A Performance Perspective for Graviton Based Lambda Functions
Introduction
AWS Lambda is constantly growing and, in time, things start to push limits. AWS has been working on new hardware architectures to make price-performance more efficient. AWS Graviton2 is a custom silicon-based 64-Bit ARM processor that’s coming to AWS Lambda which will help Serverless customers to have the same leverage as other services that can be built on AWS Graviton2. Lambda functions powered by Graviton2 offer up to 34% better price performance over x86-based Lambda functions.
As Thundra, we didn’t want to miss the opportunity to become a launch partner for such improvement and decided to test it out ourselves. The purpose of this article is to run simple benchmarks on AWS Graviton2 in comparison to x86_64 based architecture.
If you want to learn more about the AWS Graviton2 launch for AWS Lambda, make sure to visit the announcement page.
Benchmarks
Our test scenario consists of the below Python script that calculates the nth Fibonacci number to exhaust the CPU. It’s a bit cliche, but it’s simple. To keep the test simple and consistent, we’ve decided to calculate the 35th number as the sweet spot of our testing.
```python
import os
import multiprocessing
def handler(event, context):
nth = 35
processes = list()
for i in range(int(os.getenv('PROCESS_COUNT', 1))):
print(f'starting process #{i}.')
process = multiprocessing.Process(
target=calculateFib, args=(nth,))
processes.append(process)
process.start()
for i, process in enumerate(processes):
print(f'joining process #{i}.')
process.join()
print(f'finished process #{i}.')
response = {
"statusCode": 200
}
return response
def calculateFib(nth):
if nth <= 2:
return 1
else:
return calculateFib(nth - 2) + calculateFib(nth - 1)
```
We’ve run this script for all the combinations of (Process Count, RAM, Architecture) populated from the following data.
- Process Count: 1, 2, 4, 8, 16
- RAM: 256, 512, 1024, 2048, 4096, 8192, 10240 MB
- Architecture: x86_64, arm64 (AWS Graviton2)
After running all the combinations, we’ve noticed a couple of trends. The most important one of these is probably the billed duration time. There is a significant difference between the two architectures. Assuming there is some error margin and that there are other things at play here, it’s still impressive with Graviton2.
As one might guess, duration time will decrease as the RAM gets bigger. For single-process invocation, you don’t get much gain from the CPUs after 2048 MB RAM. For 16 process invocations, the margin gets bigger as you can see in the graphs below.
The raw data for billed duration gives you the bigger picture.
|
x86_64 |
256 MB |
512 MB |
1024 MB |
2048 MB |
4096 MB |
8192 MB |
10240 MB |
|
1 |
18952 |
9571 |
4752 |
2709 |
2746 |
2828 |
2731 |
|
2 |
38428 |
19242 |
9687 |
4803 |
2742 |
2737 |
2790 |
|
4 |
76149 |
37965 |
19007 |
9498 |
4904 |
2777 |
2804 |
|
8 |
155968 |
76986 |
38184 |
18914 |
9578 |
4891 |
4359 |
|
16 |
305743 |
156496 |
76358 |
38699 |
19274 |
9703 |
7723 |
|
arm64 |
256 MB |
512 MB |
1024 MB |
2048 MB |
4096 MB |
8192 MB |
10240 MB |
|
1 |
15570 |
7830 |
4105 |
2481 |
2478 |
2513 |
2494 |
|
2 |
30856 |
15482 |
8259 |
4062 |
2482 |
2482 |
2572 |
|
4 |
61394 |
30588 |
15378 |
7899 |
4270 |
2491 |
2498 |
|
8 |
121505 |
61284 |
30533 |
15398 |
7942 |
4430 |
3697 |
|
16 |
244346 |
121374 |
61116 |
30660 |
15409 |
8124 |
6541 |
In the end, we can clearly see the advantage of AWS Graviton2 on CPU-intensive jobs. Overall, when the AWS Lambda is running on arm64 we’d get a considerable amount of cost-saving with the right configuration.
Outro
In conclusion, we can see the benefit of AWS Graviton2 for both price and performance. Of course, in real life, we would have much more complicated scenarios and things to worry about. However, it should give us a rough idea of how much improvement we can expect. Getting started with AWS Graviton is pretty easy if you just follow this GitHub repository. To be able to debug your applications quickly, go to https://start.thundra.io/ and sign up for a free account.
SHARE
×
SUBSCRIBE TO OUR BLOG
Get our new blogs delivered straight to your inbox.
