"..the per-die capacity of the MX500's 64-layer 3D TLC is actually lower than that of the 32L 3D TLC.."
Why is that the case? Shouldn't doubling the number of layers lead to greater capacity and does this mean that there are more ICs per GB in the newer MX500? I'm super confused.
"The Crucial MX500 uses Micron's 256Gb 64L 3D TLC part and consequently returns to the standard drive capacities and overprovisioning ratios, instead of the unusual configurations caused by the 384Gb die capacity in the MX300's 32L 3D TLC. The slightly lower usable capacities of the MX500 than the MX300 means the new drives have slightly higher prices on a per-GB basis, but the MSRPs are still very competitive against current street prices for the competition."
So the new chips have more layers per package, but less overall capacity. I'll guess the 384Gb die hamstrung performance on relatively smaller-capacity drives by offering fewer memory channels for the processor to work with simultaneously. Plus, I'll guess that it was a bit of an oddball size for the algorithms to deal with.
For the 32L node, IMFT was more or less prioritizing MLC use cases by making a 256Gb MLC die that could be used as a 384Gb TLC part. For the 64L node, TLC is the priority and they're making both 256Gb TLC and 512Gb TLC parts. The latter should be cheaper per GB when it is available, but would be a worse choice for small consumer drives. The 256Gb die is really tiny, which makes it more appealing to the mobile market than a lot of previous IMFT parts.
"making a 256Gb MLC die that could be used as a 384Gb TLC part."
so... does this mean that NAND is all the same, and it's the controller that decides among S/M/T?? or is it the case that S can be coerced to M, and M to T, but not the other way round? is there a tute, here or elsewhere, that tells us the nittygritty about why titular NAND can/can't be used at other densities?
well, I did find one, but from 2012 (and AT: https://www.anandtech.com/show/5067/understanding-... ) "This array can be turned into either SLC, MLC, or TLC. The actual array and transistors are equivalent in all three flash types; there is no physical difference. "
some/many folks have been claiming that TLC, in particular, is physically different from SLC. is that now true?
In terms of memory cell design, SLC, MLC and TLC can be the same, but the peripheral circuit design isn't (min # of latches = # of bits per cell). More bits per cell designs are also likely to have more spare/ECC bytes per page.
In layman terms, TLC and MLC can be run as pseudo-SLC by only programming the lower page, but it's impossible to run e.g. MLC as TLC. But pseudo-SLC isn't the same thing as native SLC.
SLC/MLC chips are physically the same chips during the fab process. Maybe some Implant changes but die layout is the same. In Probe, circuits are blown to make them SLC or MLC.
The referenced quote in this thread, is that a due to the usage of die vs chip/package. The chips in the images can have multiple die in one single chip/package and thus the 'density' of the chip/package changes. More layers still means more bits/mm^2.
I'm wondering, would it be worth adding a PS4 Pro/'BoneX benchmark to SATA SSD reviews? The 1TB is fast becoming a worthwhile, if luxurious, upgrade for the top consoles. It may be slightly unaffordable now, but possibly not for long?
Generally, any good SATA SSD is going offer much the same performance as any other model. The gains when going from the second (or even fifth) best SSD on SATA to the absolute best are tiny. Plus, given the capacity you would want for a console, I think going for a 'mediocre' SSD will offer all the performance you need without spending twice as much for less than one second load time differences.
this is true once you have a SSD (any) in a console your purely limited by the CPU speed and ram in the PS4 or xbox it self (some youtubers have tested it) even to the point the SSHD from seagate and toshiba after the second load is nearly the same as a SSD (but it has to have read the data before so if you load the same game or save twice for it work well witch mite happen if you die a lot and fall back to a checkpoint save) if you switch between games SSHDs can at times offer not as much benerfit
still can't beat SSDs for consistency as it always be the same speed
Say Mr. Tallis, I'm sure I'm not alone when I ask this (feel free to chime in, guys) but there are those of us out here who want the greater capacity of a 1tb drive whilst not wanting to fork over 1tb drive prices. I can RAID 0 a pair of ~500gb drives and get better performance than a single 1tb drive, while spending $60-$120 less. What I'm saying is, I would love to see how your arsenal of ~500 gb drives perform in RAID. I have a feeling others would like to see the same information.
You can't get a pair of 500GB drives for $60 less than the price of a 1TB drive from the same product line, unless you're looking at an unusually good sale on the 500GB drives and not looking for the best price on a 1TB drive. Every 1TB SATA drive I price checked this week was cheaper per GB than its half-TB counterpart.
As for RAID testing, my collection of drives almost never includes more than one of each. At the moment I'm finishing up a NVMe RAID review using a set of drives that was loaned by a vendor specifically for this review. Between the disappointing results I'm getting from that project and the prices I'm seeing that don't make SATA RAID economical either, it's not a priority for me to test SATA RAID.
