The Intel SSD 600p (512GB) Review

Intel’s SSD 600p was the first PCIe SSD using TLC NAND to hit the consumer market. It is Intel’s first consumer SSD with 3D NAND and it is by far the most affordable NVMe SSD: current pricing is on par with mid-range SATA SSDs. While most other consumer PCIe SSDs have been enthusiast-oriented products aiming to deliver the highest performance possible, the Intel 600p merely attempts to break the speed limits of SATA without breaking the bank.

The Intel SSD 600p has almost nothing in common with Intel’s previous NVMe SSD for consumers (the Intel SSD 750). Where the Intel SSD 750 uses Intel’s in-house enterprise SSD controller with consumer-oriented firmware, the Intel 600p uses a third-party controller. The SSD 600p is a M.2 PCIe SSD with peak power consumption only slightly higher than the SSD 750’s idle. By comparison, the Intel SSD 750 is a high power and high performance drive that comes in PCIe expansion card and 2.5″ U.2 form factors, both with sizable heatsinks. 

The Intel SSD 600p is our first chance to test Silicon Motion’s SM2260 controller, their first PCIe SSD controller. Silicon Motion’s SATA SSD controllers have built a great reputation for being affordable, low power and providing good mainstream performance. One key to the power efficiency of Silicon Motion’s SATA SSD controllers is their use of an optimized single core ARC processor (via Synopsys), but in order to meet the SM2260’s performance target, Silicon Motion has finally switched to a dual core ARM processor. The controller chip used on the SSD 600p has some customizations specifically for Intel and bears both Intel and SMI logos.

The 3D TLC NAND used on the Intel SSD 600p is the first generation 3D NAND co-developed with Micron. We’ve already evaluated Micron’s Crucial MX300 with the same 3D TLC and found it to be a great mainstream SATA SSD. The MX300 was unable to match the performance of Samsung’s 3D TLC NAND as found in the 850 EVO, but the MX300 is substantially cheaper and remarkably power efficient, both in comparison to Samsung’s SSDs and to other SSDs that use the same controller as the MX300 but planar NAND.

Intel uses the same 3D NAND flash die for its MLC and TLC parts. The MLC configuration that has not yet found its way to the consumer SSD market has a capacity of 256Gb (32GB) per die, which gives the TLC configuration a capacity of 384Gb (48GB). Micron took advantage of this odd size to offer the MX300 in non-standard capacities, but for the SSD 600p Intel is offering normal power of two capacities with large fixed size SLC write caches in the spare area. The ample spare area also allows for a write endurance rating of about 0.3 drive writes per day for the duration of the five year warranty.

Intel 3D TLC NAND, four 48GB dies for a total of 192GB per package

The Intel SSD 600p shares its hardware with two other Intel products: the SSD Pro 6000p for business client computing and the SSD E 6000p for the embedded and IoT market. The Pro 6000p is the only one of the three to support encryption and Intel’s vPro security features. The SSD 600p relies on the operating system’s built-in NVMe driver and Intel’s consumer SSD Toolbox software which was updated in October to support the 600p.

For this review, the primary comparisons will not be against high-end NVMe drives but against mainstream SATA SSDs, as these are ultimately the closest to ‘mid-to-low range’ NVMe as we can get. The Crucial MX300 has given us a taste of what the Intel/Micron 3D TLC can do, and it is currently one of the best value SSDs on the market. The Samsung 850 EVO is very close to the Intel SSD 600p in price and sets the bar for the performance the SSD 600p needs to provide in order to be a good value.

Because the Intel SSD 600p is targeting a more mainstream audience and more modest level of performance than most other M.2 PCIe SSDs, I have additionally tested its performance in the M.2 slot built in to the testbed’s ASUS Z97 Pro motherboard. In this configuration the SSD 600p is limited to a PCIe 2.0 x2 link, as compared to the PCIe 3.0 x4 link that is available during the ordinary testing process where an adapter is used in the primary PCIe x16 slot. This extra set of results does not include power measurements but may be more useful to desktop users who are considering adding a cheap NVMe SSD to an older but compatible existing system.

• Thanks to Intel for the Core i7-4770K CPU
• Thanks to ASUS for the Z97 Deluxe motherboard
• Thanks to Corsair for the Vengeance 16GB DDR3-1866 DRAM kit, RM750 power supply, Carbide 200R case, and Hydro H60 CPU cooler

Our performance consistency test explores the extent to which a drive can reliably sustain performance during a long-duration random write test. Specifications for consumer drives typically list peak performance numbers only attainable in ideal conditions. The performance in a worst-case scenario can be drastically different as over the course of a long test drives can run out of spare area, have to start performing garbage collection, and sometimes even reach power or thermal limits.

In addition to an overall decline in performance, a long test can show patterns in how performance varies on shorter timescales. Some drives will exhibit very little variance in performance from second to second, while others will show massive drops in performance during each garbage collection cycle but otherwise maintain good performance, and others show constantly wide variance. If a drive periodically slows to hard drive levels of performance, it may feel slow to use even if its overall average performance is very high.

To maximally stress the drive’s controller and force it to perform garbage collection and wear leveling, this test conducts 4kB random writes with a queue depth of 32. The drive is filled before the start of the test, and the test duration is one hour. Any spare area will be exhausted early in the test and by the end of the hour even the largest drives with the most overprovisioning will have reached a steady state. We use the last 400 seconds of the test to score the drive both on steady-state average writes per second and on its performance divided by the standard deviation.

The Intel 600p’s steady state random write performance is reasonably fast, especially for a TLC SSD. The 600p is faster than all of the SATA SSDs in this collection. The Intel 750 and Samsung 960s are in an entirely different league, but the OCZ RD400 is only slightly ahead of the 600p.

Despite a decently high average performance, the 600p has a very low consistency score, indicating that even after reaching steady state, the performance varies widely and the average does not tell the whole story.

Default
25% Over-Provisioning

Very early in the test, the 600p begins showing cyclic drops in performance due to garbage collection. Several minutes into the hour-long test, the drive runs out of spare area and reaches steady state.

Default
25% Over-Provisioning

In its steady state, the 600p spends most of the time tracing out a sawtooth curve of performance that has a reasonable average but is constantly dropping down to very low performance levels. Oddly, there are also brief moments of unhindered performance where the drive spikes to exceptionally high performance of up to 100k IOPS, but these are short and infrequent enough to have little impact on the average performance. It would appear that the 600p occasionally frees up some SLC cache, which then immediately gets used up and kicks off another round of garbage collection.

With extra overprovisioning, the 600p’s garbage collection cycles don’t drag performance down as far, making the periodicity less obvious.

The Destroyer is an extremely long test replicating the access patterns of very IO-intensive desktop usage. A detailed breakdown can be found in this article. Like real-world usage and unlike our Iometer tests, the drives do get the occasional break that allows for some background garbage collection and flushing caches, but those idle times are limited to 25ms so that it doesn’t take all week to run the test.

We quantify performance on this test by reporting the drive’s average data throughput, a few data points about its latency, and the total energy used by the drive over the course of the test.

The average data rate sustained by the 600p on The Destroyer shows that the drive is almost entirely limited by the speed of the 3D TLC NAND, and it is barely any faster than the Crucial MX300 that uses the same NAND.

The average service time of the 600p is much lower than most TLC SSDs, so the NVMe PCIe interface is providing some latency benefit even when the drive’s throughput doesn’t need anything faster than SATA. The 600p is still only as good as a mid-range MLC SATA SSD and doesn’t come close to the low latency of high-end NVMe drives.

The number of high latency…