**Benchmarks and Performance Comparison: M1 vs i9-12900HK**
In this article, we will explore the performance comparison between Apple's latest M1 Pro chip and Intel's 12th gen i9-12900HK processor, using various benchmarks and tests. We will dive into the results of these tests to understand how the two processors compare in terms of overall system performance, responsiveness, and specific workload-based performance.
**CrossMark Benchmark**
One such benchmark that we tested was CrossMark, a cross-platform benchmark that aims to measure overall system performance and responsiveness. The test has a native M1 version for MacOS, which allows us to directly compare the performance of the M1 Pro and i9-12900HK. In this test, we found that the i9-12900HK scored 30% higher than the M1 Pro. This test also had three sub-scores titled productivity, creativity, and responsiveness, which sounded a bit arbitrary to me. Take the responsiveness score for example, where the i9 was 74% ahead, but I'm not sure if it means the i9 is 74% more responsive. The test's results do suggest that the i9-12900HK is significantly faster than the M1 Pro in this aspect.
However, we also tested another benchmark called Speedometer, which runs in a web browser to test the responsiveness of web applications and scripts. After averaging 5 test runs, both laptops scored exactly the same. This result was quite surprising, as my Threadripper 1950X system scored 126 here. Maybe I need an upgrade just to browse the Internet.
**Geekbench Benchmark**
We also tested Geekbench, a popular benchmarking tool that measures CPU performance in various workloads. In this test, we found that the i9-12900HK was scoring just 4% higher than the M1 Pro for single-threaded workloads and 14% higher for multi-threaded workloads. These results might seem significant, but considering the i9 has double the total thread count, it's not as impressive as it seems.
**Other Workload-Based Performance**
We tested various other workload-based performance benchmarks to get a more comprehensive view of how the two processors compare. The biggest difference we found was in AES encryption and decryption, where the i9-12900HK performed significantly better. However, there were some tests that showed no significant difference between the two processors.
One such test was the LLVM compilation test, which showed that the M1 Pro was actually faster than the i9-12900HK. However, there were other tests that didn't show a significant performance difference between the two processors. We would have loved to tested more compiling benchmarks, but finding tests that support both M1 and Windows was a challenge.
**Conclusion**
In conclusion, based on our tests, it's clear that last year's M1 Pro can't offer quite as much performance as this year's i9-12900HK, particularly in multi-threaded workloads. However, if you prioritize other factors such as size, weight, power consumption, and price, then the performance differences matter less.
We're also looking forward to testing Apple's new upcoming M2 chips later this year and comparing them against AMD's new Ryzen 6000 laptop processors. These comparisons will provide more insights into the latest developments in processor technology and how they compare against each other.
**Future Comparisons**
We'll be keeping an eye on these two laptops and their respective processors, as well as Apple's upcoming M2 chips and AMD's new Ryzen 6000 laptop processors. If you're interested in staying up-to-date with our comparisons, make sure to subscribe to our channel for future updates and reviews.
**Additional Resources**
For more information on the i9-12900HK and GE76 laptops, check out our videos over here next.
"WEBVTTKind: captionsLanguage: enI’m going to show you how Intel’s new 12th gen laptop CPU compares against Apple’s M1 Pro.To do this testing I’ve got Apple’s MacBook Pro 16 with the M1 Pro, a $2500 USD machine, and MSI’s GE76 gaming laptop with Intel’s i9-12900HK, a much larger and more expensive $4200 machine.These are the main differences between these processors. Apple’s M1 Pro is available with 8 and 10 core options, but I’ve got the higher 10 core here which has 8 performance and 2 efficiency cores. The i9 on the other hand has 14 cores with 6 performance cores but 8 efficiency cores, resulting in its total thread count being double compared to the M1 Pro at 20.Now you could argue that this isn’t a fair apples to apples comparison, because only one of the laptops is from Apple…But seriously, it’s because the Intel i9-12900HK is Intel’s top end laptop option while the M1 Pro comes in behind Apple’s top end M1 Max. Now from the testing that I’ve seen, the M1 Pro and M1 Max aren’t too different in terms of processing performance. The main difference is the GPU which isn’t the focus of this comparison.Likewise Intel’s Core i7-12700H that comes in below this still has the same core and thread count, so it’s probably going to perform pretty similarly anyway. So all things considered, I think this is a pretty fair comparison between Intel 12th gen and Apple’s M1 Pro.There’s a fairly big size difference between these two laptops. The MacBook Pro 16 is much smaller in every dimension, so less space for cooling comparatively.Of course there will be thinner Intel based laptops featuring the i9 processor, but they’ll have lower power limits and won’t perform as well as what I’m showing here.When running a multi threaded CPU benchmark, the larger