Is that a software raid or hardware raid you used for your NVMe raid review. I'm guessing software as I'm not even sure any NVMe raid solutions exist. In that case, I imagine we're back to needing dedicated RAID controller to deal with the order of magnitude increase in drive performance compared to HDD's, and then obviously you're left questioning who needs double the increase in sequential performance when a single drive is usually more than fast enough for any sequential workload.
Check m.2 PCIe/NVMe 1TB and 2TB prices. But you'll need an MB with 2 M.2 slots with 4x PCIe connected to CPU, are those even available? Otherwise latencies will dominate the performance (like for M.2 which only provide SATA, either slots or drives).
I would like to see a review of old SSDs (Indlinx Barefoot, first Intel SSDs, Sandforce SSDs, crucial c300, Samsung 830 for example) and how they would compete against today's drives. The test parcours is much more detailed and challenging for the drives. Can older drives pass this test reasonable good enough or is an upgrade to a modern ssd an major improvement, besides of capacity.
Most decent SLC SATA drives should be fine, but there were terrible examples like Kingston V-series which were not much better than HDDs to begin with. But I'd like to see the tests too.
am using your charts to compare the previous results of the MX200 vs the "new" MX500, and most of the numbers and data have changed to show the MX200 as "better" than it once was, if properly tested, this should not have been shown as such.
IMHO, if a drive at a specific time using a specific set of drivers etc had say a data rate of 217.85 (crucial MX200 500gb) why does this review "change" to show this same drive now being 235 (a gain of around 8%)
numbers should not be becoming "magically better" IMO, unless firmware was changed without notification, or tests were not run properly the first time around or something like that?
it is quite hard to do direct comparisons when your data does not agree with itself ^.^
MX500 is supposed to be "a fair bit quicker" in most things than MX200 or MX300 (around 18% for the 500gb models..20 odd % for the 1tb models) something is not agreeing with itself based on your data/numbers/review, this is troublesome to say the least, so I suppose my question in this regard is, how "factual" is the testing methodology being used, such as latency numbers (where the original data of MX200 read higher numbers in ms than this review does making this review appearing as if the mx200 is that much faster than it was previously (less latency) and the power consumption was also higher previously whereas in this review it uses "less" power, the heck is up with this?
if one screwed up original testing methodology "I understand" but, if one just changes data willy nilly to suit vendors or make the older or newer stuff appear better than it actually is, screw that noise (we already have plenty enough companies/corps that do that crud daily..like Ngreedia, Apple, Intel, MSFT etc)
Just saying, had my eye on the MX500, my MX100 256gb has been working very well for me last 2.5-3 years or something like that (still at 99% life) MX200 500gb still snappy (though not quite as snappy as the MX100 for some reason and I have never had it "loaded" currently is at 92% life and yet is at least a 9 month newer drive, have not copied.deleted anywhere close to as much as it claims I have which is confusing...MX500 is supposed to have that much more endurance, but, if the MX200 seems to "chew" endurance more than it should, that is not a good thing.
I have taken all the steps I can to reduce writing to the drive (unless needed) have trim on since day 1 etc...
Anyways, compare the "original" MX200 data you reviewed dated may 22, 2015, with this review dated february 2, 2018, quite a difference in appearances from how so so the drive appeared (MX200) vs the way it "now" appears (seems like not quite a poor drive, though it never was crud drive)
would say IMO, double check your info (past info) so you are not making silly mistakes if possible (I know no one is perfect by all means, but I highly doubt you keep hundreds of drives on hand so you can test them a year or 2 down the line, most review sites do not, so does this mean pulling numbers out of thin air?)
take care, love your site by all means, lots of good reading ^.^
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
25 Comments
Back to Article
PeachNCream - Friday, February 2, 2018 - link
"..the per-die capacity of the MX500's 64-layer 3D TLC is actually lower than that of the 32L 3D TLC.."Why is that the case? Shouldn't doubling the number of layers lead to greater capacity and does this mean that there are more ICs per GB in the newer MX500? I'm super confused.
jtd871 - Friday, February 2, 2018 - link
"The Crucial MX500 uses Micron's 256Gb 64L 3D TLC part and consequently returns to the standard drive capacities and overprovisioning ratios, instead of the unusual configurations caused by the 384Gb die capacity in the MX300's 32L 3D TLC. The slightly lower usable capacities of the MX500 than the MX300 means the new drives have slightly higher prices on a per-GB basis, but the MSRPs are still very competitive against current street prices for the competition."So the new chips have more layers per package, but less overall capacity. I'll guess the 384Gb die hamstrung performance on relatively smaller-capacity drives by offering fewer memory channels for the processor to work with simultaneously. Plus, I'll guess that it was a bit of an oddball size for the algorithms to deal with.
