
CEVA Announces NeuPro Neural Network IP
Ahead of CES CEVA announced a new specialised neural network accelerator IP called NeuPro.
Tensilica Vision DSPs for Imaging, Computer Vision, and Neural Networks
VeriSilicon’s Vivante VIP8000 Neural Network Processor IP Delivers Over 3 Tera MACs Per Second
神经网络DSP核的一桌麻将终于凑齐了
The v-MP6000UDX processor from Videantis is a scalable processor family that has been designed to run high-performance deep learning, computer vision, imaging and video coding applications in a low power footprint.
Chinese AI Chip Maker Cambricon Unveils New Cloud-Based Smart Chip
Chinese artificial intelligence chip maker Cambricon Technologies Corp Ltd has unveiled two new products, a cloud-based smart chip Cambricon MLU100 and a new version of its AI processor IP product Cambricon 1M, at a launching event in Shanghai on May 3rd.
Cambricon release new product page, including IP, Chip and Software tools
AI Chip Explosion: Cambricon’s Billion-Device Ambition
On November 6 in Beijing, China’s rising semiconductor company Cambricon released the Cambrian-1H8 for low power consumption computer vision application, the higher-end Cambrian-1H16 for more general purpose application, the Cambrian-1M for autonomous driving applications with yet-to-be-disclosed release date, and an AI system software named Cambrian NeuWare.
Chinese AI chip maker Horizon Robotics raises $600 million from SK Hynix, others
Chinese chip maker Horizon Robotics said on Wednesday it had raised $600 million in its latest funding round, bringing its valuation to $3 billion, amid a push from Chinese companies and the government to boost the semiconductor industry.
Dec. 20, Horizon Robotics annouced two chip products, "Journey" for ADAS and "Sunrise" for Smart Cameras.
Bitcoin Mining Giant Bitmain is developing processors for both training and inference tasks.
Bitmain’s newest product, the Sophon, may or may not take over deep learning. But by giving it such a name Zhan and his Bitmain co-founder, Jihan Wu, have signaled to the world their intentions. The Sophon unit will include Bitmain’s first piece of bespoke silicon for a revolutionary AI technology. If things go to plan, thousands of Bitmain Sophon units soon could be training neural networks in vast data centers around the world.
Chipintelli's first IC, CI1006, is designed for automatic speech recognition application.
Sequoia, Hillhouse, Yitu Technology Join $68M Series A Round In Chinese AI Chip Maker ThinkForce
Unisound raises US$100 million to fund AI, chip development
China’s AISpeech Raises $76M on Advanced Speech Tech; Eyes AI Chips
Chinese AI startup Rokid will mass produce their own custom AI chip for voice recognition
The world leading computer vision processing IC and system company, NextVPU, today unveiled AI vision processing IC N171. N171 is the flagship IC of NextVPU’s N1 series computer vison chips. As a VPU, N171 pushes the Edge AI computing limit further from many aspects. With powerful computing engines embedded, N171 has unprecedent geometry calculation and deep neural network processing capabilities, and can be widely used in surveillance, robots, drones, UGV, smart home, ADAS applications, etc.
Canaan's Kendryte is a series of AI chips which focuses on IoT.
Enflame Tech is a startup company based in Shanghai, China. It was established in March 2018 with two R&D centers in Shanghai and Beijing. Enflame is developing the deep learning accelerator SoCs and software stack, targeting AI training platform solutions for the Cloud service provider and the data centers.
Chinese tech startups Cloudpick, EEasy Tech snag Intel Capital funding
EEasy Technology Co. Ltd is an AI system-on-chip (SoC) design house and total solution provider. Its offerings include AI acceleration; image and graphic processing; video encoding and decoding; and mixed-signal ULSI design capabilities.
Cerebras is notable due to its backing from Benchmark and that its founder was the CEO of SeaMicro. It appears to have raised $25M and remains in stealth mode.
Wave’s Compute Appliance is capable to run TensorFlow at 2.9 PetaOPS/sec on their 3RU appliance. Wave refers to their processors at DPUs and an appliance has 16 DPUs. Wave uses processing elements it calls Coarse Grained Reconfigurable Arrays (CGRAs). It is unclear what bit width the 2.9 PetaOPS/s is referring to. Some details can be fund in their white paper.
After HotChips 2017, in the next plateform article "First In-Depth View of Wave Computing’s DPU Architecture, Systems", more details were discussed.
Graphcore raised $30M of Series-A late last year to support the development of their Intelligence Processing Unit, or IPU. Resently, co-founder and Chief Technology Officer, Simon Knowles, was invited to give a talk at the 3rd Research and Applied AI Summit (RAAIS) in London, showing interesting ideas behind their processor.
