When you're looking for a new component for your PC, you want to make sure that you at least understand most of the features that it's going to bring to your machine.
While some like the memory and GPU speed for a graphics card can be easily compared, other features such as its memory bandwidth and SLI compatibility can be more of a challenge.
This is where NVIDIA's 'Resizable BAR' comes in; while a standard since 2008, it's recently appeared again due to the company announcing support for it to its 30 Series of GPU's.
With that, here's what 'Resizable BAR' is, alongside which GPU's its available for.
NVIDIA Resizable BAR Explained
First; we need to go to their competitor, AMD, where it was rumoured that an 'Infinity Cache' feature would be on their 6000 Radeon GPU line.
This was confirmed to be the 'AMD Smart Access Memory' feature, which allows the CPU to access the 'virtual memory' directly and now in 256MB chunks.
This is the same as 'Resizable BAR', but it's only come into prominence in recent years, due to more graphically-intensive games requiring more powerful hardware and methods to render its environments at full-speed.
The feature allows the CPU and the GPU to directly communicate with one another in a direct 'lane', that removes the 256MB buffer chunk in one go.
Which Graphics Cards is BAR on?
NVIDIA announced this week at CES that the 30 Series of their graphic cards will fully support the 'Resizable BAR' feature by the Summer.
This will be possible by both a firmware update for the GPU and a BIOS update to Motherboards.
As this feature requires other components for it to work, the motherboard needs to know that this feature can access the memory lane, so this may take time across several manufacturers.
It should be noted, that it may be some time before benchmarks are released that show the performance improvements here, but it could bring as much as 10% in an increase in framerates.