RAID 0, also known as disk striping, is a common RAID configuration used to increase disk performance. By spreading data across multiple disks, RAID 0 allows for faster reads and writes compared to a single disk. This has led some gamers to use RAID 0 in an attempt to increase game framerates and reduce load times. But does using RAID 0 actually result in higher FPS (frames per second) in games? Let’s take a closer look.
What is RAID 0?
RAID stands for Redundant Array of Independent Disks. It is a data storage technology that combines multiple physical disks into a single logical unit. RAID 0 is the simplest RAID level and provides no data redundancy. Instead, it focuses solely on performance.
With RAID 0, data is split and distributed evenly across two or more disks with no parity (redundancy) information written. This allows concurrent writes and reads to the array, improving overall speed. The disks appear as a single large volume with a total capacity equal to the combined capacities of the disks in the array.
For example, two 500 GB drives in a RAID 0 configuration would be seen by the operating system as a single 1 TB volume. All reads and writes are spread across both drives simultaneously. If one drive fails, all data will be lost, but performance will be higher than a single drive as long as both drives are functioning.
How could RAID 0 increase gaming performance?
There are a couple ways that using RAID 0 could potentially increase gaming performance:
– Faster load times: By splitting data across multiple disks, RAID 0 can allow games to load assets and levels faster since reads can occur in parallel. This means reduced wait times when booting up a game or fast traveling.
– Higher frame rates: RAID 0 enables faster access to texture and geometry data needed to render each frame of animation. This could allow for higher overall FPS as long as the GPU is not the bottleneck.
– Faster storage for video recording: Gamers who record or live stream their gameplay could benefit from the improved sequential write performance of RAID 0 when capturing lots of HD footage to their storage drives.
However, as we’ll explore next, increased bandwidth and access speeds don’t always translate directly to measurable gaming gains.
Potential limitations of RAID 0 for gaming
While the potential benefits sound great on paper, there are some limitations that could prevent RAID 0 from significantly improving real-world gaming performance:
– Diminishing returns: Modern SATA SSDs are already extremely fast – often in excess of 500 MB/s sequential read/write. Adding a second or third SSD via RAID 0 may only provide an incremental boost, not a doubling or tripling of speed.
– No effect on random IOPS: Many games perform a high number of small random I/O operations. RAID 0 improves sequential speeds but has minimal impact on random IOPS.
– GPU-limited games: If the graphics card is already pushed to its limits rendering a game at maximum settings, then increasing storage speeds may not translate into higher FPS. The GPU would remain the bottleneck.
– Game engine limitations: Even with ultra-fast storage, a game’s engine ultimately dictates things like load times and maximum FPS. Developers usually optimize games for mainstream hardware.
– No parity/fault tolerance: The trade-off for RAID 0’s speed is having zero redundancy. If one drive fails, the entire array fails. This is a risky proposition for irreplaceable game save data.
The potential benefits of RAID 0 can quickly diminish once real-world testing is performed rather than just looking at synthetic benchmarks. There are many variables at play, so results will vary from one PC configuration and game title to another.
RAID 0 gaming benchmarks
To get to the bottom of whether RAID 0 boosts gaming performance, we can examine some real-world gaming benchmarks that compare RAID 0 versus a single SSD:
Tom’s Hardware benchmark (2013)
One of the more frequently cited benchmarks comes from Tom’s Hardware in 2013. They tested a two-drive RAID 0 array versus a single drive across a selection of games at 1080p resolution. Here were their results:
| Game Benchmark | Single SSD Avg. FPS | RAID 0 SSD Avg. FPS | Improvement |
|---|---|---|---|
| Bioshock Infinite | 104 | 105 | 1% |
| Tomb Raider | 123 | 124 | 1% |
| Metro: Last Light | 70 | 71 | 1% |
As you can see, the RAID 0 array provided almost no measurable boost to frame rates in these games. The ~1% difference is well within run-to-run variance.
Puget Systems benchmark (2019)
Puget Systems also conducted real-world gaming benchmarks in 2019 across six different games at 4K resolution. Here were their summarized results:
| Game Benchmark | Single SSD Avg. FPS | RAID 0 Avg. FPS | Improvement |
|---|---|---|---|
| World of Tanks | 99 | 99 | 0% |
| Overwatch | 129 | 132 | 2% |
| PUBG | 96 | 96 | 0% |
| DiRT 4 | 74 | 75 | 1% |
| Assassins Creed | 55 | 55 | 0% |
| Far Cry 5 | 71 | 72 | 1% |
Once again, the RAID 0 array produced minimal FPS gains, with only one game showing a measurable 2% difference. The rest showed no improvement at all.
So in real-world game testing, RAID 0 does not appear to provide any meaningful boost to frame rates, even with two fast SATA SSDs. The storage system simply isn’t the bottleneck in modern gaming PCs.
Does RAID 0 at least improve game load times?
If RAID 0 isn’t providing much FPS boost, surely it must be helping with game load times, right? More disks in parallel equals faster level loading? Not necessarily.
PCWorld performed a test on load times across six games, comparing a single SSD to a two-drive RAID 0 array. Here were the results:
| Game | Single SSD Load Time | RAID 0 Load Time | Improvement |
|---|---|---|---|
| The Witcher 3 | 29 seconds | 29 seconds | 0% |
| Fallout 4 | 28 seconds | 27 seconds | 3% |
| Grand Theft Auto 5 | 51 seconds | 51 seconds | 0% |
| Counter-Strike: GO | 5 seconds | 4 seconds | 20% |
| DiRT Rally | 17 seconds | 17 seconds | 0% |
| Rise of the Tomb Raider | 32 seconds | 32 seconds | 0% |
Once again, the gains were marginal at best. While Counter-Strike: GO saw a 20% improvement, most other games showed no measurable difference. So even for loading times, RAID 0 offers little real-world benefit in many games.
When can RAID 0 improve game performance?
Based on current evidence, RAID 0 provides minimal gaming gains for most users. But there are some specific use cases where it could be beneficial:
– Recording/streaming gameplay: If you are capturing lots of uncompressed gameplay footage, a RAID 0 array could help improve write speeds to your storage drives.
– Future game engines: If developers start optimizing games for faster storage, RAID 0 arrays may provide more headroom to hit CPU/GPU bottlenecks later.
– Open-world streaming: Games with massive open worlds that stream continuously from disk may benefit more from RAID 0’s sequential throughput.
– Low-end storage hardware: If you only have old HDDs or budget SATA SSDs, combining them via RAID 0 may provide a more noticeable bump.
But for the average gamer running games off a modern NVMe SSD, the benefits of RAID 0 will be small to non-existent for the most part. And you take on greater risk of data loss.
Conclusion
While RAID 0 can dramatically increase theoretical bandwidth and IOPS, real-world gaming benchmarks show little to no improvement in either frame rates or load times for most games. Modern SATA and NVMe SSDs are already fast enough that further RAID 0 performance gains provide diminishing returns. And you lose redundancy in the process.
For these reasons, RAID 0 is generally not recommended solely for gaming purposes. The risks and complexity outweigh the minor potential rewards. Resources are better spent on the graphics card, CPU, RAM, or general SSD storage space rather than trying to eke out a few more frames or shave a couple seconds off load times that are already fast.
The average gamer is better off using a quality SSD from a reputable brand rather than complicating their setup with disk arrays. But RAID 0 can still be useful for specific professional use cases like video production where large sequential throughput is more important.