Original Link: https://www.anandtech.com/show/11844/the-team-group-tforce-cardea-240gb-ssd-review



Team Group is a major vendor of DRAM modules, and they also have a variety of SSDs. They are one of the few SSD vendors that uses Phison turnkey solutions for some of their SSDs but still also designs their own drives using controllers from other vendors like Marvell. Most brands that use Phison have abandoned in-house drive or firmware development.

The Team T-Force Cardea is an enthusiast-oriented M.2 PCIe SSD carrying a large heatsink that makes it more or less a desktop-only SSD. The heatsink is not so tall that it will interfere with graphics cards when installed in a nearby slot, but this heatsink is far more substantial than the low-profile heatspreaders we've seen on other M.2 PCIe SSDs. Underneath the heatsink, the T-Force Cardea is a reference design Phison E7 NVMe SSD with 15nm MLC NAND.

The 240GB Team T-Force Cardea is the lowest capacity SSD we've tested that uses the Phison E7 controller. This gives us the chance to see how well the performance holds up with less spare area on the drive and fewer NAND dies to use in parallel. The 240GB T-Force Cardea is still large enough to fully populate the NAND channels on the E7 controller, so it isn't in an awkward position like the Intel SSD 600p.

SSDs using the Phison E7 controller have all used Toshiba 15nm planar MLC NAND; the controller doesn't support LDPC error correction or anything similar, so it is generally not regarded as a good match for TLC NAND. Despite using MLC, Phison E7 drives have often been some of the most affordable consumer NVMe SSDs, even when compared against drives using planar or 3D TLC NAND. We tend to regard Phison E7 SSDs as entry-level NVMe products, but a desktop-only drive with a heatsink clearly aspires to be more than that.

Team T-Force Cardea Specifications
Capacity 240GB 480GB
Form Factors M.2 2280 PCIe x4 with heatsink
Controller Phison E7
NAND Toshiba 15nm MLC
Sequential Read 2600 MB/s 2650 MB/s
Sequential Write 1400 MB/s 1450 MB/s
4KB Random Read  180k IOPS 180k IOPS
4KB Random Write  140k IOPS 150k IOPS
Encryption No
Write Endurance 335 TB 670 TB
Warranty Three years
Price $129.99 $219.99

The performance specifications for the two capacities of the T-Force Cardea are very similar, suggesting that the larger model won't feel any faster. The warranty period of three years is typical for this market segment, and the write endurance ratings are equivalent to about 1.2 drive writes per day over that period, which is very good for a consumer drive.

Each Phison E7 drive we've tested has used a different firmware revision that makes different performance tradeoffs. Our most recently reviewed Phison E7 SSD was the Corsair Neutron NX500 with firmware version 4.5, which seemed to prioritize peak benchmark performance over low queue depth real-world performance. The Patriot Hellfire's version 2.1 firmware seemed to be pretty well-rounded with reasonable priorities. The Team T-Force Cardea uses Phison's version 3.6 firmware. We also have results for the Zotac SONIX with 1.0 firmware.

Aside from comparing against other Phison E7 SSDs, the most relevant competitor to the 240GB Team T-Force Cardea is the 250GB Samsung 960 EVO. The 960 EVO uses the most advanced NVMe SSD controller currently available, but is held back somewhat by using 3D TLC instead of the 3D MLC of the 960 PRO, and the 250GB capacity often suffers from the small size of its SLC cache and spare area. At the moment, the 960 EVO is a bit cheaper than the T-Force Cardea.

AnandTech 2017 SSD Testbed
CPU Intel Xeon E3 1240 v5
Motherboard ASRock Fatal1ty E3V5 Performance Gaming/OC
Chipset Intel C232
Memory 4x 8GB G.SKILL Ripjaws DDR4-2400 CL15
Graphics AMD Radeon HD 5450, 1920x1200@60Hz
Software Windows 10 x64, version 1703
Linux kernel version 4.12, fio version 2.21


AnandTech Storage Bench - The Destroyer

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, 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. These AnandTech Storage Bench (ATSB) tests do not involve running the actual applications that generated the workloads, so the scores are relatively insensitive to changes in CPU performance and RAM from our new testbed, but the jump to a newer version of Windows and the newer storage drivers can have an impact.

