01
At the end of 2022, Raspberry Pi CEO Eben Upton stated that we would not see the Raspberry Pi 5 in 2023, and it would take a year to improve the Raspberry Pi inventory after the global chip shortage. However, he seems to want to surprise the many loyal Raspberry Pi fans.
The official release of this board has now been announced, and it will be available for purchase on October 23 (it may be delayed in China).
The Raspberry Pi 4 is already considered a powerful single-board computer, and the Raspberry Pi 5 claims its processing power is two to three times that of the Raspberry Pi 4. The Raspberry Pi 5 initially comes in 4GB and 8GB RAM versions (with 1GB and 2GB models to be released later), and its dimensions are basically the same as the 4B, but it adds some long-awaited features such as a built-in real-time clock, PCIe 2.0, and a power button, etc.
More importantly, the Pi 5 features a new quad-core 2.4 GHz Cortex-A76 Arm CPU (the old model initially had a Cortex-A72 with a frequency of 1.5 GHz, but firmware updates increased it to 1.8 GHz), and the new southbridge promises to improve USB 3 throughput and is equipped with a new VideoCore VII GPU (the VideoCore VI on the Raspberry Pi 4 has a frequency of 500 MHz, while the VideoCore VII has a frequency of 800 MHz).
There are many small improvements throughout the board, including a built-in fan connector with mounting holes, dual camera connectors, and a MicroSD card reader compatible with faster cards.
The retail prices for the 4GB and 8GB models are $60 and $90, respectively, only $5 more than the Raspberry Pi 4 models with the same memory capacity, indicating a very high cost-performance ratio. But how well does the Raspberry Pi 5 perform, and how does it work with the existing HAT (Hardware Attached on Top) and accessory ecosystem? For those of us who do not have the board yet, we found an overseas maker’s practical blog to take a look together.
02
The Raspberry Pi 5 looks very similar to the Raspberry Pi 4 and 3B+. At first glance, it appears to be another classic Raspberry Pi board design, reminiscent of the B+ redesign in 2014, but look closely! The “extremely advanced” 3.5mm audio/video jack has disappeared, so this generation of Raspberry Pi does not have an analog video/audio interface. That said, we haven’t used that jack since the Raspberry Pi 3 = =
The Raspberry Pi 5 camera and display connectors are smaller, using a 15-pin connector from the Raspberry Pi Zero instead of the original 22-pin connector. Note that there are now two adjacent connectors, allowing us to connect two cameras, two DSI displays, or a mixed connection.
The third connector is for PCIe devices. This is a PCIe 2.0 x1 interface for connecting fast peripherals, meaning NVMe solid-state drives can be used. We asked Upton about this interface, and he confirmed that it can accept NVMe SSDs of all sizes, but it requires a specially designed M.2 HAT, which is not ready at the time of this review release.
The Raspberry Pi 5 retains the dual micro HDMI ports introduced with the Raspberry Pi 4. Each port can support 4K 60Hz output, but to be honest, we have never used a Raspberry Pi with dual displays. Between the two micro HDMI ports is a single UART (Universal Asynchronous Receiver-Transmitter) connector, which can be used for debugging the previously released Raspberry Pi Pico Debug kit or for establishing a UART connection with other microcontrollers.
Sharp-eyed students will notice that the positions of the USB and Ethernet ports have swapped. The Raspberry Pi 5 seems to have drawn inspiration from earlier boards. This means that the Raspberry Pi 5 will require a new case. Aside from the changes in the interfaces, there are some subtle differences that make the cases designed for the old Raspberry Pi B model incompatible.
Interestingly, the Raspberry Pi 5 product profile states, “If you use a case, do not completely cover the case.” We will discuss the reasons for this later in the review.
Let’s first take a look at the two new features of the Raspberry Pi 5.
First, we have a battery backup for the real-time clock. The Raspberry Pi 5 can now keep accurate time without an NTP server or an expansion board that occupies GPIO space.
