One of the fundamental components of any computer, be it a bulky desktop or a 6-inch smartphone is memory, also known as RAM (short for Random Access Memory). RAM is a volatile form of storage and every task that your computer performs is stored on it but as soon as the power is turned off, that data gets erased.

Now if you are an enthusiast or just someone who’s up-to-date with the latest trends in technology, you probably have come across the terms DDR4, GDDR5 and HBM. People are usually ignorant when it comes to the different kinds of computer memory, and often confuse GDDR5 for DDR5. In this post, I’ll explain each of these memory types and differentiate them from the rest, along with examples to make their operation clear.

Double Data rate Generation Four (DDR4)

Okay a bit of basic info first. Nearly every kind of memory is based on dynamic random access memory (DRAM). DRAM is slower than static ram (SRAM), because it has to be continuously refreshed by the CPU or the memory controller but at the same time it is a whole lot cheaper which is the primary reason for its widespread use.


DDR4 is the latest iteration of DRAM. Released in 2014, it doesn’t offer significantly faster speeds than it’s predecessor but instead reduces the voltage and power draw by quite a margin. There are many DDR4 and DDR3 modules that have nearly the same speed. Bear in mind though, as DDR4 memory gains more widespread adoption, the bandwidth will also increase.

DDR4 Is Not All That Fast…Yet

The primary advantages of DDR4 memory over DDR3 are higher DIMM sizes (upto 64 GiB, comparatively DDR3 is limited to 16) and lower voltage requirements. Although the memory is also relatively faster than DDR3, it’s not something that’ll have a notable impact on performance, at least not in the current PC landscape.


With that out of the way, lets talk about GDDR5 memory, the predominant memory standard on video cards (although now it’s being phased out). Both GDDR5 and GDDR4 are based on DDR3 memory. So while the core specifications are the same, there are three principle differences between GDDR5 memory and DDR3.


  • DDR3 runs at a much higher voltage than GDDR5, 1.25-1.65 volts to be exact. GDDR5 on the other hand is usually limited to 1V
  • Both DDR4 and DDR3 use a 64-bit memory controller per channel which results in a 128 bit bus for dual channel memory and 256 bit for quad. GDDR5 memory on the other hand leverages a puny 32 bit controller per channel
  • However, while CPU memory configurations have faster per channel transactions, GPUs can support any number of 32 bit memory channels. This is the reason many high end GPUs like the GeForce GTX 1080 Ti and GTX 1080 have a 384 bit and 256 bit bus width, respectively. 

Both these cards are connected to 1GB memory chips via 8 (for 1080) and 12 (for the Ti) 32 bit memory controllers or channels. GDDR5 can also operate in what is called clamshell mode, where each channel instead of being connected to one memory chip is split between two. This also allows manufacturers to double the memory capacity and makes hybrid memory configurations like the GTX 660 with it’s 192 bit bus width possible.

  • Another core difference between DDR3 and GDDR5 memories involves the I/O cycles. Just like SATA interface, DDR3 can only perform one operation (read or write) in one cycle. GDDR5 can handle input (read) as well as output (write) on the same cycle, essentially doubling the bus.
  • All this might put DDR3/4 memory in a bad light, but this configuration actually suits it. CPUs are heavily sequential when it comes to their workloads and GPUs on the other hand rely on parallel processing. So while CPUs benefits from low latency, GPUs require a much higher bandwidth with loose timings.

High Bandwidth Memory (HBM)

First popularized by AMD Fiji graphics cards, high bandwidth memory or HBM is a low power memory standard with a wide bus. HBM achieves substantially higher bandwidth compared to GDDR5 while drawing much lesser power in a small form factor.


HBM adopts clocks as low as 500 Mhz to conform to a low TDP target and makes up for the loss in bandwidth with a massive bus (usually 4096 bits). AMD’s Radeon RX Vega cards are the best example of HBM2 implementation in consumer hardware. HBM2 solved the 4GB limit of the HBM1, but limited yields coupled with memory shortage prevented AMD from capitalizing on the GPU front.  


GDDR5X is a half generation upgrade of sorts. It increases the bandwidth of GDDR5 by increasing the base clock and doubling the prefetch (operations per clock). GDDR5X increases the bandwidth to upto 12 Gbps and also brings down the voltage by a bit.


As far as GDDR6 is concerned, the specifications are still murky but we do know the targeted bandwidth and voltages. GDDR6 is expected to hit a record bandwidth of 16 Gbps while cutting down the voltage by 10% compared to GDDR5. NVIDIA’s upcoming Turing graphics cards are expected to use this new memory and hit bandwidths of upto 768 GB/s.


LPDDR4 is the mobile equivalent of DDR4 memory. Compared to DDR4, it offers reduced power consumption but does so at the cost of bandwidth. LPDDR4 has dual 16 bit channels resulting in a 32 bit total bus, in comparison DDR4 has an 8 word prefetch or a 64 bit channel. Therefore, LPDDR4 RAM halves the bus but makes up for this with a measly operating voltage of 1.1-1.2V.


This allows for a greater power efficiency in smartphones and battery standby times of 8-10 hours. Micron’s LPDDR4 RAM tops out the standard with a 2133 MHz clock for a transfer rate of 4266 MT/s while Samsung follows shortly after with a clock of 1600MHz and a transfer rate of 3200 MT/s.

So that’s most of what you need to know about RAM or dynamic memory. If you have any doubts or think we missed something, do let us know in the comments section below!


Leave a Reply