We quantify performance on this test by reporting the drive's average data throughput, the average latency of the I/O operations, and the total energy used by the drive over the course of the test.

ATSB - The Destroyer (Data Rate)

The Team T-Force Cardea has a better overall data rate on The Destroyer than the SATA drives or the Samsung 960 EVO, but the larger Phison E7 drives and the 256GB MLC-based NVMe drives are much faster.

ATSB - The Destroyer (Average Latency)ATSB - The Destroyer (99th Percentile Latency)

For both average and 99th percentile latency during The Destroyer, the T-Force Cardea is the slowest Phison E7 SSD. The Phison drives all have decent 99th percentile latency, while most of Samsung's drives have more significant outliers.

ATSB - The Destroyer (Average Read Latency)ATSB - The Destroyer (Average Write Latency)

The rankings for average read and write latencies are mostly the same, with the T-Force Cardea coming in last among the Phison E7 drives but still scoring much better than the Samsung 960 EVO.

ATSB - The Destroyer (99th Percentile Read Latency)ATSB - The Destroyer (99th Percentile Write Latency)

The Phison E7 drives do fairly well at keeping latency outliers under control, though there are major differences between firmware versions. The T-Force Cardea's 99th percentile read latency is much better than the Samsung 960 EVO but also clearly slower than the other Phison E7 drives. On the write side, the T-Force Cardea's 99th percentile write latency is not the slowest among the Phison E7 drives, and all the Phison E7 drives score better than the competition.

ATSB - The Destroyer (Power)

The Team T-Force Cardea has good power efficiency on The Destroyer, with lower total energy usage than any other NVMe SSD in this bunch except the discontinued Samsung 950 PRO.



AnandTech Storage Bench - Heavy

Our Heavy storage benchmark is proportionally more write-heavy than The Destroyer, but much shorter overall. The total writes in the Heavy test aren't enough to fill the drive, so performance never drops down to steady state. This test is far more representative of a power user's day to day usage, and is heavily influenced by the drive's peak performance. The Heavy workload test details can be found here. This test is run twice, once on a freshly erased drive and once after filling the drive with sequential writes.

ATSB - Heavy (Data Rate)

The Team T-Force Cardea is roughly tied with the Samsung 960 EVO for average data rate on the Heavy test. The T-Force Cardea is slower than the larger Phison E7 SSDs, but the Patriot Hellfire's performance when full is the same as the T-Force Cardea's.

ATSB - Heavy (Average Latency)ATSB - Heavy (99th Percentile Latency)

The Patriot Hellfire's latency is quite poor when it's full, but otherwise the larger Phison E7 drives have much lower average and 99th percentile latency than the T-Force Cardea. The Samsung 960 EVO performs worse on both metrics, while the OCZ RD400 and Samsung 950 PRO are at the top of the charts.

ATSB - Heavy (Average Read Latency)ATSB - Heavy (Average Write Latency)

Splitting the average latency by reads and writes, we see that the Samsung 960 EVO's problem is mostly on the write side, where the T-Force Cardea beats it easily. The Patriot Hellfire's latency issues when full are worse for writes than for reads.

ATSB - Heavy (99th Percentile Read Latency)ATSB - Heavy (99th Percentile Write Latency)

The Phison E7 drives had good overall 99th percentile latency scores, so its not surprising to see them do well when looking at reads and writes individually. The T-Force Cardea's 99th percentile read latency is worse than the larger Phison E7 drives but half that of the Samsung 960 EVO. For writes, the T-Force Cardea and the 960 EVO are tied when the test is run on a full drive, but the T-Force Cardea's 99th percentile write latency is half that of the 960 EVO when the test is run on an empty drive.

ATSB - Heavy (Power)

The Phison E7 drives all show poor energy efficiency on the Heavy test, though the T-Force Cardea is the least power-hungry among them.



