Interesting article as always. I had been hoping for a larger price gap between this and the Samsung 950 Pro. At current prices, I think the choice is fairly obvious unless you need a 1TB SSD.
"Unlike most cheap adapter cards, the RD400's adapter draws power from the PCIe slot's 12V supply and converts it to the 3.3V required by the M.2 drive."
PCIe slots provide 10W of 3.3V power directly. (I believe this was originally done to make converting legacy cards via a bridge chip easier.) Why would the card need to do any DC-DC conversion?
I'd assume that if they convert they can get cleaner 3.3V than if they use the feed through the PCIe slot, or that they can design in better resilience to sudden power demand changes...
Basically, by converting I would have thought they can better fit the power supplied to the drive to the demands it makes.
Yeah but they're filtering the power either way so converting from 12V -> 3.3V is less efficient than using 3.3V to start with. But getting back to what Byrn was saying... Byrn, they don't have a choice: This drive draws too much power to use the 3.3V supply.
Look at idle power figures in this article. ~2.5W @ 12V. At 3.3V that would already be pushing it (right around 9W already). Under a load it's going to draw too much. So they had to use the 12V rail.
I'm sure a multi billion dollar conglomerate decided to convert the 12v PCIe rail over drawing from the weak 3.3v rail for a reason. Probably something as simple as firmware flashing compatibility. Flashing firmware causes enormous voltage spikes that would easily surpass 10 watts @ 3.3v
It's a bit weird to state that the Samsung is the fastest when it loses in write and mixed workloads. Your own benchmark is read heavy but maybe not all readers care most about read perf.
Read seems to be the most common trait people look for, so if they had to pick one I would agree on that. Kinda odd to give it the title though with no single drive leading in more than a couple benchmarks, it's only "the fastest" in relatively specific categories
The Samsung 950 Pro is clearly faster on the ATSB Heavy test which writes more than twice as much data as it reads. If you have a workload that is so much more write heavy that the RD400 comes out ahead, then it's quite atypical and you shouldn't expect our general-audience recommendations to apply.
If you're mostly writing data that never gets read back again, you're one of the use cases that really still does work well with much cheaper spinning rust.
And yet you have more read IOs. Would be interesting to see what you get if you remove at least games and web from that test. Another factor to consider is when you need the perf and when you don't. I wouldn't buy a drive like this for web browsing or watching video.Would prefer the drive not to choke when you really put pressure on it. Write is also trailing far behind in data rates and that's not great, do you really like such an unbalanced drive? When your verdict is all about read perf ,maybe a more granular verdict is better. If you would do the math in % for read, write and mixed who wins? Maybe i am being lazy but can't seem to find any info on multitasking and the system used. Do any of the tests run multiple things at once and are you using an 8 cores? With Zen arriving soon 8 cores/16 threads should become much more popular as the die should be pretty small -y guess somewhere between 99 and 131mm2. The Samsung feels like a mobile SoC that throttles.Does great when you don't really need the perf but lets you down when you need it. That being said, looking forward to Samsung's new drives (961), on paper those are much better.
Have you actually used an SM951 or 950 Pro? I've been running one for a while and I don't find the SoC comparison very accurate, mine has a decent amount of airflow going over it (tho it's still near a hot GPU)...
AT's own tests proved throttling, when it does happen, wasn't a big deal.
I guess if you're constantly dumping GBs upon GBs of data on one from an equally fast source then yeah, it's gonna be throttle city, but that's not the usage case I see for them.
You misunderstood, the comparison wasn't about the cause but the effect. Here the weakness is write and mixed workloads but the effect is similar, it lets you down when you need it most.
A double-sided M.2 2280 can usually fit four packages of flash, each containing a stack of up to 16 dies that are typically 128Gb (16GB). That multiplies out to a practical limit of 1024GB for now. Newer 3D NAND such as Micron's will be available in 256Gb MLC dies, enabling 2TB M.2 2280 drives (or 3TB with TLC).
We're not quite there yet. Micron's 3D TLC is 384Gb and everybody else seems to be going for a 256Gb TLC that will be a smaller die than their 256Gb MLC. A 4TB M.2 would require either a 512Gb die or denser packaging.
"Upgrading from a mechanical hard drive to a SSD alleviates a major performance bottleneck but the experience of moving from SATA SSDs to PCIe SSDs is not as revolutionary. I suspect most consumers would be better served with a larger SSD of moderate performance than a cramped but blazing fast PCIe drive," Thank you!
So are we saying NVMe is only really useful for enterprise applications? There just aren't consumer use cases where drive speed is now the limiting performance factor?
Is the flash controller made on the same memory process or is it made on a separate logic process? I think its made on a separate (logic) process and if so would that be 28nm for most controllers? And is the manufacturing out sourced to TSMC or in-house for most?
The PCIe NVMe controllers are mostly 28nm from what I've heard. SATA controllers can be anything from 40nm to +55nm. Like nearly all logic manufacturing, it's outsourced to TSMC and the like.
