Team Group GE Pro 2 TB

[W1zzard]

The new Team Group GE Pro M.2 NVMe SSD with Innogrit's IG5666 controller enters the PCIe Gen 5 arena, challenging the Phison E26. In our review, we put it through synthetic and real-life tests, and measure power draw to see if it’s a worthy competitor.

Show full review

 

[GabrielLP14]

That "amazing" power efficient..... gives me the chills LOL

 

[Philaphlous]

Goodness... 15W power consumption just for a SSD??? My entire laptop is currently on battery power and its drawing 7.5W from the battery...screen and all... pretty sad performance.

 

[oiawshboiwsahnoidaws]

That idle power consumption... just wow

 

[chrcoluk]

Ok here is my thoughts.

No thermal analysis test, I guess this is for the reasons mentioned at start, how do you test it fairly? Maybe test it on the lowest end board you can find and stick it under a slab pretending to be a heatsink, probably not a great result, but press the point SSD manufacturers need to be shipping heat sinks on high wattage drives.

I noticed the fsync graph, something I dont recall noticing before on older reviews, I had 2 observations from this interesting test. (a) my OS drive did very badly on it (980 pro), the OS is important in this area as it does many fsync writes for registry and logs, but perhaps even more interesting the top half a dozen or so drives were also drives that had high idle power so wake up latency may be having a bearing. So are some vendors intentionally disabling low power states to get better performance?

Dont like the way the NAND storage industry is going, its like they have adopted a enterprise attitude where heat and power consumption dont matter, this drive hitting 15W is suddenly making my enterprise DC P4600 seem an efficiency king in comparison.

 

[W1zzard]

I noticed the fsync graph, something I dont recall noticing before on older reviews

Thsi was added a few months ago

I had 2 observations from this interesting test. (a) my OS drive did very badly on it (980 pro), but perhaps even more interesting the top half a dozen or so drives were also drives that had idle power

I don't think idle power and QD1 FSync IOPS are correlated. But it's definitely true that top rated drives could perform poorly here, which is why I've added this test. For example, if you run MySQL and wonder "I'm stuck at 500 MB/s, but the drive can do 5000 MB/s? What's going on?" FSync QD1 is killing you. For every transaction, to provide consistency guarantees, MySQL will flush the data to disk, so that a power outage can't revert the transaction. For example, if the power goes out/system crashes, on your bank's DB servers, your account will still have the money that it received 0.1 seconds ago.

 

[InVasMani]

Random I/O good while price and power not so appealing. If pricing becomes a bit more favorable later it might end up being a pretty good option. It was good to see the random I/O was favorable at least.

 

[Chrispy_]

That price for YTMC NAND and that power draw make this an "avoid at all costs"

At least until the price comes down and they fix the horrible power draw in firmware.

 

[edlee]

where is the most important chart, thermal throttling and thermal image/hotspot?

Im guessing byleaving it out, it was really terrible.

I dont believe its an excuse that they didnt supply a heatsink, since most nvme drives dont come with heatsinks, so why is that a pass on this drive??

i want to see unaided heat temp numbers, and pick one nvme heatsink use that for all your testing

Thermalright HR-09 Pro and HR-10 Pro are excellent and cheap options you can use to test these gen5 nvme drives

 

[W1zzard]

where is the most important chart, thermal throttling and thermal image/hotspot?

As explained, no heatsink included, so how should I test thermals?

Im guessing byleaving it out, it was really terrible.

Based on the power numbers, this will be the most-demanding SSD in terms of cooling

since most nvme drives dont come with heatsinks

Gen 4 without heatsink I just test without heatsink. Gen5 will overheat within 15 seconds

Thermalright HR-09 Pro and HR-10 Pro are excellent and cheap options you can use to test these gen5 nvme drives

Hmm .. good idea .. maybe I could establish some kind of "standard heatsink" testing.. HR-10 Pro looks interesting, but I rather have something without a fan. I just ordered a HR-10 non-Pro, will update the review late next week

 

[edlee]

based on your review of the MSI m480 pro, i bought it and its been great.

And i decided not to use the motherboard heatsink since i believe it will still thermally throttle the drive. and i bought a Thermalright HR-09 Pro, since its passively cooled heatsink and i didn't need to run another cable for the fan which the HR-10 pro uses. And the temperature does not exceed 55 degrees when hammering the read and writes.

 

[sLowEnd]

That power consumption is nuts for an m.2 ssd
Completely unacceptable

 

[edlee]

As explained, no heatsink included, so how should I test thermals?


Based on the power numbers, this will be the most-demanding SSD in terms of cooling


Gen 4 without heatsink I just test without heatsink. Gen5 will overheat within 15 seconds


Hmm .. good idea .. maybe I could establish some kind of "standard heatsink" testing.. HR-10 Pro looks interesting, but I rather have something without a fan. I just ordered a HR-10 non-Pro, will update the review late next week

The HR-09 Pro gets much better thermals than the HR-10 (non-pro), and it doesnt have a fan

 

[ARF]

And i decided not to use the motherboard heatsink since i believe it will still thermally throttle the drive.

