graphcore

graphcore

@graphcoreai

Followers6.3K
Following658

We invented the Intelligence Processing Unit (IPU) to let innovators create next generation machine intelligence.

Bristol, England
Joined on August 12, 2016
Statistics

We looked inside some of the tweets by @graphcoreai and here's what we found interesting.

Inside 100 Tweets

Time between tweets:
11 hours
Average replies
0
Average retweets
2
Average likes
4
Tweets with photos
30 / 100
Tweets with videos
0 / 100
Tweets with links
0 / 100

Rankings (sorted by number of followers)

129. in country United Kingdom and category Electronics

Bulk Synchronous Parallel (BSP) algorithms are parallel computational algorithms which run in 3 main phases: Compute, Synchronise & Exchange. Find out more about BSP and the @graphcoreai IPU ⬇️ https://t.co/5CGvI7quXs @_ProfCalculus #ParallelCompute #AI #Programming #AIHardware

One week to go until Graphcore Research's poster session at #NeurIPS2020! The authors of the paper "Improving Neural Network Training in Low Dimensional Random Bases" will discuss novel ways to train neural networks efficiently at scale. Read the blog 👉 https://t.co/VXtClrliEU

Which AI breakthroughs can we expect to see in 2021? 

Catch @graphcoreai's Carlo Luschi at our virtual booth at #NeurIPS2020 for insights into future directions for machine intelligence ⬇️🔮

 https://t.co/8exLnneaFN

#GraphcoreResearch #NeurIPS #AIResearch #AI #ML #IPU https://t.co/dEjFhHjdZ8

Which AI breakthroughs can we expect to see in 2021? Catch @graphcoreai's Carlo Luschi at our virtual booth at #NeurIPS2020 for insights into future directions for machine intelligence ⬇️🔮 https://t.co/8exLnneaFN #GraphcoreResearch #NeurIPS #AIResearch #AI #ML #IPU https://t.co/dEjFhHjdZ8

Join @graphcoreai online at #NeurIPS2020 next week! We’re holding a series of talks at our booth on future directions of AI and new approaches to machine intelligence. Check out the schedule here💡👉 https://t.co/XutEGm0m9l #GraphcoreResearch #NeurIPS #AI #IPU #NeuralNetworks

Designed from the ground up for high performance training and inference workloads, the IPU-M2000 is @graphcoreai's core building block for AI infrastructure.

🤖 1 PetaFlop of AI compute
🧠 up to 450GB Exchange Memory
💬 2.8Tbps IPU-Fabric communications

https://t.co/9Ztpod0Kj3 https://t.co/gZSLM7o1Xv

Designed from the ground up for high performance training and inference workloads, the IPU-M2000 is @graphcoreai's core building block for AI infrastructure. 🤖 1 PetaFlop of AI compute 🧠 up to 450GB Exchange Memory 💬 2.8Tbps IPU-Fabric communications https://t.co/9Ztpod0Kj3 https://t.co/gZSLM7o1Xv

Get to know the Poplar SDK – our complete software stack co-designed with the IPU for scalable machine intelligence. Poplar takes care of all the scaling and optimisation on our MK2 IPU systems, allowing #developers to focus on the model and results. https://t.co/1kojNtNGI1

Designed to eliminate communication bottlenecks, @graphcoreai's IPU-Fabric™ allows up to 64,000 IPUs to operate on machine intelligence workloads as a single, high-performance and ultra-fast cohesive unit. https://t.co/hAcCMSxKVu #IPUFabric #IPUSystems #Supercomputing #AI

Apply before Sunday at midday for these @graphcoreai graduate roles! 🤖🔎

Graduate Product Test Engineer https://t.co/CTXJ9Liq8s

Graduate Silicon Engineer https://t.co/xXHMWM7LVN

Graduate Customer Software Engineer https://t.co/TwWG7XdAJd

#WorkThatMatters #AICareers #GradJobs https://t.co/MTGnkATJ3L

Apply before Sunday at midday for these @graphcoreai graduate roles! 🤖🔎 Graduate Product Test Engineer https://t.co/CTXJ9Liq8s Graduate Silicon Engineer https://t.co/xXHMWM7LVN Graduate Customer Software Engineer https://t.co/TwWG7XdAJd #WorkThatMatters #AICareers #GradJobs https://t.co/MTGnkATJ3L

Some impressive accelerations reported in @Ford's recent paper tracking pedestrian direction and speed using DNNs and @graphcoreai IPUs.