The second new feature is a power button! Some may dismiss this feature, but it is a frequently requested feature, and some makers have even invented their own products or purchased third-party products. When the Raspberry Pi is powered on, pressing the button will bring up the shutdown/logout menu. Pressing it again will trigger a safe shutdown. This shutdown is more like a sleep mode, with the Raspberry Pi consuming 1.4 watts. Pressing the power button will start the Raspberry Pi 5. You can also program the operating system to set the button to perform other functions, as it is a momentary button rather than a hard switch that cuts power.
The only thing that hasn’t changed is the 40-pin GPIO interface. Introduced on the Raspberry Pi B+ in 2014, the 40-pin GPIO also introduced the HAT (Hardware Attached on Top) standard. HAT is a standard similar to Arduino Shields, providing standard design templates and electronic standardization for a growing selection of accessories. The GPIO of the Raspberry Pi 5 is basically the same as previous models, but some changes have been made along the way, which will be detailed later.
▲ Comparison of Raspberry Pi 5 and Raspberry Pi 4
Raspberry Pi 5 Heat Generation and Power Consumption Performance
The Raspberry Pi 5 is the hottest of all the Raspberry Pi boards we’ve used.
When idle, without any additional cooling devices, its temperature hovers around 50.5 degrees Celsius, with a power consumption of about 2.7 watts.
During stress testing, the Raspberry Pi 5 quickly hits thermal limits (triggered at 82°C), throttling the CPU speed to reduce heat. During stress testing, our temperature reached 86.7°C (7 watts) and the CPU dropped from 2.4 GHz to 1.5 GHz. Interestingly, we observed the CPU speed fluctuating several times (2311 MHz, 2256 MHz, 2201 MHz, and 2146 MHz), possibly to balance speed and temperature.
You might be wondering, “How does the Raspberry Pi 4 perform?”
In tests, the idle operating temperature of the Raspberry Pi 4 is 45.7°C, with a power consumption of 1.02 watts. During stress testing, the temperature rose to 79.8°C, while the power consumption reached 6.2 watts.
This means that the Raspberry Pi 5 consumes about 1 watt more power than the Raspberry Pi 4 while gaining more powerful computing capabilities. The Raspberry Pi 4 has a lower stress testing temperature without a cooling system, while the Raspberry Pi 5’s temperature can be reduced to 59.3°C with a heat sink installed, maintaining good processing performance.
The Raspberry Pi 5 is also designed for “overheating,” which means we need to install heat sinks to achieve optimal performance. We tested the official Raspberry Pi active cooler, and it performed well. This cooler consists of aluminum heat sinks (with a nice anodized Raspberry Pi logo) and a fan.
It can cool the Arm-based SoC, RAM, and the new RP1 chip (which we will discuss further later) in a way that differs from the connection method of the threaded hole. It has its own mounting holes. One is next to the USB C port, and the other is between the GPIO and USB ports. The cooler uses push-in plastic/nylon pins to secure it. You can carefully use pliers or a plastic pen tube to squeeze and push the plastic pins to remove the cooler.
This official cooler connects to a new fan connector located next to the USB port, which is a better improvement than previous Raspberry Pi models where installing a fan would occupy GPIO pins, sometimes hindering the use of HATs. When the CPU temperature reaches 50°C, the fan will start working. When idle, the cooler keeps the Raspberry Pi 5’s temperature at 39.5°C, with a power consumption of 2.6 watts. During stress testing, the temperature rises to 59.3°C (with a power consumption of 6.8 watts), far below the thermal limit point. The power consumption of the Raspberry Pi 5 with a cooling system is slightly lower than that without it.
The Raspberry Pi 5 introduces a new feature: a power button. I know this isn’t big news for many. We had power buttons on home computers in the 70s, but the Raspberry Pi has never had this feature until now.
