GIGABYTE Aorus 12000 Specifications

GIGABYTE Aorus 12000 Performance

For this review, we will be looking at the 1TB model. Comparables are a mixture of popular Gen4 SSDs and the first wave of Gen5 SSDs. The other tested Gen5 SSDs are equipped with the Phison E26 platform:

• Samsung 990 Pro
• ADATA Legen 970
• Seagate FireCuda 540
• Gigabyte Aorus 10000
• Solidigm P44 Pro

For testing, we use two platforms. The consumer test platform supports PCIe Gen4/Gen5 SSDs and is sometimes leveraged for lighter consumer-based tests such as BlackMagic DiskSpeed Test and CrystalDiskMark, and our main platform, a Dell PowerEdge R760, which overlaps with our enterprise tests. For ultimate flexibility, we worked with Serial Cables, who supplied us with an 8-bay PCIe Gen5 JBOF for U.2/U.3, M.2, and E1.S/E3.S drive testing. This allows us to test all current and emerging drive types on the same test hardware.

Dell PowerEdge R760 Configuration

• Dual Intel Xeon Gold 6430 (32 cores/64 threads, 1.9GHz base)
• 1TB DDR5 RAM
• Ubuntu 22.04

VDBench Workload Analysis

When benchmarking storage devices, application testing is best, and synthetic testing is second. While not a perfect representation of actual workloads, synthetic tests help baseline storage devices with a repeatability factor that makes it easy to compare competing solutions. These workloads offer a range of testing profiles ranging from “four corners” tests and common database transfer size tests to trace captures from different VDI environments.

These tests leverage the common vdBench workload generator, with a scripting engine to automate and capture results over a large compute testing cluster. This allows us to repeat the same workloads across various storage devices, including flash arrays and individual storage devices. Our testing process for these benchmarks fills the entire drive surface with data, then partitions a drive section equal to one percent of the drive capacity to simulate how the drive might respond to application workloads. This differs from full entropy tests, which use 100 percent of the drive and take them to a steady state. As a result, these figures will reflect higher-sustained write speeds.

Profiles:

• 4K Random Read: 100% Read, 128 threads, 0-120% iorate
• 4K Random Write: 100% Write, 64 threads, 0-120% iorate
• 64K Sequential Read: 100% Read, 16 threads, 0-120% iorate
• 64K Sequential Write: 100% Write, 8 threads, 0-120% iorate
• VDI Profiles

Starting with 4K random read, the GIGABYTE Aorus 12000 Gen5 drive was far behind the leaders and close to its predecessor in performance, hitting 770K IOPS at 165.5ms. The Gen4 Solidigm P44 Pro SSD was the best-performingdrive here.

The Aorus 12000 Gen5 SSD faired the same when it came to writes, taking second last only surpassing the Solidigm drive. Here, it peaked with a speed of 264k IOPS and a latency of 198.8µs, which was slower than its older Aorus counterpart.

When switching to the sequential read 64k workload, the Aorus 12000 performed moderately better, placing 3^rd overall. It ended the test at a speed of 5.77GB/s and a latency of 346.1µs.

Unfortunately, the Aorus 12000 fell way back to last place in sequential writes. The drive peaked at just a hair over 1GB/s at a latency of over 1,000µs at the end of the test. Its predecessor performed much better with 1.4GB/s.

Next, we looked at our VDI benchmarks, designed to tax the drives further. These tests include Boot, Initial Login, and Monday Login. Starting with boot, the Aorus 12000 had fairly weak results, peaking at around 140K IOPS at 236.1ms.

Next up is VDI Initial Login. Though the Aorus 12000 wasn’t nearly as bad as the Solidigm P44 Pro, it still had noticeably weaker results than its predecessor posting just 50K IOPS at 603.3ms in latency.

Lastly, we have the VDI Monday Login benchmark, which actually showed some pretty solid results. Taking third place, the Aorus 12000 peaked at 41K IOPS at a latency of 388.7ms.