AnandTech Storage Bench - Light

Our Light storage test has relatively more sequential accesses and lower queue depths than The Destroyer or the Heavy test, and it's by far the shortest test overall. It's based largely on applications that aren't highly dependent on storage performance, so this is a test more of application launch times and file load times. This test can be seen as the sum of all the little delays in daily usage, but with the idle times trimmed to 25ms it takes less than half an hour to run. Details of the Light test can be found here. As with the ATSB Heavy test, this test is run with the drive both freshly erased and empty, and after filling the drive with sequential writes.

ATSB - Light (Data Rate)

On the Light test, the Team T-Force Cardea has a much lower average data rate than the larger Phison E7 drives or any of the other NVMe SSDs in this bunch.

ATSB - Light (Average Latency)ATSB - Light (99th Percentile Latency)

The average and 99th percentile latency scores of the T-Force Cardea are fine when the test is run on an empty drive, but are much worse when the drive is full. The Samsung 960 EVO is also strongly affected by a full drive, but its empty-drive performance is so good that the full-drive scores are still better than any of the Phison E7 drives, especially the T-Force Cardea.

ATSB - Light (Average Read Latency)ATSB - Light (Average Write Latency)

Splitting the average latency into reads and writes reveals that the T-Force Cardea's most acute problem on the Light test is write latency when the drive is full.

ATSB - Light (99th Percentile Read Latency)ATSB - Light (99th Percentile Write Latency)

The full-drive write latency penalty of the T-Force Cardea is even more apparent when looking at 99th percentile write latency than average write latency. Otherwise, its latency compares favorably against the other Phison E7 drives and is much lower than the SATA drives. The Samsung 960 EVO has a problem with read latency when full, but its write latency stays very good even when full.

ATSB - Light (Power)

The two M.2 Phison E7 SSDs are tied for power consumption, along with the OCZ RD400. The only drives in this bunch that are more power-hungry are the add-in card form factor Phison E7 drives. Samsung's drives are all much more efficient, especially the 850 EVO.



Random Read Performance

Our first test of random read performance uses very short bursts of operations issued one at a time with no queuing. The drives are given enough idle time between bursts to yield an overall duty cycle of 20%, so thermal throttling is impossible. Each burst consists of a total of 32MB of 4kB random reads, from a 16GB span of the disk. The total data read is 1GB.

Burst 4kB Random Read (Queue Depth 1)

The Phison E7 drives vary widely in their QD1 random read performance. The Team T-Force Cardea isn't the fastest, but it's above average and faster than the SATA drives.

Our sustained random read performance is similar to the random read test from our 2015 test suite: queue depths from 1 to 32 are tested, and the average performance and power efficiency across QD1, QD2 and QD4 are reported as the primary scores. Each queue depth is tested for one minute or 32GB of data transferred, whichever is shorter. After each queue depth is tested, the drive is given up to one minute to cool off so that the higher queue depths are unlikely to be affected by accumulated heat build-up. The individual read operations are again 4kB, and cover a 64GB span of the drive.

Sustained 4kB Random Read

On the longer random read test that includes some moderate queue depths, the T-Force Cardea is the fastest Phison E7 drive and slightly faster than the best SATA drive, but clearly slower than most NVMe drives.

Sustained 4kB Random Read (Power Efficiency)

In addition to being the fastest Phison E7 drive on the random read test, the T-Force Cardea is also the most power efficient. However, every other M.2 or SATA SSD in this bunch is more efficient than the Phison E7 drives. The Samsung 850 EVO, while slightly slower, delivers twice the efficiency.

At higher queue depths, the random read performance of the Team T-Force Cardea is identical to that of the larger Patriot Hellfire, while the Corsair Neutron NX500 hits the highest speeds out of the Phison E7 drives. The Samsung 960 EVO has a small lead over the T-Force Cardea, and that lead widens with increasing queue depth.

Random Write Performance

Our test of random write burst performance is structured similarly to the random read burst test, but each burst is only 4MB and the total test length is 128MB. The 4kB random write operations are distributed over a 16GB span of the drive, and the operations are issued one at a time with no queuing.