This button is a “soft” power button that calls a script to select logout/shutdown/restart, and pressing it again will soft shut down the Raspberry Pi into sleep mode. In sleep mode, the power consumption of the Raspberry Pi 5 is 1.3 watts, about half of its idle power consumption. This allows us to save more energy. But we can do even better. The Raspberry Pi engineering team has provided instructions on how to reduce sleep power consumption, and we saw amazing results!
Now our sleep power consumption can be reduced to 0.05 watts, much lower than before. In sleep mode, the 5V (GPIO) pins remain high, but the 3.3V pins are at low. This might cause issues for your favorite HATs, as they may not shut down with the Raspberry Pi. The revised HAT specification, HAT+, will address this issue, and Raspberry Pi will provide relevant instructions.
Can we use existing heat sinks and fans to cool the Raspberry Pi 5? The answer is yes, but due to the board layout changes, not every heat sink will fit. For example, the Akasa Gem Pro and 52Pi’s Ice Tower heat sinks cannot be used directly.
For active cooling, we tried Pimoroni’s Fan Shim. This is a board that slides onto the GPIO, still providing interfaces and won’t interfere with other operations. When idle, the Raspberry Pi 5’s temperature stays around 29.6°C, very quiet, although not as quiet as the new active cooler, but almost silent. The Fan Shim provides impressive cooling performance of 60.4°C (compared to the official Raspberry Pi active cooler’s 59.3°C), and if you already have a Fan Shim, it is a viable alternative to the official active cooler. If not, consider purchasing the official active cooler, but we still don’t know its price.
For passive cooling, we bought a set of ordinary self-adhesive heat sinks from Amazon. We stuck them on the SoC, PMIC, and Wi-Fi chips. The idle temperature of the Raspberry Pi 5 is 41.1°C, while the temperature without them is 50.5°C.
When we conducted stress tests with the heat sinks, the temperature rose to 85.6°C, only 1.2°C lower than the test without cooling. So, buying cheap heat sinks is a waste of money. We recommend either buying active cooling or waiting for a suitable custom passive cooling solution to hit the market.
Can the Raspberry Pi 5 be overclocked?
Yes. But how fast it can go depends on your luck (known as Silicon lottery).
Overclocking is not difficult; it only requires some adjustments to the configuration files. In our tests, we successfully overclocked the CPU to 3 GHz, and we also succeeded in raising the frequency to 3.2 GHz, but the speed reported by the system differed in neofetch and vcgencmd. Neofetch reported 3.2 GHz, while vcgencmd reported 3 GHz. After communicating with the Raspberry Pi engineering team, we believe the 3.2 GHz speed is incorrect, so we omitted this data in the review.
No matter how much you overclock, you need good cooling. Passive cooling with small heat sinks is far from sufficient, as you need to actively cool it to keep the temperature below the thermal limit of 80°C. When running at a frequency of 3 GHz, the idle temperature of the Raspberry Pi 5 is 46.6°C, with a power consumption of 3 watts. Under stress testing, the temperature of the Raspberry Pi 5 reached 69.2°C, with a power consumption of 10 watts.
Raspberry Pi 5’s 64-bit Operating System
Our review sample came with a microSD card running the latest Raspberry Pi operating system, but this time it was a card running a 64-bit operating system, with a kernel version of 6.1.0. The 64-bit Raspberry Pi operating system release has long been behind the popular 32-bit version. This is mainly because many older Raspberry Pi boards only support 32-bit operating systems. However, support for 64-bit operating systems has been available since the Raspberry Pi 3.
The new operating system is based on Debian 12 code-named “Bookworm,” released in July 2023. Bookworm brings many changes, one of the biggest being the new version of Python. In previous versions, we saw Python 3.9 installed by default, but in Bookworm, we see an upgrade to Python 3.11, which also changes how Python modules are installed.
In the past, we generally installed Python modules system-wide or for individual users. While this method is very convenient for many users, it also introduces a risk: it may corrupt or conflict with your Python installation in the base operating system. This is mainly because the Python package manager coexists with Python modules installed via pip, which may cause conflicts.