FunBunny2 - Friday, February 2, 2018 - link
large node?more area dedicated to control functions?
Billy Tallis - Friday, February 2, 2018 - link
For the 32L node, IMFT was more or less prioritizing MLC use cases by making a 256Gb MLC die that could be used as a 384Gb TLC part. For the 64L node, TLC is the priority and they're making both 256Gb TLC and 512Gb TLC parts. The latter should be cheaper per GB when it is available, but would be a worse choice for small consumer drives. The 256Gb die is really tiny, which makes it more appealing to the mobile market than a lot of previous IMFT parts.FunBunny2 - Saturday, February 3, 2018 - link
"making a 256Gb MLC die that could be used as a 384Gb TLC part."so... does this mean that NAND is all the same, and it's the controller that decides among S/M/T?? or is it the case that S can be coerced to M, and M to T, but not the other way round? is there a tute, here or elsewhere, that tells us the nittygritty about why titular NAND can/can't be used at other densities?
FunBunny2 - Saturday, February 3, 2018 - link
well, I did find one, but from 2012 (and AT: https://www.anandtech.com/show/5067/understanding-... )"This array can be turned into either SLC, MLC, or TLC. The actual array and transistors are equivalent in all three flash types; there is no physical difference. "
some/many folks have been claiming that TLC, in particular, is physically different from SLC. is that now true?
Kristian Vättö - Sunday, February 4, 2018 - link
In terms of memory cell design, SLC, MLC and TLC can be the same, but the peripheral circuit design isn't (min # of latches = # of bits per cell). More bits per cell designs are also likely to have more spare/ECC bytes per page.In layman terms, TLC and MLC can be run as pseudo-SLC by only programming the lower page, but it's impossible to run e.g. MLC as TLC. But pseudo-SLC isn't the same thing as native SLC.
Spatty - Wednesday, February 14, 2018 - link
SLC/MLC chips are physically the same chips during the fab process. Maybe some Implant changes but die layout is the same. In Probe, circuits are blown to make them SLC or MLC.The referenced quote in this thread, is that a due to the usage of die vs chip/package. The chips in the images can have multiple die in one single chip/package and thus the 'density' of the chip/package changes. More layers still means more bits/mm^2.
PeachNCream - Monday, February 5, 2018 - link
Thanks Billy! It makes more sense now.colonelclaw - Friday, February 2, 2018 - link
I'm wondering, would it be worth adding a PS4 Pro/'BoneX benchmark to SATA SSD reviews? The 1TB is fast becoming a worthwhile, if luxurious, upgrade for the top consoles. It may be slightly unaffordable now, but possibly not for long?jordanclock - Friday, February 2, 2018 - link
Generally, any good SATA SSD is going offer much the same performance as any other model. The gains when going from the second (or even fifth) best SSD on SATA to the absolute best are tiny. Plus, given the capacity you would want for a console, I think going for a 'mediocre' SSD will offer all the performance you need without spending twice as much for less than one second load time differences.leexgx - Tuesday, February 20, 2018 - link
this is true once you have a SSD (any) in a console your purely limited by the CPU speed and ram in the PS4 or xbox it self (some youtubers have tested it) even to the point the SSHD from seagate and toshiba after the second load is nearly the same as a SSD (but it has to have read the data before so if you load the same game or save twice for it work well witch mite happen if you die a lot and fall back to a checkpoint save) if you switch between games SSHDs can at times offer not as much benerfitstill can't beat SSDs for consistency as it always be the same speed
tamalero - Friday, February 2, 2018 - link
Those prices feel really nice!Hurr Durr - Sunday, February 4, 2018 - link
It's not nice until we hit around 10 cents for the gigabyte.zirk65 - Friday, February 2, 2018 - link
Looks similar to the MX200, but with bigger pipes and better thermals.MX200 = MLC NAND / Marvel Controller
MX500 = TLC NAND / Silicon Motion Controller
Yet I wonder much is different between the controller uArch these days, outside of specific I/O and power functions.
mode_13h - Friday, February 2, 2018 - link
Last I checked, Crucial's MX-series SSDs featured end-to-end data protection. Does Samsung offer anything like that, in their consumer drives?letmepicyou - Friday, February 2, 2018 - link
Say Mr. Tallis, I'm sure I'm not alone when I ask this (feel free to chime in, guys) but there are those of us out here who want the greater capacity of a 1tb drive whilst not wanting to fork over 1tb drive prices. I can RAID 0 a pair of ~500gb drives and get better performance than a single 1tb drive, while spending $60-$120 less.What I'm saying is, I would love to see how your arsenal of ~500 gb drives perform in RAID. I have a feeling others would like to see the same information.
