 |
| Comparison: AMD Ryzen 9 CPU 3950X vs 3900X |
AMD has released a new Ryzen 9 3950X 16 core processor, but is it worth paying more than 1200 39XX cores? , we'll look at how they differ in different games and applications, whether in stock or overclocked, to see what you can get.
AMD Ryzen 9 3950x
Compare these two main processors, because the 3950X is basically the next step to the 3900X. So the goal is to help you decide which option to use. 3900X costs $ 500, 3950X 50% off the price of $ 750. However, due to high demand, availability is low and price is usually higher. Two processors were tested in the same system. I am testing MSI x570 Unify motherboard, which handles all of these processors easily. Use Steve in "Hardware Unboxing" 3900X to test VRM with Unify and do well compared to other products. Unlike most other MSI cards that I use on the channel, there is no RGB here, and I'm sure I will be satisfied. The card is over $ 300 and the update price is shown in the description.
However, it does include great features like a 2.5GB Ethernet slot, 6 WLAN slots, and 3 PCIe ports for 4M.2. For other components, I used 16GB DDR4 3200 CL14 memory and used Nvidia RTX 2080 Ti to reduce GPU bottlenecks. I am using Deepcool Castle 240ex all-in-one liquid cooler on processors with Noctua NT-H1 paste for comparison with Apple. We tested both processors and enabled Precision Boost Overdrive (PBO) so that we can improve performance by making simple changes.
This is basically similar to overclocking, but the results differ for each motherboard and cooling solution. With that in mind, we will first look at the differences between different applications, as well as performance and temperature, then play games at 1080p and 1440p, and finally we compare the performance of each measurement in dollars. Let's start with Cinebench R20. In the top half of the chart, I activated PBO via Ryzen Master and got results in the bottom half of the results. Not surprisingly, the 3950X has a multi-core result, it has 4 cores. In terms of inventory, it is 27% faster than 3900X, and once PBO is activated, the result increases by 30%. Results for each reference vary, as 3950X is still 4.5% higher than inventory.
After activating PBO, I found that the only reference rating for 3950X is a little difficult, it improved 3900X, which means that 3950X occupies only 3.3% area. Although the Cinebench R15 has been replaced by the latest R20 model, I would also like to take the results of this issue into account given that many other products still use it, which gives me more comparable numbers. With PBO enabled, the only major result for 3950X remained the same, while 3900X rose 3.3% with PBO. Of course, there's also a truly amazing 3950X multicore technology, 26.5% higher than the stock market's 3900X technology and up to 29% when PBO is activated. Blender benchmarks tested for BMW and Classroom cars. Since another workload that wants to increase the kernel level, the profit of 3950X is good, which is 26% higher than the shares of 3900X. After activating PBO, 3950X is now 30% faster than 3900X,
but we can see that the result of 3900X does not change significantly by activating PBO. Use Handbrake to convert a 4K file to 1080p and another 1080p file to 720p. This is another workload that uses more cores, but I don't see much difference in the other multi-core tests that have been introduced yet. In stock, 1080p files are 10% faster than 3900X, while 4K files are only 3% faster than 3950X. With PBO, 4K file speed increased by 3950x and speed by 3%, while 3900X achieved greater advantage and reduced lead by 3950x by 8.5%. A video was exported to a 4K laptop test using Adobe Premiere. In stock, 3950X performed 7% faster than 3900X, but PBO enabled 3900X improved in just one second, and 3950X is now 1.4% faster than PBO off.
This means that if PBO is enabled on both chips, the 3950X can do video export tasks at a rate 8.4% faster than 3900X. I've also tested the effect of the distortion stabilizer in Adobe Premiere, which mainly handles playable video files, but the process is often less formatted. The 3950X speed increased by 5%, but it is interesting to note that the 3950X actually performs worse after PBO activation, as seen in the Cinebench R20. That is, if both PBO software is enabled, I don't see a clear difference between the two. I was 7 zip to test the pressure speeds and decompress. This is the only test where 3900X has exceeded 3950X, which represents only 1% of pressure, and almost before activating PBO. Decompression process is a completely different story, but 3950X is 30% faster than 3900X when running this process, and 34% faster PBO is 34% faster, which represents the biggest performance improvement for any app you've tested. Another V-Ray standard is another workload that will significantly increase the number of major accounts.