BootBench

BOOT-BENCH-1 is a workload profile adopted by OCP to profile SSDs designed for server boot duty. While this is intuitively a job for enterprise SSDs, client SSDs are often selected for their performance, capacity, and cost combination. The boot drive issue is germane not just to hyperscalers but also to server and storage system providers, as they face similar challenges.

This boot workload executes a relatively intense test plan that fills the entire drive with writes before testing a read-heavy workload sequence. For each test, it performs a 32K random read async operation alongside a 15MiB/s synchronous 128k random write and a 5MiB/s synchronous 128k random write/trim background workload. The script starts with the random-read activity at a 4-job level and scales up to 256-jobs at its peak. The final result is the read-operations performed during its peak run.

The OCP goal for this benchmark is a pass/fail at 60K read IOPS. Most drives we test will far exceed the minimum, but the results are instructive regardless.

Unfortunately, like its predecessor, the Arorus 12000 is another SSD to get a “DNF” due to its poor latency and IOPS that did not make the cut-off. To date, many of the Phison E26 Gen5 consumer SSDs have failed this test, with a couple barely passing.

Here is the current boot leaderboard:

CrystalDiskMark Speed Test

We performed a lighter-weight test on the GIGABYTE Aorus 12000 to demonstrate its Gen5 speeds. Using CrystalDiskMark, the drive recorded sequential transfer speeds exceeding 10GB/s for both read and write, showing 11.7GB/s for reads and 9.5GB/s for writes. Interestingly, the CrystalDiskMark results aligned precisely with what was outlined in the specification sheet; a rare coincidence that, while not particularly meaningful, is something we seldom see.

CrystalDiskMark’s higher queue depth (compared to BlackMagic) allows us to showcase the best-case scenario for the drive. These test results among comparable drives were so closely matched that any real-world performance differences would likely be negligible.

Blackmagic Disk Speed Test

We measured performance inside a Windows 11 environment on our consumer test platform via the popular Blackmagic test. Here, Aorus 12000 was able to hit 7.5GB/s read (higher than the average) and 9.1GB/s write (lower than the average).

Conclusion

The AORUS 12000 Gen5 SSD comes as an iteration of its predecessor, the AORUS 10000, with Micron’s new 232-layer NAND standing out as its most significant upgrade. It also boasts a Mean Time Between Failures (MTBF) rating of 1.6 million hours and adopts the standard M.2 2280 form factor. Ultimately, while this drive is priced competitively, its performance fails to live up to the expectations set by its PCIe Gen5 interface. For users seeking a budget-friendly Gen5 SSD, the AORUS 12000 Gen5 might be an appealing option, but those who prioritize performance will likely find better value in other models.

 

 

When it comes to actual performance metrics, the AORUS 12000 fell behind in most of our benchmark tests, including 4K random read/write evaluations and VDI benchmarks. Results in these categories were not only lower than those of competing drives but also frequently lagged behind its own predecessor. This outcome is not entirely surprising, though: this lack of a competitive edge is consistent across other Gen5 SSDs equipped with the same controller—they still struggle to outperform high-end Gen4 SSDs.

 

It’s important to note that we tested the 1TB model of the AORUS 12000 for this review. In contrast, we benchmarked the 2TB variant of the AORUS 10000, and all other drives included in our comparisons were also 2TB models. This difference in storage capacity could potentially have impacted the performance metrics, as SSDs often display varying performance traits based on their capacity—due to differences in NAND configuration and other related factors. As such, we might have observed slightly different results if GIGABYTE had provided the 2TB version for benchmarking.

 

Nonetheless, while the AORUS 12000 may hold some appeal with its optional thermal guard heatsink and sturdy build quality, it largely amounts to a disappointing update over the AORUS 10000 rather than a meaningful step forward. At its current price point, it does undercut the FireCuda 540 (priced at $180), which uses older NAND technology, but the cost savings are not substantial enough to offset its relative performance shortcomings. This makes it difficult to recommend over more powerful alternatives—or even its own predecessor, the AORUS 10000—for both general consumers and performance enthusiasts alike.