Burst 4kB Random Write (Queue Depth 1)

The burst random write performance of the Team T-Force Cardea is second-fastest among the Phison E7 drives, but relatively slow among the wider field of NVMe competition.

As with the sustained random read test, our sustained 4kB random write test runs for up to one minute or 32GB per queue depth, covering a 64GB span of the drive and giving the drive up to 1 minute of idle time between queue depths to allow for write caches to be flushed and for the drive to cool down.

Sustained 4kB Random Write

When longer run times and higher queue depths come into play, the T-Force Cardea and other Phison E7 drives look better. The Intel 750's random write performance can't be matched, but the Cardea is well ahead of Samsung's small NVMe drives.

Sustained 4kB Random Write (Power Efficiency)

The Team T-Force Cardea and the Patriot Hellfire are tied for first place in random write power efficiency, with a clear lead over the OCZ RD400 and Samsung's drives.

The random write performance of the T-Force Cardea scales very will with increasing queue depth until the drive runs out of spare area and needs to perform garbage collection. The larger Phison E7 drives were able to make it through the entire test without performance falling off, while the smaller Samsung NVMe drives never reached particularly high performance to begin with.



Sequential Read Performance

Our first test of sequential read performance uses short bursts of 128MB, issued as 128kB operations with no queuing. The test averages performance across eight bursts for a total of 1GB of data transferred from a drive containing 16GB of data. Between each burst the drive is given enough idle time to keep the overall duty cycle at 20%.

Burst 128kB Sequential Read (Queue Depth 1)

The burst sequential read performance of the Samsung NVMe drives is unmatched, with the 960 EVO about three times faster than the Team T-Force Cardea. It looks like Samsung's drives are quite effective at pre-fetching data in a way the other NVMe SSDs can't.

Our test of sustained sequential reads uses queue depths from 1 to 32, with the performance and power scores computed as the average of QD1, QD2 and QD4. Each queue depth is tested for up to one minute or 32GB transferred, from a drive containing 64GB of data.

Sustained 128kB Sequential Read

The sustained sequential read performance of the Team T-Force Cardea is lower than the other (larger) Phison E7 drives and all the rest of the NVMe SSDs in this bunch, leaving the Cardea with just a 12% advantage over the fastest SATA SSD.

Sustained 128kB Sequential Read (Power Efficiency)

The power efficiency of the T-Force Cardea is also sub-par, but not by quite as large a margin. Samsung as usual is generally at the top, and the larger Phison E7 drives manage somewhat better efficiency than the T-Force Cardea die to their higher performance.

The T-Force Cardea takes its time ramping up to full performance on our sustained sequential read test, needing a queue depth of at least 4 to get close to its maximum. That eventual limit is also well below the performance of important competitors like the Samsung 960 EVO.

Sequential Write Performance

Our test of sequential write burst performance is structured identically to the sequential read burst performance test save for the direction of the data transfer. Each burst writes 128MB as 128kB operations issued at QD1, for a total of 1GB of data written to a drive containing 16GB of data.

Burst 128kB Sequential Write (Queue Depth 1)

The burst sequential write performance of the Phison E7 drives are all quite close, and second only to the Samsung 960 EVO, albeit by a wide margin.

Our test of sustained sequential writes is structured identically to our sustained sequential read test, save for the direction of the data transfers. Queue depths range from 1 to 32 and each queue depth is tested for up to one minute or 32GB, followed by up to one minute of idle time for the drive to cool off and perform garbage collection. The test is confined to a 64GB span of the drive.

Sustained 128kB Sequential Write

 

On the sustained sequential write test, the T-Force Cardea's capacity disadvantage comes into play, but it still manages to perform well and beat the Samsung 960 EVO by over 50%.

Sustained 128kB Sequential Write (Power Efficiency)

The power efficiency of the T-Force Cardea is good relative to most of its competition, though it is beat by some of the larger Phison E7 drives and by the Samsung 950 PRO.

The main failing of the Team T-Force Cardea on the sustained sequential write test seems to be at QD1, where the larger Phison E7 drives are already operating close to their full speed. At QD2 or higher, the Cardea is only slightly slower than the other Phison E7 drives.



