“Draft:Inaccel”的意思、由来-开放百科全书

词条

Draft:Inaccel

释义

References

{{AFC submission|d|corp|u=Ckachris|ns=118|decliner=JC7V7DC5768|declinets=20181119205456|ts=20181113112739}} {{AFC submission|d|corp|u=Ckachris|ns=118|decliner=Whispering|declinets=20181112134905|small=yes|ts=20181112105234}} {{AFC comment|1=Need SIGCOV, independent non passing coverage from newspapers, books, journals . Please see Help desk or Teahouse for more help if need be. JC7V (talk) 20:54, 19 November 2018 (UTC)}}
{{Infobox company
|logo = File:Inaccel-logo.png
|name = inaccel
|industry = Computer software
|founder = Christoforos Kachris, Elias Koromilas, Ioannis Stamelos
|location_city = Wilmington, Delaware
|location_country = United States
|foundation = {{Start date|2018}}
|area_served =
|num_employees = 6
|homepage = {{URL|https://www.inaccel.com/|www.inaccel.com}}
}}

InAccel is an ISV company that develops hardware accelerators for FPGAs on the cloud like AWS.

The company is mainly specialized on developing hardware accelerators in the form of IP cores for machine learning applications.

Currently InAccel is offering 2 hardware accelerators for machine learning:

logistic regression
k-means clustering

The hardware accelerator provided from InAccel can be used to speedup applications built on Apache Spark.

The provided core can achieve up to 10x speedup for widely used applications like machine learning.

In June, 2018, InAccel demonstrated at Spark Summit.^[1] how to utilize FPGAs without code changes for Spark.

Based on one of the widely used benchmarks for classification using MNIST, it showed that the accelerators can achieve 3x speedup for logistic regression and 2x lower TCO^[2].

InAccel was founded in 2018 and it is based on Delaware, USA^[3] and Athens, Greece.

InAccel secured a seed funding of 500,000 euros from Marathon VC in July 2018^[4].

InAccel is part of the Xilinx Alliance and is also an official standard AWS technology partner^[5]

References

1. ^Spark Summit 2018, https://databricks.com/session/apache-spark-acceleration-using-hardware-resources-in-the-cloud-seamlessly
2. ^Accelerating Data Science with FPGAs, https://medium.com/@kachris/a-new-way-to-speedup-your-machine-learning-applications-f843a0ee2829
3. ^Bloomberg, https://www.bloomberg.com/research/stocks/private/snapshot.asp?privcapId=573401117
4. ^Crunchbase, https://www.crunchbase.com/organization/inaccel
5. ^AWS technology partner, https://aws.amazon.com/marketplace/seller-profile?id=c162d4fc-abfe-4574-98cd-44438e6e66cb

Category:Big data companies

随便看

开放百科全书收录14589846条英语、德语、日语等多语种百科知识，基本涵盖了大多数领域的百科知识，是一部内容自由、开放的电子版国际百科全书。