For this reason the 3950X score is almost 31% higher. Once PBO was activated, the 3950X is now 32% ahead of the 3900X, which is also one of the biggest noticeable improvements in all tests. In a sense, the Corona standard is similar, it also uses basic processors to create the scene, so the 3950X has a huge improvement of 28% in stock. After PBO is activated, the speed of the two is just a few seconds, leaving 3950X by a difference of 28%. He also started Geekbench due to some requests: the 3950X single core isn't 1% faster, and the 3950X multi-core score with PBO enabled is about 13% higher. These are the differences between the 3950X and 3900X processors in all of these applications, as we can see that this really depends on the specific workload. I know I'm mixing single-core and multi-core tests here, so an average 15% increase doesn't make sense.
We can see that display apps like Blender, Cinebench, Corona and V-Ray, like processor core cores available in the 3950X, were the only kernel results that were good as well. As accuracy increases, the performance of a single-core result decreases. As we've seen in some cases, the 3950X seems to be working a little better here, but this makes up for the fact that PBO can still improve it. Perform multi-core workloads. I have not tested manual overclocking because I found that PBOs generally work better because we haven't set the speed limit less than the high workload. I also measured the overall system performance on the wall when the mixer was running by default. Despite the 3950X price, it actually consumes less power than 3900X, although it has 4 more cores and tests can be done faster. This is a very impressive result, and thanks to the quality of the high-quality silicon, the 3950X seems to be able to work under low pressure. We've counted 3950X, but we're still enabling PBO, but I've found 3900X a little different, so 3950X has even more improvements in PBO activation.
These are the processor temperatures in which the mixer test itself is activated and the type of results reflects the just obtained tensile strength results. Although the 3950X supports four additional charges, it still runs cooler, also due to the fact that the operating voltage 3950X is less than 3900X. When the PBO is activated, the temperature rises relatively easily, but since I use a coolant with a diameter of only 240 mm, we can increase the temperature with better cooling. If I don't see other people as uncommitted devices, then I've reported that 3950X uses less power and runs cooler, I think I'm crazy. Then we go to scoring games. I tested these games at 1080p and 1440p resolution. These are just stock results, then we'll review PBO. Red Dead Redemption 2 tested using in-game measurement tools. I have 1080p results at the bottom of the chart and 1440 results at the top of the chart. This doesn't seem to make much difference, but this game saw one of the biggest changes in every game tested: 1080p, average frame rate 3950X increased 2.3%, average FPS frame rate at 1440px 3%, 2 also. %. Battlefield V tested in campaign mode. Again, the difference doesn't look like much, but at 1080p, this game recorded the biggest improvement on 3950X at a rate of 3.1 SPF higher than 3900X, but with an increase of at least 6% to 1%. At 1440 pixels, the results were similar and the average frame rate of 3950X increased by about 3%.
Assassin's Creed Odyssey has been tested with the included benchmark, and I generally think it's a heavy CPU test, but the extra kernel on the 3950X seems useless, just 2.2% improvement over 1080 pixels and 2% average frame rate at 1440p. The console was tested with high presets and the results again were very similar. I think you did not know the processor used in the blind game test. At 1080 pixels, 3950X is 2.3% higher than 3900X and slightly decreased to 1.7% in 1440. Tomb Raider Shadows were tested using standard gaming tools. At 1080 pixels, the average SPF increased slightly by 1.5% and by 0.9% to 1,440 pixels. Fortnite has been tested with exactly the same restart file on both processors using the restart feature. The difference isn't that interesting here: with 3950X, the average FPS at 1080p is only 0.6% higher, and at 1,440 pixels, the difference is almost zero. PUBG was also tested with recovery features, using same recovery feature in all cases. Interestingly, the 3900X resolution is a little higher this time at 1080p, but like most other games we've seen so far, the difference between the two is very small, just a few frames higher. With a resolution of 1440 pixels, the 3950X was rebooted and got 0.23% support, which is still good in the error range anyway. "Rainbow Six: Siege" has been tested according to game standards, and the only other test that the 3950X actually does is 3900X at 1080p, but that's just a small part. By 1440, the distance between the two became so narrow that it was not important at that time.
