This report provides a comprehensive comparison between Apple's latest M4 chip and its predecessor, the M3 chip, with specific focus on their integration in the iPad Pro 2024 models.
The M4 chip is Apple’s latest system on a chip (SoC) that builds on the power-efficient performance of its predecessors. It incorporates significant improvements in CPU, GPU, Neural Engine, and memory system. The chip features 28 billion transistors constructed using advanced second-generation 3-nanometer technology, offering enhanced efficiency and output. The new display engine in the M4 supports precision, color accuracy, and brightness uniformity, making it suitable for the Ultra Retina XDR display.
As of now, the M4 chip is available exclusively in the iPad Pro 2024 models. The integration of this chip allows the iPad Pro to achieve groundbreaking performance figures, making it the most powerful device of its kind. The chip configuration includes up to a 10-core GPU, 16-core Neural Engine, and the capacity to perform 38 trillion operations per second. The CPU configurations vary based on storage options; models with 1TB and 2TB storage feature four high-performance cores, while the 256GB and 512GB versions come with three high-performance cores.
The M4 chip is expected to be incorporated into other Apple devices in the near future. It is anticipated to be introduced in the 14-inch MacBook Pro, both sizes of the MacBook Air, the iMac, and the Mac mini by late 2024. The MacBook Air is projected to gain access to the M4 chip by spring 2025. Unlike the M4, its predecessor M3 has been integrated into various products, including the 14-inch MacBook Pro, 13-inch and 15-inch MacBook Air, and the 24-inch iMac, while never making its way to any iPad model.
Both the M3 and M4 chips are manufactured using a 3nm process. This is a notable improvement over the M2 chip, which used a 5nm process. The shift to a 3nm process enables better performance and efficiency. The M4, however, is a second-generation 3nm chip, offering small improvements in efficiency and output over the M3.
The M3 chip features an 8-core CPU with four performance cores and four efficiency cores, complemented by either an 8-core or a 10-core GPU. The M4 chip upgrades this with a 9-core or 10-core CPU (with three or four performance cores and six efficiency cores) and a 10-core GPU. This evolutionary step results in performance enhancements, though the increase is relatively minor.
The M4 chip includes 28 billion transistors compared to the M3's 25 billion. Additionally, the M4 has a memory bandwidth of 120Gbps, an enhancement from the M3's 100Gbps. These improvements contribute to the incremental performance gains observed in the M4.
The M4’s Neural Engine performs 38 trillion operations per second (TOPS), compared to the M3’s 18 TOPS and the M2’s 15.8 TOPS. However, this performance figure for the M4 is at INT8 precision, while the M3’s was at FP16. When normalized to the same precision, the M4 offers a modest 5% improvement over the M3. The Neural Engine in the M4 is essential for AI and machine learning workloads and can perform at speeds unmatched by any neural processing unit in any AI PC to date.
Unlike previous chips, the M4 includes a dedicated display engine for managing the OLED display in the new iPad Pro. This engine allows for stunning precision, color accuracy, and brightness uniformity, making the M4 particularly well-suited for high-performance display tasks.
The M4 chip, tested on benchmarking website Geekbench, achieved a score of 9,234 in the ML 0.6.0 test. This score, which measures machine learning capabilities, is about 23% higher than the M2 and approximately 10% better than the M3. These benchmarks indicate an incremental performance improvement rather than a significant leap. It is important to note that Geekbench tests are synthetic benchmarks rather than real-world tests, so these results may not directly translate to actual usage scenarios.
While synthetic benchmarks provide some insight, predicting real-world performance of the M4 compared to the M3 is challenging without comprehensive reviews. Notably, the iPad Pro with the M4 chip lacks active cooling, which is present in devices like the MacBook Pro, potentially affecting performance under heavy loads. Moreover, the M4 chip's Dynamic Caching, mesh shading, and hardware-accelerated ray tracing capabilities, which were also present in the M3 series, are new to the iPad Pro line but do not contribute new advantages over the M3.
The inclusion of Dynamic Caching, mesh shading, and hardware-accelerated ray tracing in the M4 is carried over from the M3 series. These features enhance graphics performance, particularly for gaming and graphic-intensive applications. Previously absent in the M2-equipped iPad Pro, these features now make the M4-equipped iPad Pro more competitive. However, as these are not new features compared to the M3, they do not provide additional specific advantages in comparison to the M3 chip.
