NVMe throttling from 2GB/s to 0.5GB/s is usually not thermal, it is an SLC cache exhaustion that brings a drive back to MLC/TLC packing mode. One can get samsung pro or a similar ssd and it will retain 2GB/s indefinitely. On cheaper drives, having a lot of free space (i.e. slc cache) may help. Not sure though what process can generate such sustained write bandwidth beyond few things like copying or video transcoding.

Which is why I replaced the cheap drive in my build desktop with a Samsung Pro NVME. It made a significant difference.

Multithreaded builds (make -j 24) can really hammer the drive. Read and write interleaved, which uses up cache in both directions.

I've used their Pro NVME drives as my OS drive for the last few years. Never had any issues with read write speeds either

Two words: Heat Sinks

A well proven way to move heat out of a silicon package to prolong its ability to perform at its highest potential. And in a desktop, you've got room for 'em.

What I’m told you need heatsink for PCIe SerDes and maybe CPU on it but NAND likes warmer temperature

On an NVME? You don't need heatsinks of an NVME. You have much bigger issues in your case if you think heatsinks will help on an NVME.

Are you suggesting NVME heatsinks are a marketing gimmick? I beg to differ.

No, that's wrong. A heatsink on an NVMe helps to let it run its top speed longer, or forever. Has nothing to do with issues in your case, those things just get warm, and before it gets too hot the heat has to be transported away or they will throttle.

Specifically, a lot of NVMe are right next to GPU placed in PCI slot 1, which is a pretty constant source of heat under load.

I installed an EK spreader sandwich on my Samsung NVMe drive, and it made a massive (20° c) difference over stock. It was previously a bare stick with no surface area/thermal sink to pull heat away.

What's an "EK spreader sandwich"? Can you send a link to yours as well?

No. NVME's are rated to run at 0-70 degrees. If your case makes your NVME run at > 80+ degrees you should prob fix the airflow in your case as your prob running your CPU/GPU rather hot.

You do not need a case for that! The NVMe heats up on its own. There are a bunch of reviews going into that topic, because it just matters for high end NVMe drives. See for example https://www.computerbase.de/2020-09/samsung-980-pro-ssd-test... - though the 980 Evo did well without an additional cooler, and here it's indeed depending on the airflow.

You seem to assume it's the case that heats up to 80C or more, that's not true and also not necessary for NVMe SSDs to go over their limit. Your CPU/GPU has fans moving the heat away, that's a better position to be in...

I said if your nvme is reaching 80+ degrees then you should fix the airflow in the case. I did not say the case is 80 degrees. Any semi decent airflow is 100% capable of running an nvme without causing any performance issues.

As you saw in the article I linked that's just not true. Those temps are with airflow, and SSDs go over their temp limit.

Nope. That article states the the ssd was 'naked' which leads me to believe they removed the sticker off it. Those stickers/labels are use to dissipate heat from the controller, so you, you know, don't need a heatsink. They act as a heatsink.

https://www.youtube.com/watch?v=KzSIfxHppPY&t=375s

No! Why would you assume crazy modifications like that? They write "naked", it's naked as in not using the heatsink the mainboard supplies. https://www.gigabyte.com/us/Motherboard/X570-AORUS-MASTER-re... shows the cooler they are talking about.

You will just have to accept that you've been wrong on this topic. Move on, it happens.

No. Lol. Cooling anything other than the controller is pointless. You don’t get throttling from nand. I have the gigabyte x570 elite. That nvme cover does nothing. And unless you’re benchmarking the nvme I highly doubt you’re going to heat up the controller enough to notice it. So yup. I agree. You’re wrong. Let’s move on.

Man, they are benchmarking the whole SSD. I actually sent you the graph. The SSDs go over 80C and then throttle. Not all of them, but enough of them to be a real problem. The NVMe covers actually work in reducing the temperature, that's also in the graph. You are wrong and you seem to have no knowledge about this topic at all, you really should not be so confident.

Read and learn! And then accept when an initial assumption turns out to be wrong.

It’s not an assumption. NAND does not throttle. The controller does. Unless you run the SSD constantly like in a benchmark you are unlikely to heat the controller up to make it throttle. The airflow of a case is enough to keep an nvme within limits that everyday usage is not going to be hindered in anyway shape or form. You don’t need a heat sink. Linking to a constant load benchmark doesn’t change that. You’re wrong.