Borderlands 3 was also tested with in-game performance measurement tools: at 1080p, the average FPS increased slightly by only 1%, and increased by 3% with a 1440p resolution setting. It might not look like this, but Ghost Recon Breakpoint increased the average FPS level by more than 2.2% to 1080p despite dropping by 1% regardless of the processor. At 1440p the average SPF is closer and there is no really big difference. In 10 tested games, we only increased the frame rate from 3950X to 1080 pixels by 1.3%. Some 3900X games are better, I don't know if they are just basic differences or other factors, like occasional basic performance, and sometimes even the 3950X that we saw earlier, but in reality, the battlefield also the best case 5 is not important either. At 1440 pixels, there are no games in the 3900X competition, while the average improvement in 3950X has slightly improved and the average improvement was 1.4%. In fact, there is not much difference between the two. I think we can say with certainty that when it comes to games, spending more money is close enough to 3950X, 3900X doesn't make sense anyway, because the game is starting to tap additional hearts, the future could be a change. If your entire system is designed for gaming, you will get better service than other products like Intel 9900K, and you can use the link in the upper right corner to see how it is stacked. Some of the rendering methods indicate that neither of these two core processor processors can benefit from additional use of basic production. However, we found it to work fine even if the absolute results aren't the best, well, to play a role.
Watched Battlefield 5 with and without Precision Boost Overdrive. Both have minor improvement with PBO enabled, but it's about 1fps, so nothing surprising. Of course, this may be different for games, but since we've seen such accurate results in all games, I think there's not a lot of willingness to use PBO. Then check the value. This is CPV from the top ten games tested in stock. Technically speaking, the 3900X is equipped with a cooler that you can use. Hope Deepcool AIO works better, but maybe you can get the most results with the included Wraith Prism cooler. That's why I have a 3900X and the original coolant with me. If you purchased a 3950X without purchasing a cooler, please take unnecessary measures because the 3950X will not come with a cooler and you will need to purchase one.
This increases costs. If we set aside $ 100 for 3950X to get a decent AIO, that would cost more than 3900X. This also applies if we bought a cooler for the 3900X device, as the game performance is very similar between the two, but the 3950X cost is not less than $ 250. It has nothing to do with the game. It performs multi-core graphics performance Cinebench R20 per dollar. Basically, each number on the graph indicates the number of points that Cinebench can exchange in dollars.
it's better. The value of 3900X continues to increase as the cost of 3950X increases by at least 50%. However, in dense networked applications like Cinebench, the speed increased by about 30%. If you really need the extra multi-core performance of the 3950X, you have to pay more to get it.
Faced with all these conditions, which of these CPUs do you get, i.e. 3900X processors with 1200 cores or 1650X 16 core processors? Personally, because I don't play games, I may rely on 3950X for video editing, but my Adobe Premiere test doesn't make much of a difference to prove its legitimacy, which is good for the future. Realistically, 3900X has higher value and great performance.
In games, the CPU works well with little difference. However, if you are only interested in games, you cannot choose between these two processors and consider other products such as Intel i9-9900K. If you want a good product, the 3900X or 3950X can work well, depending on the amount you spend and the number of additional cores and additional performance. 3900X has higher value and excellent performance. If you are ready to spend more money and run apps connected to a large network, you can use 3950X. Let me know which AMD processor you have chosen and why you chose 3900X or 3950X in the comments.