The M4 chip represents a moderate step up from the M3, featuring a 9-core or 10-core CPU (with additional efficiency cores), a 10-core GPU, and a 16-core Neural Engine capable of performing 38 TOPS. In contrast, the M3 chip has an 8-core CPU, up to a 10-core GPU, and a Neural Engine performing 18 TOPS. The M4 also contains 28 billion transistors, up from 25 billion in the M3, and a higher memory bandwidth of 120Gbps compared to the M3's 100Gbps. Additionally, the M4 introduces a dedicated display engine to accommodate the new OLED display in the 2024 iPad Pro. These enhancements result in improved efficiency and performance, although differences are subtler compared to the leap from M2 to M3.
The integration of Apple's M4 chip in the iPad Pro 2024 models reveals a distinct correlation between storage capacity and CPU performance. Notably, the iPad Pro models with higher storage (1TB and 2TB) come with a more robust 10-core CPU, while the models with lower storage (256GB and 512GB) have a 9-core CPU configuration. This ties the storage capacity directly to the computational power available to the user.
RAM configurations also vary based on storage capacity. The iPad Pro models with 256GB and 512GB storage are equipped with 8GB of RAM, while the models with 1TB and 2TB storage come with 16GB of RAM. This differentiation impacts the overall performance and multitasking capabilities of the devices.
A historical review shows that Apple has previously used similar strategies. For instance, the iPhone 13 series had different GPU core counts across its models based on storage and RAM configurations. This trend continues with the iPad Pro 2024, showcasing consistent practices in performance tiering within the same generation of devices.
Users are now compelled to consider both storage and RAM when purchasing the iPad Pro 2024 models, as these factors are intertwined with the performance potential of the devices. The premium tier models with 16GB RAM and 10-core CPUs are significantly more expensive, which may affect purchasing decisions and user satisfaction, especially for those looking for high performance without needing extensive storage.
The new iPad Pro models come in different pricing tiers, heavily influenced by storage and RAM configurations. According to the data, the entry-level 256GB and 512GB models come with a 9-core CPU and 8GB of RAM, while the higher-end 1TB and 2TB versions feature a 10-core CPU and 16GB of RAM. This tiered pricing strategy means that to access the most powerful version of the M4 chip, customers must invest in the higher storage models. Additionally, the price for the 11-inch iPad Pro with 1TB storage starts at $1,599, whereas the 13-inch model starts at $1,899.
The pricing strategy of the iPad Pro is notably reminiscent of Apple's approach with its MacBook series. Customers are required to pay more for higher storage and RAM configurations to get a more powerful CPU. This mirrors the practice found in the MacBook Air and MacBook Pro models, where enhanced performance is tied to costlier configurations. For instance, the higher-end variants of the MacBook Air with the M3 chip come only with higher RAM and storage, similar to the iPad Pro's strategy with the M4 chip.
The introduction of the M4 chip in the iPad Pro positions it as a more powerful competitor in the tablet market. The enhancements in CPU and GPU performance, along with the addition of a new display engine, make the iPad Pro a formidable option for creatives and professionals. However, the high costs associated with the most powerful configurations may limit its accessibility to a broader market. Additionally, the iPadOS, despite having apps like Final Cut Pro and Procreate, still lacks the comprehensive pro applications available on macOS, which could affect its reception as a laptop replacement.
Despite the advanced features and hardware of the M4 iPad Pro, user perception may be affected by its high price point and the limitations of iPadOS. The necessity to spend a significant amount to access the most powerful variant of the M4 chip and additional costs for accessories like the Magic Keyboard and cellular connectivity may deter some users. Moreover, some consumers may compare the iPad Pro's value proposition unfavorably against laptops like the MacBook Air, which offer better value for productivity applications at similar price points. The debate on whether the cost outweighs the value added by features such as the OLED display, hardware-accelerated ray tracing, and Neural Engine capabilities continues to persist.
The M3 chip, a predecessor to the M4, features an 8-core CPU with four performance and four efficiency cores, and either an 8-core or a 10-core GPU, along with a 16-core Neural Engine. It was notable for the shift from a 5nm to a 3nm manufacturing process, which enhanced performance and efficiency.
The M4 chip, used primarily in the iPad Pro 2024, offers a 9-core or 10-core CPU, a 10-core GPU, and a 16-core Neural Engine capable of 38 trillion operations per second. It features a second-generation 3nm manufacturing process and new display engine among other enhancements.
The Neural Engine in Apple's M-series chips is designed to accelerate AI and machine learning tasks. The M4's Neural Engine can perform 38 trillion operations per second, significantly higher than earlier generations, enhancing capabilities in AI-driven applications.
A semiconductor fabrication process that allows for greater transistor density and improved performance and efficiency. Both M3 and M4 chips use this process, with the M4 being a second-generation iteration providing additional enhancements.
The M4 chip represents a marginal but significant improvement over the M3, with advancements in CPU/GPU core counts, Neural Engine capabilities, and display technology. However, the differentiation based on storage capacity may affect user choices and market reception.