AMD blocks overclocking of HBM memory with Radeon “Fiji”

The upcoming Radeon R9 Fury X will be the first in the world to use memory technology High-Bandwidth Memory (HBM) from SK Hynix. The technology is seen as the successor to the current GDDR5 and comes with several advantages, where AMD mainly emphasizes that the memory circuits’ close integration with the graphics processor takes up less space on the circuit board and provides lower power consumption.

Another advantage of HBM is the potential for significantly higher bandwidth compared to GDDR5. Through a 4,096-bit memory bus and HBM clocked at 500 MHz (1,000 MHz efficient), the AMD Radeon R9 Fury X gets a bandwidth of a breathtaking 512 GB / s, compared to 384 GB / s for the new Radeon R9 390X / 390 and 336 GB / s for Nvidia’s flagship models Geforce GTX Titan X and GTX 980 Ti.

The fresh technology, however, is drawn with a clear disadvantage; AMD’s next flagship gets only 4 GB of video memory. Now it also appears that the company is locking down overclocking of the memory for all Fiji-based graphics cards. This is attributed to the fact that the technology is still new and that the memory bandwidth is already considered to be more than sufficient.

It is worth mentioning that the latch should not be in the hardware without AMD’s drivers, which means that overclocking of the HBM memory may be possible later in the future. Initially, however, only the graphics processor’s clock frequency will be able to be tuned by end users.

Source: Legitreviews.