Read the research paper 👉 https://t.co/I0XpOkTdau

Watch the video 📽️ https://t.co/JHVxiPcgTE

#NeuralNetworks #AI #ML #AutonomousDriving https://t.co/fxghfK1XAT

Some impressive accelerations reported in @Ford's recent paper tracking pedestrian direction and speed using DNNs and @graphcoreai IPUs. Read the research paper 👉 https://t.co/I0XpOkTdau Watch the video 📽️ https://t.co/JHVxiPcgTE #NeuralNetworks #AI #ML #AutonomousDriving https://t.co/fxghfK1XAT

We are delighted to be partnering with #Graphcore and are really looking forward to working together. Graphcore's Intelligence Processing Unit is specifically designed for #AI compute to accelerate machine intelligence. More information at https://t.co/ynnltM6LCE #MEGWARE https://t.co/ekpwK6OuJa

We’re excited to build on our support for customers in Germany with the addition of MEGWARE as a @graphcoreai Partner. Learn more about our expansion into Germany ⬇️ @MegwareComputer #PartnerProgram #Germany #IPU #AIAtScale #AI https://t.co/yinYhuMDG7

Our collection of world-class Developer resources was created by @graphcoreai AI experts to help developers program the IPU, port existing models and reproduce results with ease. Check out our Developer Portal 👇https://t.co/vRkrDrOvep #Programming #Poplar #ML #AI #Developer

Because of the IPU’s fine-grained compute capabilities, simplified optimisers can be used at low batch sizes to perform tasks where larger batch sizes would require more complex optimisers. Learn more about evaluating batch sizes for the #IPU 📽️ https://t.co/vnuauT0YSe #AI #ML

If you're a recent #graduate looking to get into #ArtificialIntelligence research, @graphcoreai, a #BackedbyDTC portfolio company and #AI innovator is looking to hire. More information on the available position here https://t.co/qZSMX8Fz4A

Bulk synchronous parallel is a programming methodology for parallel algorithms. Watch our latest #Developer video to learn how BSP is implemented on the @graphcoreai IPU 📽️⬇️ https://t.co/tpkxuzUYwV @_ProfCalculus #ParallelCompute #AI #Programming #AIHardware #IPU

Interested in AI and computer science? Join tomorrow's @graphcoreai STEM Women Virtual Insight Event to learn how you could contribute to future breakthroughs in machine intelligence 👇👩‍💻

https://t.co/SvNrnMfYxu

#swgraphcore20 #WomeninSTEM #WomeninAI #ComputerScience #AI https://t.co/tuyRecfsOj

Interested in AI and computer science? Join tomorrow's @graphcoreai STEM Women Virtual Insight Event to learn how you could contribute to future breakthroughs in machine intelligence 👇👩‍💻 https://t.co/SvNrnMfYxu #swgraphcore20 #WomeninSTEM #WomeninAI #ComputerScience #AI https://t.co/tuyRecfsOj

A new paper from @graphcoreai Research looks at novel approaches to training neural networks efficiently, which could pave the way for scaling to much larger models in future. Watch the video 📽️💡 https://t.co/YXc0kVseie #NeurIPS #NeurIPS2020 #AI #DeepLearning #NeuralNetworks

At #SC20, @graphcoreai's Simon Knowles discussed the future of AI and gave some fascinating insights into how we can train tomorrow's superhuman machine learning models efficiently. Watch his talk here 🤖👇 #DeepLearning #MachineLearning #IPU #AI https://t.co/eQujAIyS1L

LeMagIT
12 days ago

L’intelligence artificielle malmenée par la Covid-19 https://t.co/oDCLvWKjin

Next Page