Nvidia showed up with Hopper at MLPerf, and most competitors didn’t show except for Intel. // $NVDA $INTC $GOOGL // #AI #machinelearning #mlperf // @MLCommons
HARDWARE
-
NVIDIA Performance Tuning GPUs CPUs Roofline Model SC22
By
–
Join NVIDIA at #SC22 in this session on performance tuning with Roofline Model on #GPUs and #CPUs. Learn more:
-
LearnOpenCV Backs RAE Project on Kickstarter with Luxonis
By
–
I just backed rae on @Kickstarter https://
kickstarter.com/projects/openc
v/rae-0?ref=thanks-tweet
… @luxonis -
Implementing Dataflow Programs on Cerebras CS-2 Architecture
By
–
Join Leighton at #SC22 for a discussion on using our SDK to implement dataflow programs on the Cerebras architecture. During this event, he will walk through how to map a general matrix-vector product (GEMV) to a Cerebras CS-2, using our fabric simulator. #HPC #SC22
-
IRS Enhances Fraud Detection with NVIDIA GPUs and Cloudera CDP
By
–
Check out our webinar with @cloudera and learn how the IRS turbocharged fraud detection with NVIDIA #GPUs powering CDP: https://
nvda.ws/3gMrM48 -
Colossal-AI’s Cost-Effective AIGC Solution Accelerates Diffusion Pretraining
By
–
Almost 7X Cheaper! Colossal-AI’s Open Source Solution Accelerates AIGC at a Low-Cost Diffusion Pretraining and Hardware Fine-Tuning Can Be
-
Foundation Models Deliver 22x Faster Enterprise AI Results
By
–
With high-performance, high-accuracy results delivered to the enterprise 22x faster, Foundation Models are driving business forward. Watch @SambaNovaAI executives unveil the future of artificial intelligence from AI Hardware Summit 2022. https://
youtu.be/C2P1mKSL85g -
CoreWeave Offers NVIDIA HGX H100 Cloud Instances
By
–
NVIDIA Partner Network Elite Cloud Service Provider @CoreWeave is among the first to offer cloud instances with NVIDIA HGX H100 #supercomputing. Read the announcement: https://
nvda.ws/3hpNc7E -
Cerebras Wafer-Scale Cluster Achieves Linear Scaling Performance
By
–
Distributed training across thousands of traditional accelerators suffers diminishing returns as more compute is added. Join Natalia Vassilieva at #SC22 as she presents the Cerebras Wafer-Scale Cluster, which achieves near-perfect linear scaling across additional cores
#AI #SC22 -
Databricks optimizes cloud VMs with Azure and AMD technologies
By
–
Two of our favorite things: saving our users time AND money That's why we combined the latest technologies from @Azure
, Databricks, and @AMD to help users take advantage of the new Lasv3-series VMs with the Databricks Runtimes. See for yourself https://
dbricks.co/3TdOkIZ