Mixed Random Performance

Our test of mixed random reads and writes covers mixes varying from pure reads to pure writes at 10% increments. Each mix is tested for up to 1 minute or 32GB of data transferred. The test is conducted with a queue depth of 4, and is limited to a 64GB span of the drive. In between each mix, the drive is given idle time of up to one minute so that the overall duty cycle is 50%.

Mixed 4kB Random Read/Write

The Team T-Force Cardea has average performance on the mixed random I/O test. It's about the same speed as Samsung's 850 PRO SATA SSD, and is substantially outperformed by some of the larger Phison E7 drives and the Toshiba OCZ RD400. The Samsung 960 EVO does poorly here, with only 75% of the performance overall compared to the T-Force Cardea.

Mixed 4kB Random Read/Write (Power Efficiency)

The power efficiency of the T-Force Cardea isn't great; the SATA SSDs that come close to its performance level do so while consuming much less power. While the T-Force Cardea was 32% faster than the Samsung 960 EVO, the Samsung closes the gap a bit on efficiency, leaving the T-Force Cardea with only an 11% efficiency advantage.

The performance of the T-Force Cardea increases slowly across most of the test, until the write volume is high enough to trigger garbage collection. The larger Phison E7 drives never reach that threshold and instead show a spike in performance at the end of the test when the workload is pure writes that can be cached and combined.

The Samsung 960 EVO starts out with slightly higher performance on the random read side of the test, but degrades as the proportion of writes increases through the first half of the test, before increasing at the end and surpassing the T-Force Cardea when it is busy doing background garbage collection.

Mixed Sequential Performance

Our test of mixed sequential reads and writes differs from the mixed random I/O test by performing 128kB sequential accesses rather than 4kB accesses at random locations, and the sequential test is conducted at queue depth 1. The range of mixes tested is the same, and the timing and limits on data transfers are also the same as above.

Mixed 128kB Sequential Read/Write

The top performers on the mixed sequential I/O test are both older NVMe drives: the Samsung 950 PRO and Intel SSD 750. Several of the Phison E7 drives and the Samsung 960 EVO and 850 PRO are all tied for average speed, while the Zotac SONIX and OCZ RD400 are 100MB/s faster.

Mixed 128kB Sequential Read/Write (Power Efficiency)

The Team T-Force Cardea has much better performance per Watt than the other two Phison E7 SSDs that performed the same, and is almost as efficient as the Zotac SONIX that performed much better. Among currently-available NVMe SSDs, the Toshiba OCZ RD400 seems to the best combination of performance and power efficiency on this test.

The T-Force Cardea's performance drops slightly early in the test, but mostly shows increased performance as the proportion of writes increases. There's no sign of the total write volume being enough to force the drive to perform garbage collection outside of the idle periods between sub-tests. The Samsung 960 EVO's performance is almost a mirror of the T-Force Cardea's, with much better performance on the read-heavy half of the test than the write-heavy second half. Samsung's SATA drives are also better on the read-heavy portions. So while several drives are tied for overall average performance, the Samsung drives perform better on the kinds of I/O that consumer workloads are mostly comprised of.



Power Management

Real-world client storage workloads leave SSDs idle most of the time, so the active power measurements presented earlier in this review only account for a small part of what determines a drive's suitability for battery-powered use. Especially under light use, the power efficiency of a SSD is determined mostly be how well it can save power when idle.

SATA SSDs are tested with SATA link power management disabled to measure their active idle power draw, and with it enabled for the deeper idle power consumption score and the idle wake-up latency test. Our testbed, like any ordinary desktop system, cannot trigger the deepest DevSleep idle state.

Idle power management for NVMe SSDs is far more complicated than for SATA SSDs. NVMe SSDs can support several different idle power states, and through the Autonomous Power State Transition (APST) feature the operating system can set a drive's policy for when to drop down to a lower power state. There is typically a tradeoff in that lower-power states take longer to enter and wake up from, so the choice about what power states to use may differ for desktop and notebooks.

