M5 Ultra Scores 360,019 in Geekbench
The Geekbench Browser result identifies the processor as the Apple M5 Ultra and records a Metal GPU score of 360,019.
The system used for the result features a 36-core Apple processor and 256GB of memory. The benchmark was run on macOS 27 using Geekbench 7.
Because this is an uploaded benchmark result, it provides a real-world data point for the M5 Ultra rather than an official Apple performance claim.
The result is particularly interesting because Apple's M5 Ultra is designed for demanding professional workloads, including graphics, video production, scientific computing and local artificial intelligence.
What Apple Says About the M5 Ultra
Apple introduced the M5 Ultra in August 2026 alongside the new Mac Studio.
According to Apple, the M5 Ultra is built using a quad-die architecture and features up to a 36-core CPU and an 80-core GPU.
The chip also supports up to 512GB of unified memory and delivers 1.2TB/s of memory bandwidth.
Apple says the new architecture is designed to handle demanding professional and AI workloads while allowing users to keep large datasets and AI models in local memory.
Apple Claims Up to 40% Faster Graphics
Apple's own testing claims that the M5 Ultra delivers up to 40% faster graphics performance than M3 Ultra.
The company attributes the improvement to its next-generation GPU architecture, including updated shader technology, second-generation Dynamic Caching, hardware-accelerated mesh shading and third-generation ray tracing.
These are Apple's internal performance claims and should not be confused with the independently uploaded Geekbench score.
The benchmark result provides an additional data point, but a single benchmark submission cannot establish overall performance across every application or workload.
Why the GPU Upgrade Matters
The M5 Ultra is aimed at users who need substantial local computing power.
Apple says the GPU contains Neural Accelerators in each GPU core, allowing the chip to accelerate AI-related calculations.
The company also says M5 Ultra can deliver up to 4.5 times the peak GPU compute for AI compared with M3 Ultra in its testing.
For professional users, the combination of GPU performance, unified memory and memory bandwidth can be particularly important when working with large datasets, high-resolution video, 3D rendering and AI models.
Up to 512GB of Unified Memory
One of the biggest hardware features of M5 Ultra is its memory capacity.
Apple says Mac Studio with M5 Ultra can be configured with up to 512GB of unified memory.
The chip also provides up to 1.2TB/s of memory bandwidth, which Apple says is 50% higher than M3 Ultra.
Unlike systems where CPU and GPU memory are generally separated, Apple's unified memory architecture allows the processing components to access a shared memory pool.
This can be useful for workloads that require large amounts of data to be processed locally.
M5 Ultra and Local AI
Apple is positioning the M5 Ultra heavily around artificial intelligence.
The company says its Neural Accelerators are integrated into the GPU cores and can help accelerate AI workloads.
Apple's testing claims that Mac Studio with M5 Ultra can achieve up to 4.3 times the peak AI compute performance of M3 Ultra.
The large memory capacity is also important for local AI applications. Apple says the system can run large language models containing hundreds of billions of parameters entirely on the device.
This approach allows demanding AI workloads to be processed locally rather than relying entirely on cloud computing.
What the Geekbench Result Does and Does Not Show
The 360,019 score is an interesting early benchmark result, but it should not be interpreted as a complete measurement of M5 Ultra performance.
Benchmark results can vary depending on the exact configuration, software version, workload and testing conditions.
Geekbench measures specific workloads rather than every type of task a professional computer might perform.
For that reason, the result is best viewed as an early performance indicator.
Additional independent benchmarks will provide a broader picture of how M5 Ultra performs across gaming, creative applications, rendering, AI workloads and other professional tasks.
M5 Ultra vs M3 Ultra
Apple officially positions M5 Ultra as a substantial upgrade over M3 Ultra.
Its published specifications include a larger memory bandwidth figure, new GPU technologies and Neural Accelerators.
Apple says the M5 Ultra offers up to 40% faster graphics performance than M3 Ultra and up to 4.5 times the peak GPU compute for AI in its testing.
The newly uploaded Geekbench result provides another performance reference point, although it does not independently confirm Apple's complete range of performance claims.
| Feature | Apple M5 Ultra | Apple M3 Ultra |
|---|---|---|
| Maximum CPU cores | Up to 36 | Up to 32 |
| Maximum GPU cores | Up to 80 | Up to 80 |
| Maximum unified memory | Up to 512GB | Up to 512GB |
| Memory bandwidth | 1.2TB/s | 800GB/s |
| Graphics performance | Up to 40% faster, according to Apple | Previous generation |
| Peak AI GPU compute | Up to 4.5x, according to Apple | Reference generation |
Apple's comparison figures are based on its own testing, so actual performance can vary by application and workload.
Mac Studio With M5 Ultra
The M5 Ultra debuted inside Apple's latest Mac Studio.
Apple says the new Mac Studio starts at ₹6,29,900 in India for the M5 Ultra configuration.
The company says customer availability begins September 22, 2026, while the 512GB unified-memory configuration is scheduled to arrive later in October.
The system is targeted at professionals working with demanding workloads such as video editing, visual effects, 3D rendering, software development, data science and AI.
Bottom Line
The newly uploaded Geekbench result gives an early look at the graphics performance of Apple's M5 Ultra, with the chip recording 360,019 points in the Metal GPU benchmark.
Apple's official specifications confirm that M5 Ultra can feature up to an 80-core GPU, 512GB of unified memory and 1.2TB/s of memory bandwidth.
Apple also claims up to 40% faster graphics performance than M3 Ultra and substantially higher AI compute performance.
The Geekbench result is an interesting early data point, but more independent benchmarks will be needed to understand the chip's performance across a wider range of real-world workloads.