In a resent post, Graphcore shows "Preliminary IPU Benchmarks"
Simon Knowles, Graphcore CTO, spoke at the Scaled Machine Learning Conference at Stanford in March about 'Scaling Throughput Processors for Machine Intelligence'.
解密又一个xPU：Graphcore的IPU give some analysis on its IPU architecture.
The 2,048-core PEZY-SC2 sets a Green500 record
The SC2 is a second-generation chip featuring twice as many cores – i.e., 2,048 cores with 8-way SMT for a total of 16,384 threads. Operating at 1 GHz with 4 FLOPS per cycle per core as with the SC, the SC2 has a peak performance of 8.192 TFLOPS (single-precision). Both prior chips were manufactured on TSMC’s 28HPC+, however in order to enable the considerably higher core count within reasonable power consumption, PEZY decided to skip a generation and go directly to TSMC’s 16FF+ Technology.
Tenstorrent is a small Canadian start-up in Toronto claiming an order of magnitude improvement in efficiency for deep learning, like most. No real public details but they're are on the Cognitive 300 list.
ThinCI raises $65 million to develop its AI processors for automonomous and connected vehicles
Based in El Dorado Hills, California, ThinCI is developing silicon, but also software and a development kit that allows its hardware platform to be extended to a wide range of uses.
Founded in 2014, Newark, California startup Koniku has taken in $1.65 million in funding so far to become “the world’s first neurocomputation company“. The idea is that since the brain is the most powerful computer ever devised, why not reverse engineer it? Simple, right? Koniku is actually integrating biological neurons onto chips and has made enough progress that they claim to have AstraZeneca as a customer. Boeing has also signed on with a letter of intent to use the technology in chemical-detecting drones.
Adapteva has taken in $5.1 million in funding from investors that include mobile giant Ericsson. The paper "Epiphany-V: A 1024 processor 64-bit RISC System-On-Chip" describes the design of Adapteva's 1024-core processor chip in 16nm FinFet technology.
Knowm is actually setup as a .ORG but they appear to be pursuing a for-profit enterprise. The New Mexcio startup has taken in an undisclosed amount of seed funding so far to develop a new computational framework called AHaH Computing (Anti-Hebbian and Hebbian). The gory details can be found in this publication, but the short story is that this technology aims to reduce the size and power consumption of intelligent machine learning applications by up to 9 orders of magnitude.
A battery powered neural chip from Mythic with 50x lower power.
Founded in 2012, Texas-based startup Mythic (formerly known as Isocline) has taken in $9.5 million in funding with Draper Fisher Jurvetson as the lead investor. Prior to receiving any funding, the startup has taken in $2.5 million in grants. Mythic is developing an AI chip that “puts desktop GPU compute capabilities and deep neural networks onto a button-sized chip – with 50x higher battery life and far more data processing capabilities than competitors“. Essentially, that means you can give voice control and computer vision to any device locally without needing cloud connectivity.
Kalray Releases the Kalray Neural Network 3.0
Kalray (Euronext Growth Paris – ALKAL), a pioneer in processors for new intelligent systems, has announced the launch of the Kalray Neural Network 3.0 (KaNN), a platform for Artificial Intelligence application development. KaNN allows developers to seamlessly port their AI-based algorithms from well-known machine learning frameworks including Caffe, Torch and TensorFlow onto Kalray’s Massively Parallel Processor Array (MPPA) intelligent processor.
BrainChip Inc (CA. USA) was the first company to offer a Spiking Neural processor, which was patented in 2008 (patent US 8,250,011). The current device, called the BrainChip Accelerator is a chip intended for rapid learning. It is offered as part of the BrainChip Studio software. BrainChip is a publicly listed company as part of BrainChip Holdings Ltd.
This BDTi artical shows some information of aiWare IP of Aimotive .
Speaking of chips, AImotive and partner VeriSilicon are in the process of designing a 22 nm FD-SOI test chip, which is forecast to come out of GlobalFoundries' fab in Q1 2018 (Figure 4). It will feature a 1 TMAC/sec aiWare core, consuming approximately 25 mm2 of silicon area; a Vivante VIP8000-derivative processor core will inhabit the other half of the die, and between 2-4 GBytes of DDR4 SDRAM will also be included in the multi-die package. The convolution-tailored LAM in this test chip, according to Feher, will have the following specifications (based on preliminary synthesis results): 2,048 8x8 MACs Logic area (including input/output buffering logic, LAM control and MACs): 3.45mm2 Memory (on-chip buffer): in the range of 5-25mm2 depending on configuration (10-50 Mbits). Another interesting activity of Aimotive is Neural Network Exchange Format (NNEF).
