LukeSavenije

(Image credit: Adata) Introduction This article is partially based off Sean Webster's article on Tomshardware. There will be a lot of similar, but hopefully also some new info present in this. As I'm a personal owner of an S11 pro, I've done performance testing and a readout of the SSD for a review I've had pending for a while, but with this news I decided to bring this info to the general public. This SSD includes some of the part swaps mentioned in Sean's article. I'll do my best to go into a neutral perspective with this. What happened? Long story short, Adata did a part swap on their SX8200 pro and S11 pro series, making them essentially the same as their non-pro counterparts. The main difference in this regard used to be the controller, with non-pro using the Silicon Motion SM2262(G) and the pro variant using the Silicon Motion SM2262EN(G). Together with that, they've chosen a different NAND for the SX8200 pro, but my S11 pro here is using Micron NAND instead of the article mentioned Samsung NAND. How much impact does this have? Not a lot. SM2262EN is faster in some tasks, but this is by a small margin, which got even smaller over time with the firmware updates that Silicon Motion has pushed to the SM2262. Regardless I think this should be at least documented for compare and information to the consumer, as well as showing that Adata is currently selling 4 nearly identical drives. The main difference being that sx8200 (pro) lacks the heatsink given with the s11 (pro), but the PCB is on all 4 identical. Why did Adata do this? All things considered, likely cost was the main reason. It's common for many manufacturers to swap parts. Things like NAND/DRAM manufacturers, controllers and such sometimes just have to be swapped with revisions, but as long as this is within the description and performance of the older one, this is totally fine to do. The only problem I have with this is that Adata already has a drive that is basically identical to this one, so even if the margin is small, I'd rather have seen the SX8200/S11 pro retired and replaced by a newer revision of the SX8200/S11 instead of trying to silently partswap the popular two of them with the parts of the other two. The Software Used CrystalDiskMark 7.0 TxBENCH 0.96 beta Anvil’s Storage Utilities 1.1.0 AS SSD Benchmark 2.0.7316.34247 ATTO Disk Benchmark 4.01.0f1 The Hardware Used CPU: Intel Core i5-8600K Motherboard: MSI Z370 Pro Carbon AC RAM: G.Skill Trident Z 2x8GB 3200Mhz CL16 Storage: Samsung 860 EVO 500GB (OS) & ADATA Gammix S11 Pro (for testing) Power Supply: Cooler Master V1300 OS: Windows 10 CrystalDiskMark TxBench Anvil’s AS SSD ATTO Readout data Conclusion Adata did an odd move here, even if it's still well performing as a high end NVME, it makes the drive and Adata a bit more questionable. Regardless, this isn't an uncommon move and even if I would've preferred to see it done in a different way, it is how it is and the drive still deserves to be recommended.

too minor to care, they're the same platform, just with a modular board as difference

you're fine with a 750-850w here, and that would even leave you a ton of headroom regardless. systems like these could run on a good 650w if needed without any problem some options would include - Fractal Design ION+ - Seasonic Focus PX - Corsair HX(i) - Seasonic Prime PX/TX - Corsair AX - BeQuiet Straight Power 11 platinum - Enermax Maxtytan - BeQuiet Dark Power Pro 11 - Cooler Master V-platinum - FSP Hydro PTM amongst others

you're doing two things wrong here 1: the "50% is most efficient" myth comes from the 80+ certification. since they only do 3-4 loads, being 10%, 20%, 50% and 100%, 50% comes out as the most efficient. When looking at a test that does it per 10% instead you'll see that the efficiency curve actually peaks before or after that point generally, not on the exact load of 50%. You can check the many, many reviews of various PSUs for this, or look at a certification like Cybenetics' ETA 2: are you planning to go with a X299/TR dual GPU system any time soon? if not, you really don't need any kind of that wattage, with parts getting more efficient and the expected launch of ATX 12VO your PSU will be obselite in a couple years anyways. with that said, if you're actually looking for a reliable unit and you want to get the most efficiency out of it, don't go with a high wattage unit. grab something reasonable in wattage and run it somewhere in the region of 30-80% load, and don't expect the peak wattage to be 24/7 unless you run furmark+prime95 for fun. State your system you want to combine it with and if you actually do want the math to be precise... maybe your electronics cost?

I would personally also look at Gold - Enermax Revolution DF - Bitfenix Whisper - BeQuiet Straight Power 11 - Corsair RMx - Thermaltake Toughpower GF1 - Adata XPG Core Reactor - Chieftec/Chieftronic Powerplay - NZXT C/E - Asus ROG Strix - Leadex III - Inwin PB - Deepcool DQ-M V1 - Antec High Current Gamer (extreme) Platinum - Fractal Design ION+ - Bequiet Straight Power 11 - Corsair HXi - Corsair HX - FSP Hydro PTM - Cooler Master V(engeance) new amongst others that I probably forgot to mention

SE as far as I'm aware is still leadex platinum based, but as far as I'm aware the specific unit hasn't been reviewed yet and I mean... not that Superflower hasn't done crap on leadex, but oh well...

they used Cybenetics data for that btw, the person that made the review (breixobaloca) works with us

as long as you utilize multiple VGA rails, it should be fine we actually keep a whole methodology behind it, which you can find back in the spreadsheet, in case you're curious

we as a team read everything here, so we just reply whenever we can problem is... with the nature of "lemons" and such, we can't base our list just off one failed unit for a user, unless it's shown that it's a wider spread issue (for example a design oversight) or it's something that could affect multiple users (for example a GPU transient shutdown happening for a couple users with a specific unit where wattage isn't the issues). So it's not a bad thing you comment for, but not much we can do with it either unless it's more specific

more ryzen and navi will be added once I've got the numbers for now, Intel and Nvidia with theirs have been added

we do indeed require it to be within ATX specification for 12v and 5v for tier A, but if you mean transient shutdowns, they get moved to low priority when discovered to be a wider problem

ctrl+g is your friend it's tier C

if you get a very old batch of it, yes... but current production shouldn't

yep

modularity, the CM version is semi-modular

RMx uses Hong Hua as well. RMx will be more quiet, but less efficient at lower loads. the rest comes down to pricing and use.

can't reach full capacity under a low input voltage, nothing too worrying, but worth putting it there for

agreeing with @Arika S here, as long as it's not in an annoying way and within CS, it's a totally fine thing to do

it is

in-cable capacitors, a bit less noise, a bit worse 2% load efficiency and a bit upgraded internals compared to RM 2019