Definition of GDDR [ 2022 Guide ]

What is GDDR? It’s a type of SDRAM or Synchronous Dynamic Random-Access Memory. That means it synchronizes the input and output of data on both clock edges to reduce latency. This, in turn, reduces any delays that might occur when sending commands to your graphics card for processing.

The acronym stands for Graphics Double Data Rate since it operates at twice the speed as traditional DDR memory (DDR). Now that you know what it does and how it works, let’s talk about where you can find this type of RAM.

The GDDR generation of GDDR memory is constantly getting faster and more advanced; however, GDDR types do not correspond numerically to DDR. For example, GDDR (Generation Data Rate) Memory Version three was based on DDR (Double Date Rate) Memory version two chips while GDDR five used the newer GDDR four technology.

It’s important to keep in mind that GDDR generations are very different from their corresponding traditional RAM modules since they’re built with different architecture. So if you ever need some for your custom rig or upgrade project be sure to check out our inventory here at Crucial Technology where we offer a wide range of compatible options in both DDR and GDDR technologies.

GDDR Types





  • GDDR V



The GDDR SGRAM is GDDR version one. It was built by the same company that created GDDR, Micron Technology. The first GDDR memory was similar to GDDR III but didn’t offer much in terms of performance. GDDR I uses vertical four-transistor (four-Tr) cells and is, therefore, slower than GDDR II’s three transistors per cell architecture.


The next generation consisted of two key architectures; the first was called GDDR II which offered higher data rates using faster clock speeds at a lower voltage setting with reduced signal integrity issues that caused problems with earlier generations.

The second part of this generation came from Rambus Incorporated who devised a new type of SGRAM chip for their own graphics cards known as XDR DRAM or eXtended Data Rate Dynamic RAM. In order to counter any attacks on its patents, Rambus legally enforced its patents on GDDR II memory to be used only with XDR DRAM. This meant that GDDR III could no longer use GDDR II chips which led to a new generation being developed in order for GDDR technology to progress once again.


The introduction of DDR was an incredible development in the world of computer components because it allowed RAM modules and graphics cards alike to become faster, more efficient, and develop a lower latency rate without sacrificing voltage or adding additional heat onto their circuits.

While there were some early problems from both sides when they first began production GDDR III still managed to offer greater performance than any preceding memory types due in part to the increased bandwidth offered by this type’s higher data rates over even GDDR II.


GDDR IV was not as much of an improvement over GDDR III as previous generations but it still managed to offer some significant changes in terms of its bandwidth capabilities and data transfer rates which were more than double that of GDDR III.

It also used the same chip technology as GDDR III, however, this time they kept process sizes smaller at 0.18 µm instead of 0.25 µm because memory became so dense by shrinking their design down even further from what had been done with GDDR II’s four transistor cell architecture previously.

This allowed advanced graphics cards using this type to have a wider bus interface allowing them to use lower clock speeds without sacrificing performance or stability due to latency issues caused by increased voltage demands on GDDR III. GDDR IV also supported up to four GDDR III SGRAM chips which allowed it an even greater advantage over its predecessor types if the manufacturer was willing to pay more for them since they were so expensive at first due to low yields on production processes early on.


GDDR V brought many changes in terms of the architecture used by this type of memory, however, one major change that stands out is that GDDR V abandoned XDR DRAM entirely and went back to standard DDR technology instead.

This meant that manufacturers no longer had any need for Rambus Incorporated’s patents because now anyone could make use of both GDDR III and GDDR IV architectural designs without having their products blocked from further development or added costs associated with licensing fees. GDDR V also increased power supply voltages from two volts to three and managed data rates as high as 1250 MHz which is more than double that of GDDR III’s capabilities at 550 MHz.


GDDR VI was the final generation before GDDR became obsolete for use in modern graphics cards due to their focus on developing other memory technologies such as HBM or Wide I/O interfaces and this marked a significant change because it brought back XDR DRAM instead of DDR technology as its predecessor types did.

This time, however, they used faster clock speeds up to 2600MHz so it could match GDDR V’s performance levels while still offering greater capacity with 512MB per chip compared to 256MB previously latencies were much lower than GDDR V as well. GDDR VI also increased speeds to 16 Gbps which was double that of GDDR IV and it managed this despite increasing the number of I/O pins by 50% so they now stood at 340 or 370 depending on whether the chip was single-sided or not.

It is important to note, however, that GDDR VI only offered these advantages over its predecessor if manufacturers wanted their products using this type installed into server systems because GDDR VI has never been used in consumer graphics cards before development stopped for good shortly after its introduction where GDDR became obsolete entirely due to new emerging technologies taking priority instead.