Billy Tallis - Friday, February 2, 2018 - link
You can't get a pair of 500GB drives for $60 less than the price of a 1TB drive from the same product line, unless you're looking at an unusually good sale on the 500GB drives and not looking for the best price on a 1TB drive. Every 1TB SATA drive I price checked this week was cheaper per GB than its half-TB counterpart.As for RAID testing, my collection of drives almost never includes more than one of each. At the moment I'm finishing up a NVMe RAID review using a set of drives that was loaned by a vendor specifically for this review. Between the disappointing results I'm getting from that project and the prices I'm seeing that don't make SATA RAID economical either, it's not a priority for me to test SATA RAID.
Wardrop - Saturday, February 3, 2018 - link
Is that a software raid or hardware raid you used for your NVMe raid review. I'm guessing software as I'm not even sure any NVMe raid solutions exist. In that case, I imagine we're back to needing dedicated RAID controller to deal with the order of magnitude increase in drive performance compared to HDD's, and then obviously you're left questioning who needs double the increase in sequential performance when a single drive is usually more than fast enough for any sequential workload.peevee - Monday, February 5, 2018 - link
Check m.2 PCIe/NVMe 1TB and 2TB prices.But you'll need an MB with 2 M.2 slots with 4x PCIe connected to CPU, are those even available? Otherwise latencies will dominate the performance (like for M.2 which only provide SATA, either slots or drives).
KarlKastor - Saturday, February 3, 2018 - link
I would like to see a review of old SSDs (Indlinx Barefoot, first Intel SSDs, Sandforce SSDs, crucial c300, Samsung 830 for example) and how they would compete against today's drives. The test parcours is much more detailed and challenging for the drives. Can older drives pass this test reasonable good enough or is an upgrade to a modern ssd an major improvement, besides of capacity.velanapontinha - Saturday, February 3, 2018 - link
^this^peevee - Monday, February 5, 2018 - link
Most decent SLC SATA drives should be fine, but there were terrible examples like Kingston V-series which were not much better than HDDs to begin with.But I'd like to see the tests too.
Dragonstongue - Thursday, June 21, 2018 - link
am using your charts to compare the previous results of the MX200 vs the "new" MX500, and most of the numbers and data have changed to show the MX200 as "better" than it once was, if properly tested, this should not have been shown as such.IMHO, if a drive at a specific time using a specific set of drivers etc had say a data rate of 217.85 (crucial MX200 500gb) why does this review "change" to show this same drive now being 235 (a gain of around 8%)
numbers should not be becoming "magically better" IMO, unless firmware was changed without notification, or tests were not run properly the first time around or something like that?
it is quite hard to do direct comparisons when your data does not agree with itself ^.^
MX500 is supposed to be "a fair bit quicker" in most things than MX200 or MX300 (around 18% for the 500gb models..20 odd % for the 1tb models) something is not agreeing with itself based on your data/numbers/review, this is troublesome to say the least, so I suppose my question in this regard is, how "factual" is the testing methodology being used, such as latency numbers (where the original data of MX200 read higher numbers in ms than this review does making this review appearing as if the mx200 is that much faster than it was previously (less latency)
and the power consumption was also higher previously whereas in this review it uses "less" power, the heck is up with this?
if one screwed up original testing methodology "I understand" but, if one just changes data willy nilly to suit vendors or make the older or newer stuff appear better than it actually is, screw that noise (we already have plenty enough companies/corps that do that crud daily..like Ngreedia, Apple, Intel, MSFT etc)
Just saying, had my eye on the MX500, my MX100 256gb has been working very well for me last 2.5-3 years or something like that (still at 99% life) MX200 500gb still snappy (though not quite as snappy as the MX100 for some reason and I have never had it "loaded" currently is at 92% life and yet is at least a 9 month newer drive, have not copied.deleted anywhere close to as much as it claims I have which is confusing...MX500 is supposed to have that much more endurance, but, if the MX200 seems to "chew" endurance more than it should, that is not a good thing.
I have taken all the steps I can to reduce writing to the drive (unless needed) have trim on since day 1 etc...
Anyways, compare the "original" MX200 data you reviewed dated may 22, 2015, with this review dated february 2, 2018, quite a difference in appearances from how so so the drive appeared (MX200) vs the way it "now" appears (seems like not quite a poor drive, though it never was crud drive)
https://www.anandtech.com/show/9258/crucial-mx200-...
https://www.anandtech.com/show/12263/the-crucial-m...
would say IMO, double check your info (past info) so you are not making silly mistakes if possible (I know no one is perfect by all means, but I highly doubt you keep hundreds of drives on hand so you can test them a year or 2 down the line, most review sites do not, so does this mean pulling numbers out of thin air?)
take care, love your site by all means, lots of good reading ^.^
dpjtpa - Friday, June 14, 2019 - link
This is definitely the option I will be purchasing soon