To solve or at least mitigate this conflict and impact, Python introduced PEP 668 in version 3.11. PEP (Python Enhancement Proposal) is a proposal for improving the Python language and environment. PEP 668 aims to help manage and resolve this issue.
The goal of PEP 668 is to ensure that when we install Python extensions, we do not affect the original Python environment of the operating system. Therefore, PEP 668 recommends that when installing Python packages, we do not install them directly in the operating system but create a “little island” (virtual environment venv) to install and use Python packages. The benefit of this approach is that this “little island” is isolated from the operating system, so it won’t cause any issues. This may make it a bit complex for beginners to install Python packages, but it has a more significant impact on companies providing third-party plugins and hardware expansion boards.
Raspberry Pi 5’s Processing Performance
With faster boot times, improved microSD card performance, and a 64-bit operating system, we have a system that focuses more on speed. Raspberry Pi claims that the performance of the Raspberry Pi 5 is 2 to 3 times that of the Raspberry Pi 4, and overall, it indeed feels that way.
Application launch speeds have improved significantly. On the supplied microSD card, the Raspberry Pi 5 opens Gimp in 5.5 seconds, while the Raspberry Pi 4 takes 10.8 seconds. The Raspberry Pi 5 opens Firefox in 5.1 seconds, while the Raspberry Pi 4 takes 8.6 seconds. Boot time is about 18 seconds, while the Raspberry Pi 4 takes 38 seconds (using the same microSD card and operating system image).
In synthetic benchmark tests, the differences between generations are clear. In the single-threaded CPU test of Sysbench, the Raspberry Pi 5 performs 2,729 events per second, while the Raspberry Pi 4 performs 1,766 (the more events, the better). When we increase to four threads, the Raspberry Pi 5 wins again, with 10,912 events, while the Raspberry Pi 4 has 7,068, an improvement of 54%.
When we run the 7-Zip compression benchmark, the Raspberry Pi 5 achieves a MIPS of 9,543 for compression, while the old Raspberry Pi’s MIPS is only 4,287, an improvement of 122%. For decompression, the Raspberry Pi 5 reaches 13,231 MIPS, while the Raspberry Pi 4’s MIPS is 7,568.
We have not yet conducted a complete AI test suite, but we successfully ran the TensorFlow Lite benchmark test from the Phoronix Benchmark Suite using the SqueezeNet neural network. In this test, the fewer microseconds it takes to complete the task, the better. The Raspberry Pi 5 takes only 25,276 microseconds, while the Raspberry Pi 4 takes 80,327 microseconds, a difference of 68%.
With faster boot times, more efficient microSD performance, and a 64-bit operating system, the Raspberry Pi 5 is faster. According to Raspberry Pi’s official claim, the performance of the Raspberry Pi 5 is generally 2 to 3 times that of the Raspberry Pi 4, and our actual experience confirms that it is indeed fast.
First, opening applications is noticeably faster. On the same microSD card, the Raspberry Pi 5 opens Gimp in just 5.5 seconds, while the Raspberry Pi 4 takes 10.8 seconds. Starting Firefox, the Raspberry Pi 5 takes only 5.1 seconds, while the Raspberry Pi 4 takes 8.6 seconds. Boot time is around 18 seconds, compared to 38 seconds on the Raspberry Pi 4 with the same operating system image and SD card.
In synthetic benchmark tests, the differences between generations are very clear. In the Sysbench CPU test (single-threaded mode), the Raspberry Pi 5 can generate 2,729 events per second, while the Raspberry Pi 4 can only generate 1,766 (more events mean better performance). When we raised the test to four threads, the Raspberry Pi 5 again won with 10,912 compared to 7,068, an improvement of 54%.
When we run the 7-Zip compression benchmark, the Raspberry Pi 5 provides a compression speed of 9,543 MIPS, while the older model’s compression speed is 4,287 MIPS, a performance improvement of 122%. The decompression speed of the Raspberry Pi 5 is 13,231 MIPS, while the Raspberry Pi 4’s is 7,568 MIPS.
