Skip to content

Commit 0314c47

Browse files
vaughanknightdanuw
authored andcommitted
Minor updates, included adopters page link
Minor updates, included adopters page link
1 parent a07870f commit 0314c47

File tree

1 file changed

+3
-8
lines changed

1 file changed

+3
-8
lines changed

README.md

Lines changed: 3 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -6,8 +6,7 @@ The Carbon Aware SDK is a toolset to help you measure the carbon emissions of yo
66
![Carbon Aware Software](./images/carbon-aware-software.png)
77

88

9-
By knowing the carbon emissions of the energy that powers your applications,
10-
you and your organisation can leverage greener energy sources to reduce your CO2 emissions by:
9+
By knowing the carbon emissions of the energy that powers your applications, you and your organisation can leverage greener energy sources to reduce your CO2 emissions by:
1110

1211
* Building AI models when carbon emissions are lower
1312
* Deploying software into the cloud in locations that have greener energy sources
@@ -80,14 +79,10 @@ deployment in the greenest location.
8079

8180
The Carbon Aware SDK is being used by large and small companies around the
8281
world. Some of the world’s biggest enterprises and software companies, through
83-
to start-ups.
84-
85-
Both UBS and Vestas have used the SDK, with further details over on the [adopters page](./docs/adopters.md).
82+
to start-ups. Both UBS and Vestas have used the SDK, with further details over on the [adopters overview](./docs/adopters.md).
8683

8784
Machine Learning (ML) workloads are a great example of long running compute
88-
intensive workloads, that often are also not time critical. By moving these
89-
workloads to a different time, the carbon emissions from the ML training can be
90-
reduced by up to 15%, and by moving the location of the training this can be
85+
intensive workloads, that often are also not time critical. By moving these workloads to a different time, the carbon emissions from the ML training can be reduced by up to 15%, and by moving the location of the training this can be
9186
reduced even further, at times by up to 50% or more.
9287

9388
## What does the SDK/API provide that 3rd party data providers such as WattTime or ElectricityMaps do not?

0 commit comments

Comments
 (0)