and thicker Intel machine still ends up running hotter despite it having more room for cooling. To be fair, these aren’t directly comparable as results depend on where sensors are in the chip and how software reports them, but it gives us a rough idea that when in a sustained heavy workload, both chips may run above 90 degrees Celsius.Now hang on just a moment, although the M1 chip does appear to be running quite warm, let’s have a listen to the differences in fan noise.The MacBook was running significantly quieter, and that’s because it’s drawing less power from the wall, as shown here. In the Cinebench R23 multi core benchmark, the Intel i9-12900HK laptop was using 188% more wattage from the wall, a crazy difference.Well I suppose it’s only crazy if the i9 doesn’t also give us a corresponding performance uplift, so let’s have a look and see how both perform in different applications.Cinebench is just a nice way to quickly get a rough idea of what to expect in terms of both single and multi threaded performance - don’t worry we’ll cover more practical workloads next. The single core score was 24% higher with the i9 in this test, while the multi core score was 41% ahead. Decent improvements to be sure, but for 188% more power? The M1 chip is clearly much more power efficient.The M1 Pro’s scores are actually quite good and in line with good Intel 11th gen and Ryzen 5000 laptops from last year, it’s just that this year Intel 12th gen really pushes things to the next level by giving us 14 processing cores and 20 threads, up from the best case last year which was 8 core 16 thread chips.Performance on battery power is where things continue to get impressive for Apple. Although the i9-12900HK still has the lead here, the M1 Pro is essentially scoring the same in both single and multi threaded workloads, regardless of whether or not it’s plugged into wall power. The Intel machine lowered by about 5000 points just from unplugging the power, which further helps show the power efficiencies with the M1 design.Alright back to comparing with both plugged into wall power for max performance.Adobe Premiere is a more practical test and was tested with the Puget Systems benchmark. Adobe has an M1 native version of Premiere for MacOS, so this was used for best performance on the MacBook Pro 16. In terms of overall score difference, Intel’s top end i9-12900HK was scoring just 7% higher than the M1 Pro here.Now this test utilizes both the CPU and the GPU, and you could argue that as Intel’s i9-12900HK has its own built in graphics that it’s not fair to include the Nvidia RTX 3080 Ti, so I’ve also included results from just the Intel graphics so you can see the difference. Without Nvidia, the M1 Pro was scoring 91% higher than just the Intel processor alone with its integrated graphics. You could also argue that this isn’t fair, as it’s unlikely you’ll actually come across an i9 laptop that doesn’t have Nvidia discrete graphics, but now you’ve got details with both.Those were the overall test scores, we can drill down into specific aspects like exporting for example. In terms of actually exporting a video project, the i9 plus Nvidia combo was just 4% ahead of the M1 Pro. Personally I don’t think it matters if an export takes a few seconds or even a few minutes longer, there are more important aspects.The live playback scores were also relatively close, and this represents how smooth it is to actually playback footage in Premiere on each laptop, which I’d argue is more important. The i9 was scoring just 5% ahead of the M1 Pro in this one, while again without the Nvidia graphics the i9 with only its integrated graphics was far behind.The differences are a bit larger when looking at the effects score with the i9 plus 3080 Ti reaching 22% higher than the M1 Pro.It’s not that surprising that the GPU score has the largest gap between the two as the RTX 3080 Ti is certainly a powerful beast. This is one area where I would expect the jump up to Apple’s M1 Max to close the gap, as it has more GPU cores.I’ve also tested DaVinci Resolve. Although Resolve itself does have an M1 Native version for MacOS, unfortunately the Puget Systems benchmark doesn’t currently support it, so I’ve had to make this test up myself. Basically this is just a random 20 minute 4K 60 FPS video I imported into Resolve and then exported using the YouTube 4K preset. Even with the Nvidia graphics disabled, the 12900HK was completing the video export 35% faster compared to the M1 Pro, while the i9 plus 3080 Ti combined were chewing through the export 57% faster than the M1 Pro. Again this is certainly an area where the M1 Max’s additional GPU cores would likely narrow the gap, and of course things could be much different depending on the project and various effects and color corrections.Now it’s not all massive gains for Intel. I’ve used the Phoronix test suite to compile LLVM on both machines, and Intel’s i9 was completing this task 25% slower compared to the M1 Pro. I was a little surprised given the i9 has a much higher thread count, but at the same time perhaps there’s a good reason why so many programmers stick to the MacBook.The M1 Pro had a big defeat in my Handbrake video transcode test, where Intel’s i9-12900HK was able to complete the same task a massive 131% faster. Although there is an M1 native version of Handbrake, which I used for this test, I believe handbrake is able to take advantage of Intel’s AVX processor instructions to complete the