Although we have not conducted a full suite of AI tests, we have run the TensorFlow Lite benchmark test using the SqueezeNet neural network from the Phoronix Benchmark Suite. In this test, the lower the score (i.e., the fewer microseconds it takes for the computer to complete the task), the better. The Raspberry Pi 5 takes only 25,276 microseconds, while the Raspberry Pi 4 takes 80,327 microseconds, a difference of 68%.
Video Playback and Streaming
Like the Raspberry Pi 4, the Raspberry Pi 5 can output to two displays at a maximum resolution of 4K via its dual micro HDMI ports. For the Raspberry Pi 5, its enhanced GPU can provide a refresh rate of 60Hz on each screen, and if the display supports it, it can even use HDR.
We haven’t been able to get it to run HDR or verify that it can indeed provide a 60Hz refresh rate, but we have every reason to believe it can drive a second display (just like previous models) as it can easily output to a 4K display. The bigger question is how this board handles challenges like streaming high-resolution videos from YouTube.
Streaming video has long been a “weakness” of the Raspberry Pi, and the faster GPU and CPU are expected to improve this situation. However, in our tests, the performance of YouTube still has room for improvement. We played the YouTube video “Tears of Steel” at 1080p resolution (which actually has about 24fps pixels), and the overall effect was very smooth with very few dropped frames. When we switched to a 1080p, 60fps natural video, the picture remained very smooth, although YouTube’s “Nerds stats” showed some dropped frames.
When we kept the screen output at 4K but played a 1080p video from YouTube, both videos were very choppy and ran very slowly. The “Nerds stats” reported that both videos dropped about two-thirds of their frames. This issue occurred whether the video was played in full screen or even just in a portion of the browser window. Even resizing the video player made it feel very slow.
Perhaps future software updates or configuration adjustments will improve YouTube’s streaming playback. However, when the screen resolution is at 1080p, video playback at 1080p is also quite smooth, which is still a significant improvement over previous Raspberry Pi versions.
USB and MicroSD Card Performance, RP1 Chip
The Raspberry Pi 5 uses a new “Pi Silicon” chip, the RP1. It looks very similar to the RP2040, marking Raspberry Pi’s first foray into custom chips. The RP1 is primarily responsible for most of the I/O processing of the Raspberry Pi 5.
According to the product profile we received, the USB bandwidth provided by the RP1 is more than double that of previous models, allowing for faster transfer speeds when using USB drives with UAS (USB Attached SCSI). The RP1 also provides a dedicated four-channel 1.5 Gbps MIPI camera and display interface, increasing the total bandwidth for cameras and displays threefold. However, it is important to note that the theoretical limit of the USB 3.0 ports remains the same as the Raspberry Pi 4, at 5 Gbps, so we expect higher processing capabilities to drive higher throughput.
To find out how fast the Raspberry Pi 5’s USB 3.0 connection is, we conducted storage performance tests on the built-in MicroSD card reader and a PCIe 3.0 solid-state drive connected via USB. Using Sysbench’s file I/O test, the Raspberry Pi 5 read from a Kingston Canvas Go Plus MicroSD card at a speed of 12.75MB/s and wrote at a speed of 8.5MB/s. Meanwhile, the solid-state drive had a read speed of 31.33MB/s and a write speed of 20.89MB/s.
So how does this compare to the Raspberry Pi 4? The Kingston Canvas Go Plus had a read speed of 8.78 MB/s and a write speed of 5.85 MB/s. The solid-state drive’s read speed was 12.96 MB/s and write speed was 8.64 MB/s. Therefore, both the USB 3.0 and MicroSD card reader interfaces have increased by more than double.
