The Mushkin Source 500 GB SSD is currently one of the most affordable 500 GB drives on the market due to the company's choice to skip the DRAM chip, which enables the cost savings required to reach such a price point. On previous pages, we talked extensively about the theory behind DRAM-less designs, which basically trade some amount of performance for lower manufacturing cost.

Generally, and the Mushkin Source is no exception here, the lack of DRAM mostly affects performance of random writes, i.e. writing small data chunks spread out over a large area of the disk (multiple GB). When benched with our regular test area size of 128 GB, we saw pathetic numbers (that, of course, are still much better than with a HDD) that are nowhere near what we've come to expect from modern SSDs. After some consideration, we added a second data point for these results, testing at a smaller size of 16 GB, which should be more in-line with what the Mushkin Source will see in actual usage. At this smaller test size, the SSD does a lot better, but the synthetic results are still not as fast as competing drives with DRAM. Nearly all of today's write-intensive consumer applications fall into just two categories: large sequential writes (ISO/game/disc image/video copy, large software installation) and small random writes (small software installation, extracting or copying lots of small files). Sequential performance of the Mushkin Source is excellent, pretty much identical to more expensive drives with DRAM. The small random writes are what trouble it, but they usually are localized to a small subset of the total disk space, which is why our real-life testing shows results that are "close enough" to competing drives, especially when you consider the lower cost of the Source. Some applications do write to larger disk areas, but are constrained by the rate at which they can generate data (for example, BitTorrent downloads). In the enterprise world, things are different, of course. Large databases with many concurrent writers are the norm, so are busy file servers with many connected clients, for which the Mushkin Source is definitely not an option.

On average, in real-life, we see the Mushkin Source about 5% slower than high-end 2.5" SATA drives, which is not a lot. The top-end NVMe drives in our test group are up to 20% faster. Individual applications will see different results, but overall, the numbers are much better than what the synthetic testing results would suggest. What really matters for the Mushkin Source is its price. With just $81 for the tested 500 GB variant, it is easily 10%-20% cheaper than competing drives. While that's "just" 15 dollars or so, it could still factor into your buying decision, especially when building a rig where performance isn't that important; a media PC or system for your parents that will only see light usage anyway, for example. The low price per GB could also lure in buyers that are looking to upgrade an older HDD-based system on a budget. The biggest competition for the Mushkin Source is without any doubt the Crucial MX500, which offers better performance more consistently and in more edge cases, for just a little extra cost.