We report two idle power measurements. Active idle is representative of a typical desktop, where none of the advanced PCIe link or NVMe power saving features are enabled and the drive is immediately ready to process new commands. The idle power consumption metric is measured with PCIe Active State Power Management L1.2 state enabled and NVMe APST enabled.

Active Idle Power Consumption (No LPM)Idle Power Consumption

As with the other Phison E7 drives, the Team T-Force Cardea has a compatibility problem with out testbed when it comes to power management. Since the T-Force Cardea is intended as a desktop-only product, the overall power consumption is a very minor concern, and the heatsink eliminates any concerns about a higher idle temperature making the drive more susceptible to thermal throttling under load.

Idle Wake-Up Latency

With power management that is inoperative on our testbed, none of the Phison E7 drives show more than a trivial delay when first given I/O commands after an idle period. By contrast, the Samsung 960 EVO needs an extra 5ms to come out of its sleep state.



Conclusion

Modern 3D NAND tends to be manufactured with a high capacity per die, which makes it difficult to make a high-performance low-capacity SSD. The Team T-Force Cardea avoids that problem by using Toshiba's 15nm planar MLC NAND. The result is a drive that handles heavy workloads quite well without the performance drops that are typical of TLC SSDs. The Phison E7 SSD controller platform used by the T-Force Cardea helps keep costs much lower than most other MLC-based drives, though the heatsink prevents it from being the cheapest Phison E7 drive on the market.

The most important competitor at the low end of the capacity range is Samsung's 960 EVO 250GB. The MLC-based 960 PRO isn't available at capacities below 512GB, and the previous generation 256GB 950 PRO is out of production. Thanks to the strength of Samsung's Polaris controller, the 250GB 960 EVO manages to be the fastest drive in its class for light workloads. But when heavier workloads with a high volume of writes are involved, the T-Force Cardea comes out ahead.

We haven't seen any strong evidence that the heatsink helps the T-Force Cardea under ordinary conditions. In fact, we often see performance degrade as a test continues and the accumulated volume of writes forces the drive to continue performing garbage collection beyond the idle time our test protocol provides. Larger SSDs and many competing 240GB-class SSDs seem to be able to wrap up their garbage collection more quickly, leading to more consistent sustained performance. But this shortcoming of the T-Force Cardea is only really noticeable on our synthetic benchmarks; even our most intense tests of real-world I/O patterns have it clearly outperforming SATA SSDs and most cheaper NVMe SSDs.

The most significant performance weakness we spotted during our testing is with sequential reads. Samsung's NVMe SSDs are several times faster at queue depth 1. The gap narrows at higher queue depths, but all of the Phison E7 SSDs are still at a disadvantage here.

  250GB 500-512GB 1TB
Team T-Force Cardea $129.99 (54¢/GB) $219.99 (46¢/GB)  
Samsung 960 EVO $117.60 (47¢/GB) $234.00 (47¢/GB) $467.00 (47¢/GB)
Samsung 960 PRO   $298.00 (58¢/GB) $598.38 (58¢/GB)
MyDigitalSSD BPX $114.99 (48¢/GB) $199.99 (42¢/GB)  
Toshiba OCZ RD400 $118.79 (46¢/GB) $239.99 (47¢/GB) $567.18 (55¢/GB)
Intel SSD 600p $165.59 (65¢/GB) $199.99 (39¢/GB) $329.99 (32¢/GB)

Current pricing has the 240GB T-Force Cardea well above the Samsung 960 EVO, and just below it at the half-TB capacity class. For most users, the 960 EVO's performance profile will be a better fit, making the 960 EVO the better buy. For a particularly heavy workload at, the T-Force Cardea may be a better choice than the 250GB 960 EVO. But since it doesn't appear that the heatsink matters in ordinary use, most users can save even more money by going with a cheaper Phison E7 drive like the MyDigitalSSD BPX.

In an upcoming review, we will more thoroughly explore the thermal limits of M.2 SSDs, including when positioned near a hot-running graphics card. That may reveal the large heatsink on the Team T-Force Cardea to occasionally be a significant advantage.

Log in

Don't have an account? Sign up now