The TeamGroup L5 LITE 3D (480GB) SATA SSD Review: Entry-Level Price With Mainstream Performance
by Billy Tallis on September 20, 2019 9:00 AM EST- Posted in
- SSDs
- SATA
- Silicon Motion
- SM2258
- TeamGroup
Whole-Drive Fill
This test starts with a freshly-erased drive and fills it with 128kB sequential writes at queue depth 32, recording the write speed for each 1GB segment. This test is not representative of any ordinary client/consumer usage pattern, but it does allow us to observe transitions in the drive's behavior as it fills up. This can allow us to estimate the size of any SLC write cache, and get a sense for how much performance remains on the rare occasions where real-world usage keeps writing data after filling the cache.
 |
|||||||||
The ATSB tests already showed that the TeamGroup L5 LITE 3D doesn't lose much performance when it is full, but actually plotting its performance through the process of filling it up is surprising. The sequential write throughput does drop slightly after about 5GB, but only by 10-15MB/s, and there are no further performance drops for the rest of the fill process. This is a lot more consistent than most drives, and provides more evidence that SLC caches running out aren't a problem for this SSD.
 |
|||||||||
| Average Throughput for last 16 GB | Overall Average Throughput | ||||||||
Working Set Size
When DRAMless SSDs are under consideration, it can be instructive to look at how performance is affected by working set size: how large a portion of the drive is being touched by the test. Drives with full-sized DRAM caches are typically able to maintain about the same random read performance whether reading from a narrow slice of the drive or reading from the whole thing. DRAMless SSDs often show a clear dropoff when the working set size grows too large for the mapping information to be kept in the controller's small on-chip buffers.
 |
|||||||||
As expected, the L5 LITE 3D maintains fairly steady random read performance regardless of working set size. The DRAMless Mushkin Source starts off with significantly lower random read IOPS and declines even more as working set sizes grow to more than a few GB of active data. The three drives here with Phison controllers (one SATA, two NVMe) all show at least some decline in performance with large working set sizes, even though those drives all have the usual 1GB DRAM to 1TB NAND ratio.
Facebook
Twitter
RSS
42 Comments
View All Comments
DanNeely - Friday, September 20, 2019 - link
In the consumer segment so am I. For an OEM one I can understand more easily, it's performance is still better than a typical eMMC while being between the eMMC and a 240GB m.2. Going into sub $500 systems even a few dollars on the BOM are significant because margins are so low, and this offers a cheaper upgrade from eMMC than the 240 which like you I'd strongly recommend spending the extra few bucks for if building a system.Kristian Vättö - Friday, September 20, 2019 - link
Ryan & Billy,I have to agree with some of the other commenters here that the conclusion is inconclusive and is primarily focused on price and performance, which are obviously important, but should not be the only factors in consideration when recommending a drive. A few points that I feel were left out in the conclusion:
1. NAND switching has always been a big no-no. Why? Because it's impossible to guarantee that an end-user will receive a drive with the same BOM. While the review unit has competitive performance, what if it is the highest performing of the known four variants? For example, going from 256Gbit dies to 512Gbit dies could easily halve the write performance and substantially change the ATSB numbers. At a minimum, AT should request Team to disclose the different NAND configurations along with their internal performance data, so that a first degree conclusion of the performance between the different variants could be drawn.
2. Reliability has very little to do with the physical controller. It's a piece of silicon like a CPU and rarely fails unless subjected to extraordinary environmental conditions (heat, humidity etc). What matters are the firmware and NAND. Firmware in this case is likely just a standard SMI FW with minimal modifications, so that’s not much of a risk. But the NAND is a big question mark. It being Team branded means it’s not a qualified component from a NAND vendor, but something that has been packaged by 3rd party. That opens a possibility for using unqualified NAND i.e. dies that don’t meet the NAND vendors’ specs, such as having too many bad blocks from the beginning. Many of these Tier2/3 SSDs mix good and bad dies to drive the cost down and that’s also why most of them are 240GB instead of 250 or 256GB since the extra spare area helps to cover more bad blocks. Of course it’s no guarantee that the drive will fail prematurely, but there is always a reason why a certain product is cheaper than others.
3. Amazon/NewEgg reviews are not bulletproof, but serve as a good first degree reality check, especially if the drive has already been in the market for a while. With 27% 1-star reviews, I would personally not have the guts to recommend the drive unless it’s substantially (>20%) cheaper than any household SSD brand/model.
4. RMA process and general support are areas that have more importance now since the performance differences between SATA SSDs especially are becoming minor. Some vendors offer very good terms with e.g. advance replacement, which can be highly important if the SSD is used in a primary system. Including a paragraph on the company’s policy would be something I recommend as it can be a real headache especially with lesser known brands (e.g. long response time, need to wait weeks for a replacement etc). It may not be possible for AT to test the process, but the key aspects can be covered with a paper comparison.
None of these points mandate an overhaul of the conclusion as the Lite 3D may very well be an excellent choice for a budget-focused buyer, but at least there should be a disclaimer of the caveats to raise the question whether saving $5 or $10 is worth it over a safe, well-known Tier1 brand.
sheh - Friday, September 20, 2019 - link
Agreed.Furthermore, on the topic of random NAND types in the same model, since the drive here was provided by the manufacturer, it's almost certain they provided the best variant. For all we know it could be a NAND type they don't even use anymore.
Kenaz - Friday, September 20, 2019 - link
Find it good that also low-priced products are tested. To my knowledge Anandtech should be the first to give a Team L5 Lite 3D SSD a professional test.I wrote last year for the 120GB and 240GB model of this SSD series two user reviews on Hardwareluxx Germany and was a bit impressed by the performance for a budget drive. Since then I have recommended this SSD series as a possible budget option. Both SSDs run now for 1 / 1.5 years without problems and good S.M.A.R.T. values.
A long-term write test would be interesting to see how much terrabyte TBW the unknown NAND can withstand. Would this be possible with Anandtech? :)
Translated with www.DeepL.com/Translator
MDD1963 - Saturday, September 21, 2019 - link
$96 for 1 TB is priced attractively, but, I think I'd just spring the extra $5-10 on Crucial's MX500...; it would need to be $68.99 to get me to risk it! :)ballsystemlord - Monday, September 23, 2019 - link
Billy, having to click to see each image slows down the experience, could you guys just allow the page to load most or all of the images? Please?takeshi7 - Wednesday, September 25, 2019 - link
I have two of these 480GB drives and one 240GB drive. I was so impressed with the first one I got two more for other PCs. Really great value, but the gold color doesn't match common PC colors.MASSAMKULABOX - Wednesday, October 16, 2019 - link
Might be worth asking them when and why they change the NAND variety. I know its going to be price, but what stops them using DongCrap NAND? , what is their Baseline criteria ..## commercially sensitive Blah blah. Maybe they are hoping that three years is good enough until 480 layer Penta cell NAND takes over. Price IS the main consideration when buying sata drives tho ...MASSAMKULABOX - Wednesday, October 16, 2019 - link
£124 for 1tb AMZ/uk , so not even in the running really ...Scour - Friday, July 10, 2020 - link
I hoped to get more info about this SSD, but it looks like a flameware about the brand.I don´t know how many of these ppl who flamed about Teamgroup ever had a article from this brand, but probably not many.
I have my L3 since almost 4 years and it never had problems and it´s still faster than many newer entry-level-SSDs.