DeepScale raises $3 million for perception AI to make self-driving cars safe
Leepmind is carrying out research on original chip architectures in order to implement Neural Networks on a circuit enabling low power DeepLearning
A crowdfunding effort for Snickerdoodle raised $224,876 and they’re currenty shipping. If you pre-order one, they’ll deliver it by summer. The palm-sized unit uses the Zynq “System on Chip” (SoC) from Xilinix.
NovuMind combines big data, high-performance, and heterogeneous computing to change the Internet of Things (IoT) into the Intelligent Internet of Things (I²oT). this video is the description and demos of NovuMind FPGA AI Accelerator.
Reduced Energy Microsystems are developing lower power asynchronous chips to suit CNN inference. REM was Y Combinator's first ASIC venture according to TechCrunch.
TeraDeep is building an AI Appliance using its deep learning FPGA’s acceleration. The company claims image recognition performance on AlexNet to achieve a 2X performance advantage compared with large GPUs, while consuming 5X less power. When compared to Intel’s Xeon processor, TeraDeep’s Accel technology delivers 10X the performance while consuming 5X less power.
Deep Vision is bulding low-power chips for deep learning. Perhaps one of these papers by the founders have clues, "Convolution Engine: Balancing Efficiency & Flexibility in Specialized Computing" [2013] and "Convolution Engine: Balancing Efficiency and Flexibility in Specialized Computing" [2015].
Groq is founded by Ex-googlers, who designed Google TPU.
Groq's website claims that its first chip will run 400 trillion operations per second with 8TOP/s per Watt power efficiency.
Face Recognition System “K-Eye” Presented by KAIST
从ISSCC Deep Learning处理器论文到人脸识别产品
Kneron to Accelerate Edge AI Development with more than 10 Million USD Series A Financing
Here is the demo of their chip, usb dongle and development board.
According to this article, "Esperanto exits stealth mode, aims at AI with a 4,096-core 7nm RISC-V monster"
Although Esperanto will be licensing the cores they have been designing, they do plan on producing their own products. The first product they want to deliver is the highest TeraFLOP per Watt machine learning computing system. Ditzel noted that the overall design is scalable in both performance and power. The chips will be designed in 7nm and will feature a heterogeneous multi-core architecture.
According to the linkedin page of its CEO, former SPARC developer in ORACLE, SambaNova Systems is a computing startup focused on building machine learning and big data analytics platforms. SambaNova's software-defined analytics platform enables optimum performance for any ML training, inference or analytics models.
The red-hot AI hardware space gets even hotter with $56M for a startup called SambaNova Systems
SambaNova is the product of technology from Kunle Olukotun and Chris Ré, two professors at Stanford, and led by former Oracle SVP of development Rodrigo Liang, who was also a VP at Sun for almost 8 years.
GreenWaves Technologies develops IoT Application Processors based on Open Source IP blocks enabling content understanding applications on embedded, battery-operated devices with unmatched energy efficiency. Our first product is GAP8. GAP8 provides an ultra-low power computing solution for edge devices carrying out inference from multiple, content rich sources such as images, sounds and motions. GAP8 can be used in a variety of different applications and industries.
Light-Powered Computers Brighten AI’s Future
Optical computers may have finally found a use—improving artificial intelligence
Lightmatter aims to reinvent AI-specific chips with photonic computing and $11M in funding
It takes an immense amount of processing power to create and operate the “AI” features we all use so often, from playlist generation to voice recognition. Lightmatter is a startup that is looking to change the way all that computation is done — and not in a small way. The company makes photonic chips that essentially perform calculations at the speed of light, leaving transistors in the dust. It just closed an $11 million Series A.
TORONTO, Canada/NUREMBERG, Germany – FEB 21st, 2018 – Think Silicon®, a leader in developing ultra-low power graphics IP technology, will demonstrate a prototype of NEMA® xNN, the world’s first low-power ‘Inference Accelerator’ Vision Processing Unit for artificial intelligence, convolutional neural networks at Embedded World 2018.
Innogrit Technologies Incorporated is a startup seting out to solve the data storage and data transport problem in artificial intelligence and other big data applications through innovative integrated circuit (IC) and system solutions: Extracts intelligence from correlated data and unlocks the value in artificial intelligence systems; Reduces redundancy in big data and improves system efficiency for artificial intelligence applications; Brings networking capability to storage devices and offers unparalleled performance at large scales; Performs data computation within storage devices and boosts performance of large data centers.
Kortiq is a startup providing "FPGA based Neural Network Engine IP Core and The scalable Solution for Low Cost Edge Machine Learning Inference for Embedded Vision". Recently, they revealed some comparison data. You can also find the Preliminary Datasheet of their AIScaleCDP2 IP Core on their website.
