The variation in DDR5 CAS latency and speed between different generations can significantly impact performance.
The variation in DDR5 CAS latency and speed between different generations can significantly impact performance.
I’m planning to move up to 13600k and checked out TEAMGROUP’s T-Create Classic DDR5 CL46 32GB kit (2 x 16GB) at 5600MHz. Since I’m not aiming for a lot of RGB, the 32GB DDR5 at $97 looks like a solid choice. How does this stack up against the G.Skill RipJaws S5 Series DDR5 6000 CL36/30 or the TEAMGROUP kits you mentioned? Usually I stick with Kingston or HyperX, but with today’s prices the Kingston Fury Renegade Silver DDR5 CL32 at 6400MT/s is a close alternative, about $70-80 less than what I’m considering. The other options are around $20-30 more.
You can adjust the memory settings to achieve performance similar to what's shown in this video and others. Generally, more precise timing offers superior results compared to higher speeds. A compact 6000 kit should suffice for most needs.
In the 13th generation era, speed matters a lot, so higher frequency usually wins. A 6400 CL40 setup should outperform a 6000 CL30 one, especially if you're focused on performance. When checking timing, focus on the overall latency formula: (tCL + tRCD + 8) multiplied by 2000 divided by data rate. This gives you the main latency figure. Based on this, a CL30-38-38 package is only marginally faster than a CL36-36-36, while a CL36 is about 10% quicker than a CL40-40-40. CL30 chips tend to overclock more since they use Hynix memory; with a solid board, you can push them to 7200+MHz. If you plan to enable XMP and don’t want noticeable gains, stick with the CL36. For the best balance, choose the TeamGroup CL40—it offers great value.
This is my first experience with DDR5, though I believe I already have some background on boosting DDR4 speeds. Now I’m considering either a model with the smallest cas latency or the one rated for the highest performance. After reviewing, I’m leaning toward choosing the one with the lowest latency or top speed rating. Then I plan to attempt overclocking and fine-tune both settings. Based on what others have shared, the CL46 CL30 kit at 6000MHz seems to be the most suitable option. I’ll aim for the highest possible timing that fits the motherboard’s limits, ensuring precision and tightness.
Comparing the two, I’m unsure which is more realistic—doing both overclock attempts or focusing on one. The review mentioned a 5600MHz CL46 as the least ideal for me, so depending on the game, the difference might be noticeable. However, my own testing with 1440p and 3080p resolutions has already improved performance compared to 1080p and 4090.
Regarding the RAM speeds, I’m curious about the differences between 7200MHz 34-44-44-28 and 6000MHz 30-38-38-28. Why would I aim for these particular configurations? Because the review suggests those were the best-tuned settings that delivered strong results. Also, since I’m still learning, I’m not sure why the hardware unboxing gave a lower T-RAS compared to other options. It might be too late, but I’m following the general advice to tighten timings carefully.
Initially, I need to clarify that you’re requesting quotes so we can receive a notification. I just returned to this page. If you plan to boost performance, focus on Hynix DDR5 solutions. Both the 6000 CL30 and 6400 CL40 models are confirmed Hynix, while the 5600 CL46 kit is likely to be a mix. They’re grouped similarly, and if it was assembled within the last four months, it should support DDR5 8000 if your hardware can handle it. My setup is a 6000 CL30 Flare X5 kit, which easily runs 7800 CL36 and could likely manage 8000 CL38 with extra CPU tuning. I should note that pushing DDR5 overclocking is challenging, particularly on the 13th generation controller. Unless you’re ready to experiment extensively for stable high speeds, aim for more achievable rates like 6800MT/s (depending on your board). Also, Intel’s specs differ: some chips lack half the secondary timings—set those instead. Hynix memory features minimal tRAS, which is straightforward. Hynix diesels tend to run close to CAS latency, though you can adjust tRAS if needed. On 11th generation and older Intel boards, tRCD and tRP were fixed due to controller constraints. With Hynix DDR5, the main rule is even CAS latency across most boards, especially on Z690/Z790 from EVGA. tRP is typically one tick lower than tRCD, though this isn’t always certain. You can often swap tRP for tRAS as long as their sum matches the target. If you skip effort, simply keep timings identical. The reason tRAS never dropped below the others on DDR4 was due to tRAS and tRCD limits; DDR4 had much stricter tRAS caps (21 for Ryzen, 28 for Intel). Most DDR4 variants couldn’t reach those extremes, and other primary timings usually dropped below 21 at typical speeds.
However, the video shared here and numerous others demonstrate your mistake.