By the way, the Raspberry Pi 5’s MicroSD card reader now supports faster MicroSD cards using the SDR104 standard. SDR104 is a subset of the popular UHS-I card standard, theoretically capable of speeds up to 104 MB/s. Although very few cards indicate they support SDR104, you can find UHS-I cards that claim transfer speeds exceeding 100 MB/s. The theoretical maximum speed of the Raspberry Pi 4’s card reader is about 50 MB/s, but in practice, we have never seen cards exceeding 40 MB/s.
We tried several different MicroSD cards for the Raspberry Pi 5 and Raspberry Pi 4. Using the storage performance benchmark tool IOZone, we found that the Kingston Canvas Go Plus achieved sequential read and write speeds of 86 to 55 MB/s. The same card on the Raspberry Pi 4 had sequential write speeds of 37 to 41 MB/s.
The Raspberry Pi 4 also has an M.2 connector that allows direct connection to SSDs. This is a significant improvement, and we want to test NVMe solid-state drives using the required M.2 HAT connection. It should be faster than these already high numbers.
Using GPIO
The GPIO of the Raspberry Pi 5 is one of its biggest highlights. These 40 GPIO pins provide endless possibilities for our favorite programming languages in electronic projects. The Raspberry Pi has a deep connection with Python, which is also the preferred programming language for many projects, but we can also write GPIO code in Lua, Go, C, JavaScript, BASIC, or other languages.
We usually use the Python modules RPi.GPIO and GPIO Zero to interact with GPIO for testing. We executed GPIO Zero tests without encountering any issues, which is good news for beginners looking to get started with electronics and Raspberry Pi. Although GPIO Zero runs well, RPi.GPIO has some issues due to some underlying configuration problems.
The RPi.GPIO module was created by Ben Croston in the early days of Raspberry Pi and quickly became the standard for many Raspberry Pi projects and hardware. RPi.GPIO is likely the behind-the-scenes tool for many of your favorite Raspberry Pi HAT software modules, and on the Raspberry Pi 5, you may encounter some problems. In fact, we couldn’t properly test any third-party HATs we usually use. The official Raspberry Pi Sense HAT test was successful, possibly because it uses libgpiod instead of RPi.GPIO.
The Raspberry Pi CTO Gordon Hollingworth made a statement about PEP 668 and the compatibility of Raspberry Pi 5 HATs.
The Raspberry Pi operating system will follow the practices of systems like Debian and Ubuntu, adopting PEP 668, and encouraging users to be aware of the potential issues they may encounter when installing, updating, and deleting packages. We will provide more documentation to help users understand this change and guide them to use some tools, such as virtualenvwrapper, to make this process easier.
Any HAT that communicates using standard Linux interfaces can be used without changing software. However, many HATs’ software relies on non-portable interfaces, such as RPi.GPIO, and these software fail every time we release new hardware. During the time leading up to the Raspberry Pi 5’s release, we will work closely with manufacturers to update their software in a timely manner to accommodate the Raspberry Pi 5, which is a clear advantage of separating our product release from the market launch!
As Hollingworth mentioned, the weeks between the release and the retail launch of the Raspberry Pi 5 will give manufacturers enough time to prepare many of the best Raspberry Pi HATs and accessories to accommodate the Raspberry Pi 5. Once officially released, we will also retest some HATs.
Additionally, once the active cooler is installed, reading GPIO pins during wiring can be a bit tricky, but not impossible. Just make sure to keep wires away from the rotating fan. If you plan to use HATs or other accessories with GPIO, it is recommended to purchase a pair of 2×20 male-female header extenders. This way, your plugs will be above the cooler, keeping the ventilation clear. M2.5 support pillars and screws can be used to help stabilize the board.
Dual Camera Support
The Raspberry Pi 5 introduces support for multiple cameras, a new feature on mainstream Raspberry Pi boards. If you have used the Compute Module, you may already be familiar with multi-camera support, as it has been integrated into the Compute Module IO board since day one, but most Raspberry Pi fans may not have owned a Compute Module.
