Upgrade the memory to enhance system performance
Upgrade the memory to enhance system performance
Evening all,
I currently possess
AMD Ryzen 7 5700X 8-Core Processor (4.90 GHz) overclocked
Nvidia Geforce RTX 3060 12gb
32.0GB Ram (crucial) – speed 3466 MT/s overclocked
Windows 11 Pro 24H2
ASRock B450 Steel Legend ATX AM4 Motherboard
x2 M.2 1tb and 2 SSD drives, each 1tb
I’m considering buying these
https://www.amazon.co.uk/gp/product/B0C2...A1OLE&th=1
64gb, 32gbx2 pro
Would it be worthwhile? I mainly use a computer for Photoshop (4k/8k work) and 3 screens for sim racing ACC (resolution runs at 3840 x 2160)
Many thanks, Richard
The tech specs reference a part number CP2K32G4DFRA32A with specific latency details. It mentions a CAS latency of 22, secondary timings at 22-22-22, indicating relatively slow performance. The ratio of 3600mhz to Fclck in UCLK versus Infinity fabric is noted, suggesting tighter timing could improve efficiency. Using 4x16GB instead of 2x32GB may enhance memory interleaving and boost bandwidth for productivity apps. Dual rank dimms might add a small latency penalty (5–10ns) due to increased controller processing, but 4x16GB single-rank dimms could be a better option. Choosing tighter timings may help balance gaming needs with productivity demands. Some users are seeking benchmarks for memory performance in Ryzen systems, though exact guidance is hard to find. There’s also discussion about optimizing memory usage for games and high-performance scenarios.
Perhaps. Some say it 'stresses' the memory controller. Yet I lean toward calling it creating more work for it.
In terms of memory performance, there are two main aspects to consider: bandwidth and latency. The top-end option can reach up to a maximum of 59 gigabits per second, but real-world limits bring it down to around 55 gigabits per second. Latency-wise, the best options can hit 50 nanoseconds, whereas typical values range from 60 to 65 nanoseconds, with some reaching as low as 80 nanoseconds or less (though the lower end is often unreliable).
Given the uncertainty about overclocking capabilities of any specific RAM set, it's better to purchase a kit that matches your timing requirements. Common timings include CAS 22, 20, 18, 16, or even 14. Higher latency RAM tends to be pricier—prices can exceed £400 for a CAS 14 kit.
The motherboard support site provides a memory compatibility list, showing which RAM models have been tested with the board. These should work reliably if you locate them and they're available. Some vendors also note on this page that mixing two different RAM kits isn't recommended, even if they look similar, due to potential incompatibility issues.
Therefore, it's wise to buy one kit suited to your needs and verify its compatibility with a recommended model from the Qualified Vendor List. Whether it matters depends on luck or how consistent production runs have been. There might be alternatives or 'non-QVL' options that function because they use the same integrated circuits as QVL models.
Ultimately, you'll need to explore specification pages to uncover details about the memory's design—such as which ICs were used (e.g., Samsung B die for overclocking, or Hynix C for more flexibility). You can also search forums for recommendations on affordable alternatives that still deliver good results.
There are more affordable or free memory benchmarks available, such as downloading RAM benchmark tools.
vovsoft.com offers these options, and there are also those with RAM benchmarks at https://www.ocbase.com/download.
You can try the PassMark PerformanceTest PC Benchmark Software to test your PC's hardware speed and compare it with other machines. It covers disk, 3D, and CPU tests.
www.passmark.com