Recently got the SM950 pro 512. Large writes slow down after 30 seconds. It starts out ETA 3 minutes, 10 minutes later it's only 70% complete. I read into it; evidently these M.2 cards heat up and slow down. There is absolutely no heatsink on the card. Running them on a PCI expansion card would allow headspace for small heatsinks.
Are any of you aware of a ribbon cable/riser cable I could use to get this M.2 card off my motherboard and move it to a cooler part of my case? I'm out of PCI slots for these expansion cards.
Even with the degree of thermal throttling I've observed when not using any kind of heatsink, the 512GB Samsung 950 Pro should only take ~12-13 minutes to fill to capacity with sequential writes. I suspect that your bottleneck is whatever is the source of the data being written, not the 950 Pro.
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.
40 Comments
Back to Article
tarqsharq - Wednesday, May 25, 2016 - link
Interesting article as always. I had been hoping for a larger price gap between this and the Samsung 950 Pro. At current prices, I think the choice is fairly obvious unless you need a 1TB SSD.Chaitanya - Wednesday, May 25, 2016 - link
You have a choice of Sandisk X400 as well if you want 1TB capacity in M.2 form factor.edzieba - Wednesday, May 25, 2016 - link
There's also the OEM version of the OCZ drive, the Toshiba XG3, which is also available in a 1TB m.2 SKU.Lord of the Bored - Thursday, May 26, 2016 - link
And who DOESN'T need a 1TB SSD?MrSpadge - Friday, May 27, 2016 - link
I think you're mixing up "want" and "need".DanNeely - Wednesday, May 25, 2016 - link
"Unlike most cheap adapter cards, the RD400's adapter draws power from the PCIe slot's 12V supply and converts it to the 3.3V required by the M.2 drive."PCIe slots provide 10W of 3.3V power directly. (I believe this was originally done to make converting legacy cards via a bridge chip easier.) Why would the card need to do any DC-DC conversion?
Byrn - Wednesday, May 25, 2016 - link
I'd assume that if they convert they can get cleaner 3.3V than if they use the feed through the PCIe slot, or that they can design in better resilience to sudden power demand changes...Basically, by converting I would have thought they can better fit the power supplied to the drive to the demands it makes.
digitalgriffin - Wednesday, May 25, 2016 - link
You can filter any volt feed with enough capacitors. But you lose power efficiency when you do.Alexvrb - Wednesday, May 25, 2016 - link
Yeah but they're filtering the power either way so converting from 12V -> 3.3V is less efficient than using 3.3V to start with. But getting back to what Byrn was saying... Byrn, they don't have a choice: This drive draws too much power to use the 3.3V supply.Look at idle power figures in this article. ~2.5W @ 12V. At 3.3V that would already be pushing it (right around 9W already). Under a load it's going to draw too much. So they had to use the 12V rail.
Wardrop - Wednesday, May 25, 2016 - link
2.5 watts is 2.5 watts. If it's a higher voltage, it's less amps, and vice versa. I think you've confused watts with amps?Samus - Wednesday, May 25, 2016 - link
I'm sure a multi billion dollar conglomerate decided to convert the 12v PCIe rail over drawing from the weak 3.3v rail for a reason. Probably something as simple as firmware flashing compatibility. Flashing firmware causes enormous voltage spikes that would easily surpass 10 watts @ 3.3vjjj - Wednesday, May 25, 2016 - link
It's a bit weird to state that the Samsung is the fastest when it loses in write and mixed workloads.Your own benchmark is read heavy but maybe not all readers care most about read perf.
LostWander - Wednesday, May 25, 2016 - link
Read seems to be the most common trait people look for, so if they had to pick one I would agree on that. Kinda odd to give it the title though with no single drive leading in more than a couple benchmarks, it's only "the fastest" in relatively specific categoriesBilly Tallis - Wednesday, May 25, 2016 - link
The Samsung 950 Pro is clearly faster on the ATSB Heavy test which writes more than twice as much data as it reads. If you have a workload that is so much more write heavy that the RD400 comes out ahead, then it's quite atypical and you shouldn't expect our general-audience recommendations to apply.DanNeely - Wednesday, May 25, 2016 - link
If you're mostly writing data that never gets read back again, you're one of the use cases that really still does work well with much cheaper spinning rust.jjj - Wednesday, May 25, 2016 - link
And yet you have more read IOs. Would be interesting to see what you get if you remove at least games and web from that test.Another factor to consider is when you need the perf and when you don't.
I wouldn't buy a drive like this for web browsing or watching video.Would prefer the drive not to choke when you really put pressure on it.
Write is also trailing far behind in data rates and that's not great, do you really like such an unbalanced drive? When your verdict is all about read perf ,maybe a more granular verdict is better. If you would do the math in % for read, write and mixed who wins?
Maybe i am being lazy but can't seem to find any info on multitasking and the system used. Do any of the tests run multiple things at once and are you using an 8 cores? With Zen arriving soon 8 cores/16 threads should become much more popular as the die should be pretty small -y guess somewhere between 99 and 131mm2.
The Samsung feels like a mobile SoC that throttles.Does great when you don't really need the perf but lets you down when you need it. That being said, looking forward to Samsung's new drives (961), on paper those are much better.
