Megatransfer: The RAM revolution in 3 steps 🚀

Key Findings

• Previously, RAM was measured in MHz based on clock cycles, but now MT/s is used for data transfers.
• MT/s indicates a theoretical maximum data transfer rate, not a direct correlation to MHz.
• Rating RAM in Mbps can give a clearer picture of actual data movement compared to MT/s.

You've probably noticed that some computer specifications now list RAM speeds in MT/s or "Megatransfers per second" instead of the traditional Megahertz. Why this change? Glad you asked! 🤔

How RAM Specs Worked Before

Traditionally, RAM was rated in megahertz, where one MHz represents one million clock cycles. Computer hardware operates on a clock cycle represented by a square wave, with a peak and valley in the wave representing a single cycle. With each cycle, RAM can complete one memory operation, such as a read or write. In other words, you can perform one million transfers every second with RAM running at one MHz.

This made it fairly easy to calculate RAM speed. If you had 133MHz RAM, it was half as fast as 266MHz RAM. However, with the invention of DDR (Double Data Rate) RAM, it became possible to complete a data transfer at both the peak and valley of the wave. Effectively doubling the data rate, hence the name. 🔄

So, 133MHz DDR would effectively be as fast as 266MHz of traditional RAM. To simplify things for consumers, most RAM manufacturers chose to market their DDR RAM as "effectively" the MHz rating of regular RAM. So, you might see something like DDR4 3200MHz RAM, but the clock speed actual RAM is only 1800MHz.

The Problem with MHz as a Measure of Performance

So what's the problem? Well, for one thing, it means that the marketed specifications for RAM are technically inaccurate. More importantly, however, megatransfers are a measure of the actual amount of data that can be moved, meaning it's not a one-to-one analysis with megahertz. The MT/s specification is the theoretical maximum amount of data that can be transferred in one second. While RAM will always operate at its rated frequency, it won't always reach its maximum MT/s rate, due to varying computing workloads. 📊

There is also the future of RAM to consider, as memory QDR (Quad Data Rate) can perform four operations per clock cycle: two for reading and two for writing. This is even more complicated to express in MHz, so something that addresses real-world data volumes seems to make more sense. 🔍

Mega Transfers Make More Sense, But They're Not the Whole Picture

While I believe MT/s is a better measure of RAM performance than MHz, it's not perfect and doesn't represent the whole picture. On the one hand, RAM latency also matters, but if we're focusing on data transfer performance, the stakes are higher. 📈

You see, the actual amount of data moved in a megatransfer depends on the memory bus width. That is, the number of bits that can be moved per transfer. The main RAM in modern computers has been 64-bit for years, meaning you have a direct comparison between two RAM modules as long as they have the same bus width. 🍏

However, MT/s wouldn't be as straightforward for GPU VRAM, since GDDR (graphics DDR) uses a much wider bus. Low-end GPUs have 128-bit buses, while 256-, 384-, and 512-bit buses are more typical.

So, it might make even more sense to rate RAM in terms of Megabits per second (Mbps), which tells you exactly how much actual data the RAM can theoretically move. Many memory manufacturers already do this. If you look at the specification sheet From a Samsung memory module, you'll notice it's rated in Mbps, not MT/s. This eliminates both MHz and MT/s and tells you exactly what the total memory bandwidth is. 💾

Whether MT/s or Mbps will ultimately prevail remains to be seen, but I think it's clear that either way, MHz ratings for RAM are no longer the best way to express performance. We'll be looking at all three for a while longer, so it's best to know the difference! 🧐