THANKS FOR SIGNING UP!
We’ll make sure to share the best materials crafted for you!
Introduction
AWS Lambda is constantly growing and, in time, things start to push limits. AWS has been working on new hardware architectures to make price-performance more efficient. AWS Graviton2 is a custom silicon-based 64-Bit ARM processor that’s coming to AWS Lambda which will help Serverless customers to have the same leverage as other services that can be built on AWS Graviton2. Lambda functions powered by Graviton2 offer up to 34% better price performance over x86-based Lambda functions.
As Thundra, we didn’t want to miss the opportunity to become a launch partner for such improvement and decided to test it out ourselves. The purpose of this article is to run simple benchmarks on AWS Graviton2 in comparison to x86_64 based architecture.
If you want to learn more about the AWS Graviton2 launch for AWS Lambda, make sure to visit the announcement page.
Benchmarks
Our test scenario consists of the below Python script that calculates the nth Fibonacci number to exhaust the CPU. It’s a bit cliche, but it’s simple. To keep the test simple and consistent, we’ve decided to calculate the 35th number as the sweet spot of our testing.
```python
import os
import multiprocessing
def handler(event, context):
nth = 35
processes = list()
for i in range(int(os.getenv('PROCESS_COUNT', 1))):
print(f'starting process #{i}.')
process = multiprocessing.Process(
target=calculateFib, args=(nth,))
processes.append(process)
process.start()
for i, process in enumerate(processes):
print(f'joining process #{i}.')
process.join()
print(f'finished process #{i}.')
response = {
"statusCode": 200
}
return response
def calculateFib(nth):
if nth <= 2:
return 1
else:
return calculateFib(nth - 2) + calculateFib(nth - 1)
```
We’ve run this script for all the combinations of (Process Count, RAM, Architecture) populated from the following data.
- Process Count: 1, 2, 4, 8, 16
- RAM: 256, 512, 1024, 2048, 4096, 8192, 10240 MB
- Architecture: x86_64, arm64 (AWS Graviton2)
After running all the combinations, we’ve noticed a couple of trends. The most important one of these is probably the billed duration time. There is a significant difference between the two architectures. Assuming there is some error margin and that there are other things at play here, it’s still impressive with Graviton2.
As one might guess, duration time will decrease as the RAM gets bigger. For single-process invocation, you don’t get much gain from the CPUs after 2048 MB RAM. For 16 process invocations, the margin gets bigger as you can see in the graphs below.
The raw data for billed duration gives you the bigger picture.
|
x86_64 |
256 MB |
512 MB |
1024 MB |
2048 MB |
4096 MB |
8192 MB |
10240 MB |
|
1 |
18952 |
9571 |
4752 |
2709 |
2746 |
2828 |
2731 |
|
2 |
38428 |
19242 |
9687 |
4803 |
2742 |
2737 |
2790 |
|
4 |
76149 |
37965 |
19007 |
9498 |
4904 |
2777 |
2804 |
|
8 |
155968 |
76986 |
38184 |
18914 |
9578 |
4891 |
4359 |
|
16 |
305743 |
156496 |
76358 |
38699 |
19274 |
9703 |
7723 |
|
arm64 |
256 MB |
512 MB |
1024 MB |
2048 MB |
4096 MB |
8192 MB |
10240 MB |
|
1 |
15570 |
7830 |
4105 |
2481 |
2478 |
2513 |
2494 |
|
2 |
30856 |
15482 |
8259 |
4062 |
2482 |
2482 |
2572 |
|
4 |
61394 |
30588 |
15378 |
7899 |
4270 |
2491 |
2498 |
|
8 |
121505 |
61284 |
30533 |
15398 |
7942 |
4430 |
3697 |
|
16 |
244346 |
121374 |
61116 |
30660 |
15409 |
8124 |
6541 |
In the end, we can clearly see the advantage of AWS Graviton2 on CPU-intensive jobs. Overall, when the AWS Lambda is running on arm64 we’d get a considerable amount of cost-saving with the right configuration.
Outro
In conclusion, we can see the benefit of AWS Graviton2 for both price and performance. Of course, in real life, we would have much more complicated scenarios and things to worry about. However, it should give us a rough idea of how much improvement we can expect. Getting started with AWS Graviton is pretty easy if you just follow this GitHub repository. To be able to debug your applications quickly, go to https://start.thundra.io/ and sign up for a free account.
