Every time I build a new PC, some decisions absolutely haunt me.One of these is how much RAM to get, and since it’s now a business expense, I tend to overspec how much I’ll actually need and then feel bad about how much I just installed. At least, until I start opening dozens of Chrome tabs, and then all is forgiven.But it wasn’t always like that, and I used to carefully weigh how much capacity I needed with the MT/s and timings I wanted, and then decide if I was going to buy half now, and the chance that the particular RAM kit would exist when I had more budget to get the rest. My first custom gaming PC fell into this category,where I overspent getting the black and yellow DIMMs I wanted,then went back to Newegg after three days of anxiousness to buy another 8GB of Avexir Blitz DDR3 because I was worried it would disappear from sale.

I was absolutely correct on this occasion, as the company stopped making DDR3 shortly after, and after one or two DDR4 releases, it disappeared altogether. But I also got lucky by buying two RAM kits at different times and having them both work together perfectly, which I later discovered wasn’t always the case. As always, our usual advice applies, with getting ample memory to begin with, so your PC doesn’t need upgrades straight away.

That’s going to be 32GB of DDR5 for any new build, and really, with RAM prices being so low, you can easily go for 64GB as two 32GB modules and be happy for a long time. But what happens if you need more later? Let’s go over why mixing RAM modules hasn’t been the best idea historically, and where we are now.

RAM is RAM isn’t it?

Ahh, you’d think so but it really isn’t

Let’s get the most common misconception out of the way so we can dig in. Yes, you can mix RAM kits together. You can run different speeds, timings, and capacities and still boot, but it isn’t before you think that’s solved things.

Take laptops, for example. It’s common to see weird RAM capacities that aren’t the result of two equal SODIMMs, and the reason it’s not much of an issue is that laptops often run at J## Optimizing DDR5 RAM: A Guide to Stability and Performance

Building a new PC or upgrading your existing one with DDR5 RAM can unlock significant performance gains. However, DDR5 introduces complexities that weren’t present in previous generations. Achieving optimal stability and performance requires a bit more attention to detail. Here’s a breakdown of key considerations:

• Match specifications exactly if you have to mix kits
• Ensure your BIOS is up to date
• Avoid four-DIMM configurations if possible

This might not guarantee performance, compatibility, or anything really, but it will help and there will always be some edge cases where particular RAM kits won’t work together or with specific CPU and motherboard combinations.

### DDR5 has a strange quirk

[Image of DDR5 RAM and motherboard components]

DDR5’s internal architecture differs significantly from DDR4.It utilizes two self-reliant 32-bit channels per module, instead of a single 64-bit channel. While this theoretically doubles bandwidth, it introduces a quirk: the performance isn’t always linear when populating all memory slots.

This is where the advice above comes into play. Mixing RAM kits, even those with seemingly identical specifications, can lead to instability.The sub-timings and internal configurations within each kit might not align, causing conflicts. Similarly,an outdated BIOS might not have the necessary microcode to properly handle the complexities of DDR5,leading to errors.

using all four DIMM slots on your motherboard can sometimes reduce performance compared to a two-DIMM configuration. This is because the memory controller has to manage more channels together, potentially creating bottlenecks. While modern motherboards and CPUs are improving in this regard, it’s still a factor to consider, especially when aiming for maximum overclocking potential.

TeamGroup xtreem DDR5 Memory Review: Pushing the Limits of Speed and RGB

TeamGroup has been steadily making waves in the memory market,and their Xtreem DDR5 series is a prime example of their commitment to performance and aesthetics. We recently had the opportunity to test two kits: a 32GB (2x16GB) DDR5-7200 CL34 and a 64GB (2x32GB) DDR5-7200 CL36. This review will cover our experiences with both,focusing on performance,stability,RGB implementation,and overall value.

unboxing and Aesthetics

Both kits arrive in TeamGroup’s signature Xtreem packaging, a sleek black box with vibrant RGB accents. The memory modules themselves are visually striking. They feature a black PCB and a translucent white heat spreader with extensive RGB lighting. The RGB is implemented through individually addressable LEDs, allowing for a wide range of customization options. The heat spreader design is relatively low-profile, ensuring compatibility with most CPU coolers.

Specifications

Here’s a breakdown of the specifications for each kit:

32GB (2x16GB) DDR5-7200 CL34

Capacity: 32GB (2 x 16GB)
Speed: DDR5-7200 (PC5-57600)
Timings: CL34-40-40-96
Voltage: 1.35V
XMP 3.0 Support: Yes
RGB: Yes

64GB (2x32GB) DDR5-7200 CL36

Capacity: 64GB (2 x 32GB)
Speed: DDR5-7200 (PC5-57600)
Timings: CL36-40-40-96
Voltage: 1.35V
XMP 3.0 Support: Yes
RGB: Yes

Test Bench and Methodology

Our testing was conducted on the following system:

CPU: Intel Core i9-13900K
Motherboard: ASUS ROG Maximus Z790 Hero
GPU: NVIDIA GeForce RTX 4090
Storage: Samsung 990 Pro 2TB NVMe SSD
PSU: Corsair HX1200
Cooler: Corsair iCUE H150i Elite LCD

We used AIDA64 Extreme for memory bandwidth and latency tests, and Cinebench R23 for real-world performance evaluation. Memory was tested at both JEDEC defaults and with the XMP 3.0 profile enabled. Stability testing was performed using Memtest86 for a minimum of 4 hours.

Performance

32GB DDR5-7200 CL34:

With XMP enabled, the 32GB kit consistently achieved its advertised speed of 7200MHz and timings of CL34-40-40-96. AIDA64 reported excellent bandwidth results, with read speeds exceeding 85 GB/s and write speeds surpassing 75 GB/s. Latency was also remarkable, averaging around 65ns. In Cinebench R23, we observed a noticeable performance advancement compared to running at JEDEC defaults, with a roughly 5% increase in scores.

64GB DDR5-7200 CL36:

The 64GB kit also hit its advertised speed of 7200MHz with XMP enabled, though the slightly looser CL36 timings resulted in marginally lower bandwidth and higher latency compared to the 32GB kit. AIDA64 showed read speeds around 82 GB/s and write speeds around 72 GB/s, with latency averaging around 68ns.Cinebench R23 scores were similar to the 32GB kit when XMP was enabled, demonstrating that the increased capacity didn’t significantly impact performance in this workload.

RGB Implementation

TeamGroup’s RGB implementation is handled through their proprietary software. The software is relatively easy to use and offers a wide range of customization options, including static colors, dynamic effects, and synchronization with other RGB components. The RGB lighting is shining and vibrant, and the effects are smooth and responsive.

Stability

Both kits proved to be remarkably stable during our testing. We ran Memtest86 for over 4 hours on each kit with XMP enabled and encountered no errors. This indicates that TeamGroup has done a good job of selecting high-quality memory chips and ensuring that the XMP profiles are reliable.

Value

The TeamGroup Xtreem DDR5