I do not understand how you can still think that after my explanation and after looking at the graph I sent you. One very last try - though I assume in the best case it's for potential other readers.

Go to https://www.computerbase.de/2020-09/samsung-980-pro-ssd-test.... look at the graph. You see that a bunch of them go to the 80C line or hover above. All of them throttle (what you said does not happen). In that graph are shown, going above the limit:

1. FireCuda 520 1TB

2. Patriot Viper VP4100 1TB

3. Samsung 970 Evo 1TB

4. WD Black SN750 1TB (+ the same one with a cooler)

This is only a small part of the market of course, but it goes to show that the throttling is a real thing that happens with multiple models.

Then you had a moving goalpost there, that those SSDs do not throttle under realistic workloads. However, this is a sequential read that's only 5 minutes long. Hardly unrealistic. The hour long constant load benchmark is a different graph, however, constant load is also realistic if it's longer than 5 minutes.

If you activate the other chart modes you see the measured performance, which shows the drops linked to the too high temperature, and that they did the same thing for write performance.

You can counteract this with a lot of targeted airflow and/or a heatsink, the heatsink will at least help move the throttling to a later moment. Gamersnexus had a very impressive demonstration of this, one where they did get this wrong: They had an article about a MSI SSD heatsink where they claimed it did not help (so the SSD did throttle! Again something you said does never happen), where it then turned out that it did not work only because their applied temperature sensors (glued them to the heatsink), and IIRC they also missed the higher performance they got regardless. GN often gets it right, stuff like that happens, but it made this one memorable and highlighted the positive effect of these heatsink coolers.

I'm into this topic professionally for years now. I'm not wrong here. If you can't take my word for it, look at professional SSD reviews, they have covered this also for years now.

And sure: There are scenarios where this does not matter. Gaming. Browsing. But: In those workloads there is no significant difference to a SATA SSD anyway. These NVMe SSDs are only interesting if you have large (and thus: long) file transfers. This is what they have to get right (and some do, but not all of them).

It’s like you’re not even reading what I’m saying. So I’ll just end it here. You’re wrong.

I'm not wrong, you are.

By the way, by repeating that I'm wrong and by starting with a straight "No", by always commenting without reasoning and politeness, you made sure that I will correct you - and that I'm not buying into your strange attempts to correct your statements to something that is correctish. They don't work anyway, these SSDs throttle.

You should change your tone around here.

> They don't work anyway, these SSDs throttle.

But you don't know what causes the throttling. Its stupid to shove a heatsink on nand. It does not help. Not a single bit. Period. Under any normal daily usage, or even if you had a workstation, you're not going to be reading/writing so constantly frequently that you're ever going to cause the controller to heat up and cause throttling. If you experience any excessive heat. You have bigger issues in your case. Period. Your only proof of throttling is benchmarks running constant read/writes over a period of time. This is not real world usage and doesn't make it necessary to go out and start shoving heatsinks on every single nvme drive. If that was the case then all the laptops which have space between the nvme and the case, or motherboards which lack a 'heatshield' like the gigabyte board you linked to, would have throttling issues. Which they don't.

> You should change your tone around here.

So now you're threatening me?

-----

Anyway I'm done, not gonna sit here an argue anymore.

How is the lifetime of a SSD run around 70° vs the same SSD with a slightly bigger heatsink that runs at 60°?

I would run some tests though.

There are all kinds of vendors / models, who knows what kind of throttling they use?

The point being let the user decide.

Having a fast machine that can sustain throughput and I/O is a perfectly rational desire, and for some of us, need.

If you do a lot of video work, daily!

Every time you recompile. Many times per day.

That must be one fast compiler.

Do your compiles write multiple gigabytes of data? Within seconds?

If you're running multithreaded build jobs with ninja and have many cores then maybe?

Unless you compile multi-gigabyte targets, all writes will likely fit in a ram cache and thus cannot be a real bottleneck. That is assuming your compiler farm can read and compile at GB/s level, which is pretty unrealistic.

To test that, one can try it with ramdisk first, before getting an expensive ssd.

And that is something else that is alien to laptops. Best laptop I could find (for when I am away from my desktop) is 8c16t, that makes a massive difference for compilation.