task faster.7-Zip was used to test compression and decompression. The i9 had the lead here too, despite the latest version of 7-Zip having native M1 support. The i9 was 28% faster in compression, but then it had a much larger 66% boost when it came to decompression.With Blender we start getting into applications that don’t yet have full M1 support. While running the benchmark I confirmed that it was running via the Rosetta 2 emulation layer, so there could potentially be a performance improvement in future if this test is updated with M1 support. I guess you could argue this isn’t fair, but the fact is it’s the real current state of M1 - not all applications have been updated for optimal performance. Anyway with the overhead of the emulation layer, the i9 was completing the longer Classroom test 44% faster than the M1 Pro. Even with a native M1 version, honestly I’d still expect the i9 to win here given this is a heavy multi threaded rendering workload. I think it’s just too much to ask the 10 core 10 thread M1 Pro to beat the 14 core 20 thread i9.The corona renderer benchmark also runs via Rosetta 2 with no native M1 version, which is probably in part why the Intel i9 laptop was able to complete the task 75% faster than the MacBook. Again this is a heavy threaded test, so even with a future M1 version my money is on the i9 still winning, granted with a smaller difference.Likewise the V-Ray benchmark is also an x86 application with no M1 support as of yet, so that combined with the difference in multi threaded performance between the two chips is why the 12900HK is able to score 63% higher than the M1 Pro in this workload.VeraCrypt was used to test AES encryption and decryption and is a bit of a weird one. Although there is actually native M1 support for this version of the software, it might not be taking advantage of hardware accelerated AES, which would explain why the i9 is over 10 times faster. With x86 processors, the AES-NI instruction set can be leveraged to boost performance. Honestly I’m not sure what exists here for M1, but in any case the simple fact is that right now at the time of testing the latest version of VeraCrypt clearly can’t make use of any similar feature.CrossMark is a cross-platform benchmark that aims to measure overall system performance and responsiveness. This test has a native M1 version for MacOS, and I found the i9-12900HK to score 30% higher than the M1 Pro. This test has three sub scores titled productivity, creativity and responsiveness, which sound a bit arbitrary as I don’t really know what they test. Take the responsiveness score for example, the i9 is 74% ahead, does this mean it's 74% more responsive? Probably not, given I’d say my experience using MacOS felt smoother for the most part compared to Windows 11, but hey more number means more better right?For example if I use the Speedometer benchmark which runs in the web browser to test the responsiveness of web applications and scripts, after averaging 5 test runs both laptops actually scored exactly the same. Just for context, my Threadripper 1950X system scored 126 here, so maybe I need an upgrade just to browse the Internet.Speaking of less useful benchmarks (with regards to Crossmark that is), I left Geekbench until later because I just don’t rely on the results, but some of you ask for it. This test also has a native M1 version, and in the workloads run by this test, the i9 was scoring just 4% higher than the M1 pro for single threaded and 14% higher in multi threaded, which doesn’t seem like much given the i9 literally has double the total thread count.These are the differences in all of the applications just tested. The biggest difference was AES encryption and decryption right up the top for the reasons discussed earlier. The LLVM compilation test was the only one I tested that was faster on the M1 Pro, however that said there were a few other tests that weren’t that much better on the i9. I would have loved to have tested more compiling benchmarks, but I’m not exactly a programmer so finding tests that support both M1 and Windows was a challenge.Out of these specific tests, the i9 is over 150% faster, but if we choose not to factor in those big AES outliers then we’re instead looking at Intel being 39% faster than the M1 Pro in the rest of the tests. Now this only applies to this exact selection of workloads and isn’t a general statement as there are a random number of both single and multi threaded tests here, but regardless it’s enough to give you the idea that last year’s M1 Pro can’t offer quite as much performance when compared to this year’s i9-12900HK, particularly in multi threaded workloads where Intel 12th gen can dominate.Now that said, we of course can’t forget the differences noted earlier, like the MacBook being smaller, lighter, using less power and still getting full performance on battery while actually being cheaper, at least when up against the GE76. If those things are more important for you, then the performance differences matter less.Of course Apple are rumored to have their new upcoming M2 chips later this year, so make sure you’re subscribed for my future comparisons with that. I’m also planning on seeing how AMD’s new Ryzen 6000 laptop processors both compare against Apple M1 and Intel 12th gen, you don’t want to miss any of that! For now though you can find out more information on both of these two laptops in these videos over here next, so I’ll see you in one of those.\n"
