Skip to main content

The future of fast PC graphics? Connecting directly to SSDs

Performance boosts are expected with each new generation of the best graphics cards, but it seems that Nvidia and IBM have their sights set on greater changes.

The companies teamed up to work on Big accelerator Memory (BaM), a technology that involves connecting graphics cards directly to superfast SSDs. This could result in larger GPU memory capacity and faster bandwidth while limiting the involvement of the CPU.

A chart breaks down Nvidia and IBM's BaM technology.
Image source: Arxiv Image used with permission by copyright holder

This type of technology has already been thought of, and worked on, in the past. Microsoft’s DirectStorage application programming interface (API) works in a somewhat similar way, improving data transfers between the GPU and the SSD. However, this relies on external software, only applies to games, and only works on Windows. Nvidia and IBM researchers are working together on a solution that removes the need for a proprietary API while still connecting GPUs to SSDs.

The method, amusingly referred to as BaM, was described in a paper written by the team that designed it. Connecting a GPU directly to an SSD would provide a performance boost that could prove to be viable, especially for resource-heavy tasks such as machine learning. As such, it would mostly be used in professional high-performance computing (HPC) scenarios.

Get your weekly teardown of the tech behind PC gaming
Check your inbox!

The technology that is currently available for processing such heavy workloads requires the graphics card to rely on large amounts of special-purpose memory, such as HBM2, or to be provided with efficient access to SSD storage. Considering that datasets are only growing in size, it’s important to optimize the connection between the GPU and storage in order to allow for efficient data transfers. This is where BaM comes in.

“BaM mitigates the I/O traffic amplification by enabling the GPU threads to read or write small amounts of data on-demand, as determined by the compute,” said the researchers in their paper, first cited by The Register. “The goal of BaM is to extend GPU memory capacity and enhance the effective storage access bandwidth while providing high-level abstractions for the GPU threads to easily make on-demand, fine-grain access to massive data structures in the extended memory hierarchy.”

An Nvidia GPU core sits on a table.
Niels Broekhuijsen / Digital Trends

For many people who don’t work directly with this subject, the details may seem complicated, but the gist of it is that Nvidia wants to rely less on the processor and connect directly to the source of the data. This would both make the process more efficient and free up the CPU, making the graphics card much more self-sufficient. The researchers claim that this design would be able to compete with DRAM-based solutions while remaining cheaper to implement.

Although Nvidia and IBM are undoubtedly breaking new ground with their BaM technology, AMD worked in this area first: In 2016, it unveiled the Radeon Pro SSG, a workstation GPU with integrated M.2 SSDs. However, the Radeon Pro SSG was intended to be strictly a graphics solution, and Nvidia is taking it a few steps further, aiming to deal with complex and heavy compute workloads.

The team working on BaM plans to release the details of their software and hardware optimization as open source, allowing others to build on their findings. There is no mention as to when, if ever, BaM might find itself implemented in future Nvidia products.

Editors' Recommendations

Monica J. White
Monica is a UK-based freelance writer and self-proclaimed geek. A firm believer in the "PC building is just like expensive…
5 GPUs you should buy instead of the RTX 4070
RTX 4070 logo on a graphics card.

Nvidia's RTX 4070 is one of the best graphics cards you can buy, make no mistake about that. Some recent price drops, combined with excellent 1440p performance and features like DLSS 3.5, make it the go-to GPU for a high-end gaming experience in 2024. There are several other GPUs to keep in mind around this price, however.

The market around for graphics cards that cost $500 to $600 is hotly contested among AMD and Nvidia, and there are some other excellent options to keep in mind when shopping for a new GPU. Here are five GPUs to consider if you're in the market for the RTX 4070.
Nvidia RTX 4070 Super

Read more
What is VSync, and why do you need it?
HP Omen 40L Gaming PC on a table connected to a monitor.

If you’ve been playing PC games for a number of years, you’ve probably heard the term ‘VSync’ tossed around once or twice. Maybe you’ve also heard of G-Sync and FreeSync. For those unaware, VSync is actually short for ‘vertical synchronization’. This is a display feature that is designed to keep your gaming screen running in sync with your computer's GPU. VSync isn’t just important for PC gaming, but it’s one of the most important criteria that goes into a good gaming display.

In this article, we’re going to take a closer look at VSync (and its related technologies) to find out exactly how it works, if you should have it enabled, and how to disable it if you don’t like the optimization. 
What is VSync technology?

Read more
How 8GB VRAM GPUs could be made viable again
Screenshot of full ray tracing in Cyberpunk 2077.

Perhaps there is still some hope for GPUs with low VRAM. According to a new patent published by Microsoft, the company worked out a method that could make ray tracing and path tracing more viable in terms of how much video memory (VRAM) they use. As of right now, without using upscaling techniques, seamless ray tracing requires the use of one of the best graphics cards—but this might finally change if this new method works out as planned.

This new patent, first spotted by Tom's Hardware, describes how Microsoft hopes to reduce the impact of ray tracing on GPU memory. It addresses the level of detail (LOD) philosophy, which is already something that's used in games but not in relation to ray tracing, and plans to use LOD to adjust ray tracing quality dynamically, thus lowering the load that the GPU -- particularly its memory -- has to bear.

Read more