Hailo unveils Hailo-8, an edge chip custom-designed for AI workloads
......Hailo-8 is capable of 26 tera operations per second (TOPs) ...... In one preliminary test at an image resolution of 224 x 224, the Hailo-8 processed 672 frames per second compared with the Xavier AGX’s 656 frames and sucked down only 1.67 watts (equating to 2.8 TOPs per watt) versus the Nvidia chip’s 32 watts (0.14 TOPs per watt)......
Silicon Startup Raises ‘Prodigy’ for Hyperscale AI Workloads
Silicon Valley-based Tachyum Inc., which has been emerging from stealth over the last year and a half, is unveiling a processor codenamed “Prodigy,” said to combine features of both CPUs and GPUs in a way that offers a purported 10x performance-per-watt advantage over current technologies. The company is primarily focused on the hyperscale datacenter market, but has aspirations to support brainier applications, noting that “Prodigy will enable a super-computational system for real-time full capacity human brain neural network simulation by 2020.”
Startup AI Chip Passes Road Test
AlphaICs designed an instruction set architecture (ISA) optimized for deep-learning, reinforcement-learning, and other machine-learning tasks. The startup aims to produce a family of chips with 16 to 256 cores, roughly spanning 2 W to 200 W.
Syntiant: Analog Deep Learning Chips
Startup Syntiant Corp. is an Irvine, Calif. semiconductor company led by former top Broadcom engineers with experience in both innovative design and in producing chips designed to be produced in the billions, according to company CEO Kurt Busch.
The Goya chip can process 15,000 ResNet-50 images/second with 1.3-ms latency at a batch size of 10 while running at 100 W. That compares to 2,657 images/second for an Nvidia V100 and 1,225 for a dual-socket Xeon 8180. At a batch size of one, Goya handles 8,500 ResNet-50 images/second with a 0.27-ms latency.
Baidu Backs Neuromorphic IC Developer
MUNICH — Swiss startup aiCTX has closed a $1.5 million pre-A funding round from Baidu Ventures to develop commercial applications for its low-power neuromorphic computing and processor designs and enable what it calls “neuromorphic intelligence.” It is targeting low-power edge-computing embedded sensory processing systems.
AI startup Flex Logix touts vastly higher performance than Nvidia
Four-year-old startup Flex Logix has taken the wraps off its novel chip design for machine learning. CEO Geoff Tate describes how the chip may take advantage of an "explosion" of inferencing activity in "edge computing," and how Nvidia can't compete on performance.
Dec. 12, 2018, Tokyo Japan – Preferred Networks, Inc. (“PFN”, Head Office: Tokyo, President & CEO: Toru Nishikawa) announces that it is developing MN-Core (TM), a processor dedicated to deep learning and will exhibit this independently developed hardware for deep learning, including the MN-Core chip, board, and server, at the SEMICON Japan 2018, held at Tokyo Big Site.
AI Startup Cornami reveals details of neural net chip
Stealth startup Cornami on Thursday revealed some details of its novel approach to chip design to run neural networks. CTO Paul Masters says the chip will finally realize the best aspects of a technology first seen in the 1970s.
AI chip startup offers new edge computing solution
Anaflash Inc. (San Jose, CA) is a startup company that has developed a test chip to demonstrate analog neurocomputing taking place inside logic-compatible embedded flash memory.
Optalysys launches world’s first commercial optical processing system, the FT:X 2000
Optalysys develops Optical Co-processing technology which enables new levels of processing capability delivered with a vastly reduced energy consumption compared with conventional computers. Its first coprocessor is based on an established diffractive optical approach that uses the photons of low-power laser light instead of conventional electricity and its electrons. This inherently parallel technology is highly scalable and is the new paradigm of computing.
Eta Compute Debuts Spiking Neural Network Chip for Edge AI
Chip can learn on its own and inference at 100-microwatt scale, says company at Arm TechCon.
1. pytorch/glow
2. TVM:End to End Deep Learning Compiler Stack
3. Google Tensorflow XLA
4. Nvidia TensorRT
5. PlaidML
6. nGraph
7. MIT Tiramisu compiler
8. ONNC (Open Neural Network Compiler)
9. Multi-Level Intermediate Representation
- DAWNBench:An End-to-End Deep Learning Benchmark and Competition Image Classification (ImageNet)
- Fathom:Reference workloads for modern deep learning methods
- MLPerf:A broad ML benchmark suite for measuring performance of ML software frameworks, ML hardware accelerators, and ML cloud platforms. Now you can find MLPerf v0.5 results here.
- AI Matrix
- AI-Benchmark
- AIIABenchmark
- FPGAs and AI processors: DNN and CNN for all
- 12 AI Hardware Startups Building New AI Chips
- Tutorial on Hardware Architectures for Deep Neural Networks
- Neural Network Accelerator Inference
- "White Paper on AI Chip Technologies 2018". You can download it from here, or Google drive.