Smart decision.
I suspect that the motherboard "heatsink" with a "thermal pad" successfully killed my Lexar NM790.
Today, I use Silicon Power XPower XS70 with its own stock heatsink, and for some reason my computer feels more responsive.

 

[Wirko]

Thsi was added a few months ago


I don't think idle power and QD1 FSync IOPS are correlated. But it's definitely true that top rated drives could perform poorly here, which is why I've added this test. For example, if you run MySQL and wonder "I'm stuck at 500 MB/s, but the drive can do 5000 MB/s? What's going on?" FSync QD1 is killing you. For every transaction, to provide consistency guarantees, MySQL will flush the data to disk, so that a power outage can't revert the transaction. For example, if the power goes out/system crashes, on your bank's DB servers, your account will still have the money that it received 0.1 seconds ago.

Anyone who needs high performance with fsync should seek an enterprise SSD with actual hardware-based PLP, such as the Kingston DC2000B, announced just yesterday at TPU. Those have their PCBs filled with capacitors, hence fewer NAND chips and lower memory capacity. Other SSDs have software-emulated PLP, it probably works, but fsync IOPS is its very weak point.

Also, database engines can recover data from if the power was cut, and writing interrupted, at any given moment. Only uncommitted transactions are lost, as they should be. (I know that about Oracle but other DBs keep transaction log files with the same purpose.) That's not even the greatest danger that fsync guards against. The crucial part here is that data has to be written to permanent storage in the same order the server sent it for writing. If a certain sector is written but one that arrived earlier is stuck in some cache and lost, the mess is complete. I think the same is true of journaling filesystems such as NTFS. This is what fsync guards against.

And of course, power loss is just unacceptable in any application that juggles valuable data, financial or not.

 

[Hagal77]

I may have overlooked it, but the most important aspect, i.e. heat development, is unfortunately missing from the test. The temperature problem is what makes the Phison bad, I can't find anything about it in the test, why? It's clear that no heat sink is fitted at the factory. But you could have taken a standard one and the values. With PCiE5 NVME I am less interested in the speed, the non plus ultra are the temperatures.

greetz

 

[MachineLearning]

Thank you very much for the review.

One small error on components page:

The Innogrit IG5666 son PS5026-E26 is Phison's first PCI-Express 5.0 controller.

Phison gobbling up the competition, I see

 

[W1zzard]

Anyone who needs high performance with fsync should seek an enterprise SSD with actual hardware-based PLP, such as the Kingston DC2000B, announced just yesterday at TPU

but the $$

Also, database engines can recover data from if the power was cut, and writing interrupted, at any given moment. Only uncommitted transactions are lost, as they should be.

correct, that's because of fsync. if you turn that off, you will lose that guarantee. innodb_flush_log_at_trx_commit

And of course, power loss is just unacceptable in any application that juggles valuable data, financial or not.

and it still happens

But you could have taken a standard one and the values

which one is the standard?

One small error on components page:

Whoops, fixed, I forgot to update that section

 

[Knight47]

Smart decision.
I suspect that the motherboard "heatsink" with a "thermal pad" successfully killed my Lexar NM790.
Today, I use Silicon Power XPower XS70 with its own stock heatsink, and for some reason my computer feels more responsive.

Dont people say that Gen4 drives don't need heatsink whatsoever?

 

[Hagal77]

Clear. Passive, a small heat sink and or what comes with the mainboard anyway would be enough and then compare with a Phison.

 

[W1zzard]

Clear. Passive, a small heat sink and or what comes with the mainboard anyway would be enough and then compare with a Phison.

You are aware that all Phison E26 drives have an active, fan-cooled heatsink? What about fan noise? Some of these Phison drives run crazy whining fan speed, others run almost noise less

 

[RootinTootinPootin]

Can pass as just a regular data drive, its good to have DRAM as well, kinda "uhhmmm ok" vibe I am getting. hope this is priced right from where I am at.

 

[starfals]

The power consumption :O Its insane!

 

[ARF]

Dont people say that Gen4 drives don't need heatsink whatsoever?

Gen 4 drives definitely do need heatsinks.

 

[Hagal77]

You are aware that all Phison E26 drives have an active, fan-cooled heatsink?

i have a Corsair MP700 PRO in my Threadripper Sys, this thing gone extream hot. I have dismantled the roar cube. It's only passively cooled via the AL plate from the TRX50 Aero D, which is so-so, but great is different, now I run the heavy workloads via the KC3000 and the Aorus Gen4.
I wont only a comparison between Phsion E26 and the Inno with the same heat sink, that would be great.
Too bad there is no 8TB version, would be interesting to test.

its good to have DRAM

Yes. Without DRAM, no buy

Gen 4 drives definitely do need heatsinks.

True, my Seagate Firecuda 520 (Server) and Aorus Gen4 Copper gone fast to hot, without Heatsink no way.