SHARE
Introduction
AWS Lambda is constantly growing and, in time, things start to push limits. AWS has been working on new hardware architectures to make price-performance more efficient. AWS Graviton2 is a custom silicon-based 64-Bit ARM processor that’s coming to AWS Lambda which will help Serverless customers to have the same leverage as other services that can be built on AWS Graviton2. Lambda functions powered by Graviton2 offer up to 34% better price performance over x86-based Lambda functions.
As Thundra, we didn’t want to miss the opportunity to become a launch partner for such improvement and decided to test it out ourselves. The purpose of this article is to run simple benchmarks on AWS Graviton2 in comparison to x86_64 based architecture.
If you want to learn more about the AWS Graviton2 launch for AWS Lambda, make sure to visit the announcement page.
Benchmarks
Our test scenario consists of the below Python script that calculates the nth Fibonacci number to exhaust the CPU. It’s a bit cliche, but it’s simple. To keep the test simple and consistent, we’ve decided to calculate the 35th number as the sweet spot of our testing.
```python
import os
import multiprocessing
def handler(event, context):
nth = 35
processes = list()
for i in range(int(os.getenv('PROCESS_COUNT', 1))):
print(f'starting process #{i}.')
process = multiprocessing.Process(
target=calculateFib, args=(nth,))
processes.append(process)
process.start()
for i, process in enumerate(processes):
print(f'joining process #{i}.')
process.join()
print(f'finished process #{i}.')
response = {
"statusCode": 200
}
return response
def calculateFib(nth):
if nth <= 2:
return 1
else:
return calculateFib(nth - 2) + calculateFib(nth - 1)
```
We’ve run this script for all the combinations of (Process Count, RAM, Architecture) populated from the following data.
- Process Count: 1, 2, 4, 8, 16
- RAM: 256, 512, 1024, 2048, 4096, 8192, 10240 MB
- Architecture: x86_64, arm64 (AWS Graviton2)
After running all the combinations, we’ve noticed a couple of trends. The most important one of these is probably the billed duration time. There is a significant difference between the two architectures. Assuming there is some error margin and that there are other things at play here, it’s still impressive with Graviton2.
As one might guess, duration time will decrease as the RAM gets bigger. For single-process invocation, you don’t get much gain from the CPUs after 2048 MB RAM. For 16 process invocations, the margin gets bigger as you can see in the graphs below.
The raw data for billed duration gives you the bigger picture.
|
x86_64 |
256 MB |
512 MB |
1024 MB |
2048 MB |
4096 MB |
8192 MB |
10240 MB |
|
1 |
18952 |
9571 |
4752 |
2709 |
2746 |
2828 |
2731 |
|
2 |
38428 |
19242 |
9687 |
4803 |
2742 |
2737 |
2790 |
|
4 |
76149 |
37965 |
19007 |
9498 |
4904 |
2777 |
2804 |
|
8 |
155968 |
76986 |
38184 |
18914 |
9578 |
4891 |
4359 |
|
16 |
305743 |
156496 |
76358 |
38699 |
19274 |
9703 |
7723 |
|
arm64 |
256 MB |
512 MB |
1024 MB |
2048 MB |
4096 MB |
8192 MB |
10240 MB |
|
1 |
15570 |
7830 |
4105 |
2481 |
2478 |
2513 |
2494 |
|
2 |
30856 |
15482 |
8259 |
4062 |
2482 |
2482 |
2572 |
|
4 |
61394 |
30588 |
15378 |
7899 |
4270 |
2491 |
2498 |
|
8 |
121505 |
61284 |
30533 |
15398 |
7942 |
4430 |
3697 |
|
16 |
244346 |
121374 |
61116 |
30660 |
15409 |
8124 |
6541 |
In the end, we can clearly see the advantage of AWS Graviton2 on CPU-intensive jobs. Overall, when the AWS Lambda is running on arm64 we’d get a considerable amount of cost-saving with the right configuration.
Outro
In conclusion, we can see the benefit of AWS Graviton2 for both price and performance. Of course, in real life, we would have much more complicated scenarios and things to worry about. However, it should give us a rough idea of how much improvement we can expect. Getting started with AWS Graviton is pretty easy if you just follow this GitHub repository. To be able to debug your applications quickly, go to https://start.thundra.io/ and sign up for a free account.
