Raspberry Pi 3 Model B+ Performance vs Raspberry Pi 3 Model B

Updated 17 April 2018

The Raspberry Pi 3 Model B+ (I’ll be calling it the Pi 3+ for short) is an update to the existing Raspberry Pi 3 Model B and was released on Wednesday 14th March 2018. It has some very interesting changes:

  • The Broadcom BCM2837 has been clocked to 1.4 GHz, up from 1.2 GHz on the Pi 3, as well as getting an integrated heat spreader and improved power integrity system. This should allow for better heat dissipation and the possibility of a higher overclock
  • The built in WiFi now supports 802.11 b/g/n/ac in 2.4 and 5 GHz along with an updated antenna that should allow for better range and less packet losses than you can currently achieve (test results coming soon)
  • And rounding out, there is Gigabit Ethernet (limited to around 300 Mbit/s by the USB 2.0 controller) and a new Power over Ethernet header that will work with the incoming PoE HAT

The full list of features and specs can be found in the Official Product Release.

I've been testing both boards to see exactly what you can expect when you purchase the new Pi 3+, and how it stacks up against its predecessor. Let me know in the forum what you think as well as any other tests you'd like to see us run (more tests are coming soon!)


Overall setup

During all tests, both Raspberry Pi’s were in the open air without a case, in an air-conditioned room set at 24°C. The only devices plugged in were an Ethernet cable, HDMI monitor and a wireless keyboard and mouse combo. All tests were run 3 times and the averaged results are shown below. The Pi 3 was tested with NOOBS installed Raspbian 2017-11-29, but the Pi 3+ required an update to Raspbian 2018-03-14 before it would boot. All benchmarks were from either Roy Longbottom’s Raspberry Pi benchmarks or the official Raspberry Pi repository.

raspberry-pi-overall-performance

Actual test results shown in detail below

3b-vs-3b+

CPU performance (Result: Pi 3+ up to 19% FASTER)

SysBench uses a range of tests to ascertain the performance of a system as a whole, but we’ll be using just the CPU component in this test. The CPU benchmark gives an overall view of the CPU performance by measuring the time it takes to check for prime numbers up to 10,000, with more focused tests used below.

Dhrystone is a benchmark that uses a test program and reports how many times it can be completed per second.

The Linpack benchmarks for single and double precision performance are a standard test that has been around in various forms since 1979. The results show how many floating point operations can be performed per second.

raspberry-pi-cpu-performance

Memory performance (Result: Pi 3+ 20% FASTER)

Memory performance was tested with the SysBench memory benchmark, which shows the memory read and write speeds. Tests were conducted with a block size of 1 MB and a total size of 2 GB, using 1 or 4 threads.

raspberry-pi-memory-performance

Ethernet speed (Result: Pi 3+ 257% FASTER)

iPerf 3.1.3 was used to measure the speed between each Pi and a desktop PC over a Gigabit Ethernet connection (even though the Pi only supports 10/100) for a period of 60 seconds.

raspberry-pi-ethernet-performance

Power usage (Result: Pi 3+ up to 38% MORE)

A DP832 power supply was used to provide power to the Pi’s during almost all tests, at 5.1V like the Official Raspberry Pi power supply. The only exception to this was the 4 core SysBench test for the Pi 3+, which dropped the voltage to a point that required throttling, but switching to the Official Raspberry Pi power supply alleviated this issue. The load measurement was taken during the SysBench 4 core CPU test. All measurements show the maximum amperage the boards required.

raspberry-pi-power-usage

Temperatures (Result: Pi 3+ 4% LOWER)

Temperature readings were taken using the vcgencmd measure_temp command again during the SysBench 4 core CPU test.

raspberry-pi-temperatures


Conclusions

The changes made by the Raspberry Pi Foundation have clearly improved performance, especially with the ~2.5x Ethernet speed increase. This does however come at the cost of some extra power usage. It does all this while being able to maintain slightly lower temperatures as well.

As always, start a discussion below or in the forum to let us know if there’s a specific test you’d like to see us do with these boards, if you think we should compare something else in the Raspberry Pi Family, or just what projects you're excited about that these changes will let you accomplish.

Have a question? Ask the Author of this guide today!

Please enter minimum 20 characters

Your comment will be posted (automatically) on our Support Forum which is publicly accessible. Don't enter private information, such as your phone number.

Expect a quick reply during business hours, many of us check-in over the weekend as well.

Comments


Loading...
Feedback

Please continue if you would like to leave feedback for any of these topics:

  • Website features/issues
  • Content errors/improvements
  • Missing products/categories
  • Product assignments to categories
  • Search results relevance

For all other inquiries (orders status, stock levels, etc), please contact our support team for quick assistance.

Note: click continue and a draft email will be opened to edit. If you don't have an email client on your device, then send a message via the chat icon on the bottom left of our website.

Makers love reviews as much as you do, please follow this link to review the products you have purchased.