Impulses - Wednesday, May 25, 2016 - link
Have you actually used an SM951 or 950 Pro? I've been running one for a while and I don't find the SoC comparison very accurate, mine has a decent amount of airflow going over it (tho it's still near a hot GPU)...AT's own tests proved throttling, when it does happen, wasn't a big deal.
I guess if you're constantly dumping GBs upon GBs of data on one from an equally fast source then yeah, it's gonna be throttle city, but that's not the usage case I see for them.
At least given the current capacity/costs...
jjj - Wednesday, May 25, 2016 - link
You misunderstood, the comparison wasn't about the cause but the effect. Here the weakness is write and mixed workloads but the effect is similar, it lets you down when you need it most.stux - Thursday, May 26, 2016 - link
I use one of the Sm951s in an AngelBird PCI adapter. Takes care of the throttling issue, it never goes above 41C.And it's fantastic :)
theduckofdeath - Wednesday, May 25, 2016 - link
For a consumer PC read performance is nearly infinitely more relevant than random write performance. That's probably why. :)AnnonymousCoward - Wednesday, May 25, 2016 - link
What about boot time. Is it slow like the Intel card?mervincm - Wednesday, May 25, 2016 - link
Intel 750 SSD isn't slow at boot anymore. Later SSD firmware and NVME drivers have really helped my boot performance.Yregister - Thursday, November 3, 2016 - link
But that's on Windows, correct? I read that the 750 doesn't work on a Mac, not bootable...moheban79 - Saturday, November 12, 2016 - link
Thats not true. I got my Intel 750 booting up my hackintosh. Should be doable.adamto - Wednesday, May 25, 2016 - link
Why there is no 2T or even 4T M.2 SSD?Silma - Wednesday, May 25, 2016 - link
Because there isn't enough place on the stick. One would need to develop much denser NAND.Billy Tallis - Wednesday, May 25, 2016 - link
A double-sided M.2 2280 can usually fit four packages of flash, each containing a stack of up to 16 dies that are typically 128Gb (16GB). That multiplies out to a practical limit of 1024GB for now. Newer 3D NAND such as Micron's will be available in 256Gb MLC dies, enabling 2TB M.2 2280 drives (or 3TB with TLC).Dr.Neale - Wednesday, May 25, 2016 - link
Don't you mean 4 TB with TLC?Billy Tallis - Wednesday, May 25, 2016 - link
We're not quite there yet. Micron's 3D TLC is 384Gb and everybody else seems to be going for a 256Gb TLC that will be a smaller die than their 256Gb MLC. A 4TB M.2 would require either a 512Gb die or denser packaging.Chaser - Wednesday, May 25, 2016 - link
"Upgrading from a mechanical hard drive to a SSD alleviates a major performance bottleneck but the experience of moving from SATA SSDs to PCIe SSDs is not as revolutionary. I suspect most consumers would be better served with a larger SSD of moderate performance than a cramped but blazing fast PCIe drive," Thank you!Meteor2 - Thursday, May 26, 2016 - link
So are we saying NVMe is only really useful for enterprise applications? There just aren't consumer use cases where drive speed is now the limiting performance factor?stux - Thursday, May 26, 2016 - link
This might be the case in Windows, but I've found with OSX, one of the biggest upgrades has been sata3 to PCIe ssd gen 1 to 2 and then 3Ien 0.5 to 1 to 2GB/s
This was evident with all the recent MacBook Pro 15" upgrades and also with PCIe ssds in some Mac Pro towers.
SunnyNW - Wednesday, May 25, 2016 - link
Is the flash controller made on the same memory process or is it made on a separate logic process? I think its made on a separate (logic) process and if so would that be 28nm for most controllers? And is the manufacturing out sourced to TSMC or in-house for most?Ryan Smith - Wednesday, May 25, 2016 - link
Controllers are made on a separate logic process.Kristian Vättö - Thursday, May 26, 2016 - link
The PCIe NVMe controllers are mostly 28nm from what I've heard. SATA controllers can be anything from 40nm to +55nm. Like nearly all logic manufacturing, it's outsourced to TSMC and the like.BangkokTech - Friday, May 27, 2016 - link
Recently got the SM950 pro 512. Large writes slow down after 30 seconds. It starts out ETA 3 minutes, 10 minutes later it's only 70% complete. I read into it; evidently these M.2 cards heat up and slow down. There is absolutely no heatsink on the card. Running them on a PCI expansion card would allow headspace for small heatsinks.BangkokTech - Friday, May 27, 2016 - link
Are any of you aware of a ribbon cable/riser cable I could use to get this M.2 card off my motherboard and move it to a cooler part of my case? I'm out of PCI slots for these expansion cards.Billy Tallis - Saturday, May 28, 2016 - link
Even with the degree of thermal throttling I've observed when not using any kind of heatsink, the 512GB Samsung 950 Pro should only take ~12-13 minutes to fill to capacity with sequential writes. I suspect that your bottleneck is whatever is the source of the data being written, not the 950 Pro.BiTesterEmailer - Wednesday, July 20, 2016 - link
Informative and detailed as always.