Introduction
AWS Lambda is constantly growing and, in time, things start to push limits. AWS has been working on new hardware architectures to make price-performance more efficient. AWS Graviton2 is a custom silicon-based 64-Bit ARM processor that’s coming to AWS Lambda which will help Serverless customers to have the same leverage as other services that can be built on AWS Graviton2. Lambda functions powered by Graviton2 offer up to 34% better price performance over x86-based Lambda functions.
As Thundra, we didn’t want to miss the opportunity to become a launch partner for such improvement and decided to test it out ourselves. The purpose of this article is to run simple benchmarks on AWS Graviton2 in comparison to x86_64 based architecture.
If you want to learn more about the AWS Graviton2 launch for AWS Lambda, make sure to visit the announcement page.
Benchmarks
Our test scenario consists of the below Python script that calculates the nth Fibonacci number to exhaust the CPU. It’s a bit cliche, but it’s simple. To keep the test simple and consistent, we’ve decided to calculate the 35th number as the sweet spot of our testing.
```python
import os
import multiprocessing
def handler(event, context):
nth = 35
processes = list()
for i in range(int(os.getenv('PROCESS_COUNT', 1))):
print(f'starting process #{i}.')
process = multiprocessing.Process(
target=calculateFib, args=(nth,))
processes.append(process)
process.start()
for i, process in enumerate(processes):
print(f'joining process #{i}.')
process.join()
print(f'finished process #{i}.')
response = {
"statusCode": 200
}
return response
def calculateFib(nth):
if nth <= 2:
return 1
else:
return calculateFib(nth - 2) + calculateFib(nth - 1)
```
We’ve run this script for all the combinations of (Process Count, RAM, Architecture) populated from the following data.
- Process Count: 1, 2, 4, 8, 16
- RAM: 256, 512, 1024, 2048, 4096, 8192, 10240 MB
- Architecture: x86_64, arm64 (AWS Graviton2)
After running all the combinations, we’ve noticed a couple of trends. The most important one of these is probably the billed duration time. There is a significant difference between the two architectures. Assuming there is some error margin and that there are other things at play here, it’s still impressive with Graviton2.
As one might guess, duration time will decrease as the RAM gets bigger. For single-process invocation, you don’t get much gain from the CPUs after 2048 MB RAM. For 16 process invocations, the margin gets bigger as you can see in the graphs below.
The raw data for billed duration gives you the bigger picture.
|
x86_64 |
256 MB |
512 MB |
1024 MB |
2048 MB |
4096 MB |
8192 MB |
10240 MB |
|
1 |
18952 |
9571 |
4752 |
2709 |
2746 |
2828 |
2731 |
|
2 |
38428 |
19242 |
9687 |
4803 |
2742 |
2737 |
2790 |
|
4 |
76149 |
37965 |
19007 |
9498 |
4904 |
2777 |
2804 |
|
8 |
155968 |
76986 |
38184 |
18914 |
9578 |
4891 |
4359 |
|
16 |
305743 |
156496 |
76358 |
38699 |
19274 |
9703 |
7723 |
|
arm64 |
256 MB |
512 MB |
1024 MB |
2048 MB |
4096 MB |
8192 MB |
10240 MB |
|
1 |
15570 |
7830 |
4105 |
2481 |
2478 |
2513 |
2494 |
|
2 |
30856 |
15482 |
8259 |
4062 |
2482 |
2482 |
2572 |
|
4 |
61394 |
30588 |
15378 |
7899 |
4270 |
2491 |
2498 |
|
8 |
121505 |
61284 |
30533 |
15398 |
7942 |
4430 |
3697 |
|
16 |
244346 |
121374 |
61116 |
30660 |
15409 |
8124 |
6541 |
In the end, we can clearly see the advantage of AWS Graviton2 on CPU-intensive jobs. Overall, when the AWS Lambda is running on arm64 we’d get a considerable amount of cost-saving with the right configuration.
Outro
In conclusion, we can see the benefit of AWS Graviton2 for both price and performance. Of course, in real life, we would have much more complicated scenarios and things to worry about. However, it should give us a rough idea of how much improvement we can expect. Getting started with AWS Graviton is pretty easy if you just follow this GitHub repository. To be able to debug your applications quickly, go to https://start.thundra.io/ and sign up for a free account.
*** This is a Security Bloggers Network syndicated blog from Thundra blog authored by Oguzhan Ozdemir. Read the original post at: https://blog.thundra.io/a-performance-perspective-for-graviton-based-lambda-functions
