This month we introduce new benchmarks based on Futuremark's PCMark 8 storage suite. Most illuminating, perhaps, is that the data we generate shows that there's probably little reason to upgrade if you already own a modern SATA 6Gb/s-attached SSD.
Detailed solid-state drive specifications and reviews are great—that is, if you have the time to do the research. However, at the end of the day, what an enthusiast needs is the best SSD within a certain budget.
So, if you don’t have the time to read the benchmarks, or if you don’t feel confident enough in your ability to pick the right drive, then fear not. We at Tom’s Hardware have come to your aid with a simple list of the best SSD offered for the money.
April Updates: Our Benchmark Suite Evolves
The first time you swap an SSD in to replace a hard drive, there's a sense of euphoria as it dawns on you that this is what enthusiast-oriented computing should always be like. Where have these magical little devices been your whole life? But eventually you get used to that speed, and PCs powered by mechanical storage simply become infuriatingly slow. At that point, you swap between SATA 6Gb/s-capable SSDs and can't tell the difference. They all feel fast, and most do in fact deliver comparable performance.
There are good reasons why this is the case. Simply adding an SSD to a modern machine is enough to shift its principal performance bottleneck from storage to something else. Exceptions do exist; there were some truly bad solid-state drives before the technology really matured. Generally speaking, though, the perceptible differences between two good drives in 2014 are minor. Would you notice if Photoshop fired up in 1.33 seconds rather than 1.45? Probably not. And according to our benchmarks, just about every SSD released in the past year can boot Windows, run productivity applications, or load Skyrim in practically the same amount of time.
The following chart of Futuremark's PCMark 8 trace-based storage suite illustrates this using three of the metric's 10 workloads, including World of Warcraft, a session of Battlefield 3, and a light Adobe Photoshop photo manipulation workload. I tested nine drives and arranged them from fastest to slowest. As you can see, eight of them don't reflect much variance at all. And one of those eight includes four SSD S3500s in RAID 0. That's 2 TB of flash capable of pumping out 4x the performance of one SSD...in theory. But it finishes PCMark 8 no faster than the other seven data points that aren't SanDisk's U110.
In short, regardless of your storage device's potential, there is a point where it doesn't make sense to connect anything faster to the SATA 6Gb/s interface. Does that mean SSD performance isn't important to measure anymore? Of course not. It's just that these real-world workloads don't expose the benefits of slightly-faster storage hardware, particularly without any other operations happening in parallel.
Does this mean SSD performance doesn't really matter anymore? Absolutely not. It just doesn't much matter for those specific workloads, especially without anything else occurring in parallel. It takes a lot of concurrent storage activity to expose the advantages of one drive over another. Browsing the Web, sending email, and watching YouTube isn't hardcore enough. As your usage intensifies, though, I/Os take longer to service. Inside, your SSD is programming, reading, and erasing, all of which affect speed. While all of that is happening, Windows might decide to ask the drive to TRIM some LBAs, further impacting performance.
In order to better understand the implications of heavy drive activity with the same workloads, we need to introduce additional testing.
In my last SSD review, I added another PCMark 8-based test that characterizes performance using the same collection of traces played 18 times in a row. Namely, the first 13 iterations employ punishing random writes. The last five runs are preceded by five-minute resting periods. That's the only time during the day-long benchmark when nothing happens. A ton of data gets generated during this process. So to keep things simple, I follow just one trace (Adobe Photoshop [Heavy]) through its 18-round bout. Of PCMark's workloads, this one takes the longest to run and is the most write-heavy.
Each drive starts in a more degraded state than it would during a normal PCMark 8 run. That's a deep hole from which the SSDs cannot escape until rest periods are introduced between the Recovery Phase rounds. Adding five minutes of idle time is a big deal, explaining why the MB/s figure increases in three of my four test cases. A quick glance makes it easy to separate the products able to recover from taxing work and the ones that don't.
Our new tests appear complicated, but I'm working to simplify them as much as possible. Then again, SSDs are generally complex beasts. Some of their inner workings are discussed freely, while other aspects are black boxes we can only evaluate after hundreds of hours of benchmarking. Their behavior changes from one workload to the next, which is why we're cautious about making generalizations. Performance remains an important consideration, though we'll keep reminding you that it's only one criteria of reviewing SSDs.
Storage Reviews For The Month:
Some Notes About Our Recommendations
A few simple guidelines to keep in mind when reading this list:
- We only recommend SSDs we've actually used. Recommending SSDs we've never put hands on wouldn't be incredibly helpful.
- There are several criteria we use to rank SSDs. We try to evenly weigh performance and capacity at each price point and recommend what we believe to the best drive based on our own experiences, along with information garnered from other sites. Some people may only be concerned with performance, but that ignores the ever-present capacity issue that mobile users face ever-presently. Even on the desktop, other variables have to be considered.
- Prices and availability change on a daily basis. Our picks will be valid the month of publication, but we can't extend our choices very far beyond that time frame. SSD pricing is especially competitive, and a $15 difference can be the reason why one SSD makes the list, while another does not. As you shop, use our list as a guide, but always double-check for yourself.
- The list is based on some of the best U.S. prices from online retailers. In other countries or at retail stores, your mileage will most certainly vary.
- These are new SSD prices. No used or open-box offers are in the list; they might represent a good deal, but it’s outside the scope of what we’re trying to do.