In terms of coding, both the libcamera and Picamera2 Python modules support multiple cameras, and we successfully tested by passing camera parameters (0 or 1) to libcamera and using the correct camera constructors (also 0 or 1) in Picamera.
The camera/display uses a 15-pin connector previously used for Raspberry Pi Zero series boards, the latest of which was used on the Raspberry Pi Zero 2 W. The camera module and older Raspberry Pis used a 22-pin connector, requiring a cable or adapter change. We successfully conducted tests with both on the Raspberry Pi 5.
Support for Power Over Ethernet
The Raspberry Pi 5 supports Power Over Ethernet (PoE) but requires a new PoE HAT to use it. The PoE header was initially placed between the GPIO and Ethernet ports on the Raspberry Pi 3B+ and Raspberry Pi 4, but this position has been used to connect the fan on the Raspberry Pi 5, and the PoE header has now been moved between the camera/display and Ethernet port. So, we cannot simply make the connection with some jumpers. A replacement PoE HAT is expected to be released soon.
Raspberry Pi 5’s Emulation
As I write this content, support for emulators on the Raspberry Pi 5 is still catching up. Undoubtedly, RetroPie, Lakka, Recalbox, etc., will update their products accordingly. Once compatible images are available, we will conduct tests.
We hope the Raspberry Pi 5 can improve emulation for games after the PS2 era. Boards like Khadas’s VIM4 and Edge 2 Pro have excellent emulation performance for PS2 and PSP era games. Emulation of Gamecube/WiiU would be a great option and could make the Raspberry Pi 5 a low-cost choice for retro gaming emulation.
If your tastes lean towards earlier games, we are 100% confident that the Raspberry Pi 5 has enough processing power for 8-bit, 16-bit, and even many 32-bit game consoles. It might even improve the performance of some late arcade games (after the late 1990s) that have custom chips that need to be emulated.
Raspberry Pi Comparison
Compared to the Raspberry Pi 4, the price of the Raspberry Pi 5 has increased by $5. Is it worth it? In terms of computing power, yes. While it may not reach the heights of the Khadas VIM4, Edge2 Pro, or LattePanda Sigma, we also didn’t spend that much money.
The Raspberry Pi 5 is transitioning towards a low-cost, low-power Linux desktop computer while retaining GPIO functionality. In the past, the Raspberry Pi was often seen as a GPIO with Linux. However, now aligning the Raspberry Pi OS with Debian and Ubuntu standards, it feels more like a Linux computer. The impact of PEP688 and its effect on HATs is an example of this change. In the short term, Raspberry Pi accessory manufacturers will face difficulties in adjusting their products, but this adjustment will eventually happen. Some old accessories may not be updated, in which case the fallback is to use the Raspberry Pi 4 or earlier models.
Conclusion
In many ways, the Raspberry Pi 5 is an impeccable product. If you are a Pi fan, you surely can’t wait to buy it as soon as it is available, and considering the price of $60 or $80, you should be able to afford it (just budget a few more dollars for cooling).
There is much to love about the Raspberry Pi 5: overall stronger performance, smoother video playback, greater storage bandwidth, which is one of the best reasons to choose the Raspberry Pi 5 over the 4. Many will find the RTC (real-time clock) or power button very useful as well.
If you don’t need this performance, the Raspberry Pi 4 is still a reliable choice; this older model can accomplish more tasks without an active cooling system and has solid support even after four years on the market. If you don’t need Linux, but just need GPIO functionality, the $8 Raspberry Pi Pico W is also a very good choice. However, if you want the best single-board computer currently available from Raspberry Pi, the Raspberry Pi 5 will be your top choice.
Source: https://www.tomshardware.com/reviews/raspberry-pi-5
Author: Avram Piltch
Translated text first published in: DF Maker Community
Reprint must indicate source information
RTX4090都不让用了?美国这是玩不起?
公司里的硬件技术总监CTO,天天都在干些啥?
太真实了!研发人员都是怎么被搞没的?
回武汉后,工资降了一半。