Jump to content

Water-cooled flash drive?

GuitarBro
 Share
Posted

More ridiculous stuff has been attempted in the past, so why not try to water-cool a USB flash drive to see if it makes it go any faster under sustained load?

 

Maybe someone with too much free time can test this for science?

Hi there!

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

It would make no sense and would be pointless because you can't overclock the controller on the stick and in general the controllers inside in usb sticks are simple, dumb, without advanced power management schemes, adjusting of frequencies and so on.

Also, you only have a budget of 2.5-4.5w  (5v at 0.5A for usb 2.0 , 5v at 0.9A for usb 3.0) ... assuming the actual flash memory chip needs 0.5-1w, you're left with 2-3 watts of power available for the controller.

Most chips would have around 30-40c/w rise above ambient if there's no cooling around them, but the controller would radiate some of the heat through the circuit board and into the usb connector and maybe on the usb stick case.. but even at 2w consumption (which is silly, very big), you'd have 30c ambient temperature + 2w x 25c/w = ~ 80c chip temperature, for a chip that's rated for 110-125c

 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author

Well, suppose you just really don't like warm flash drives when they're ejected and you have to touch it then.

Hi there!

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
5 minutes ago, GuitarBro said:

Also, considering there are flash drives with SSD controllers, perhaps these could benefit?
 

(Source: https://www.anandtech.com/show/8724/exploring-performance-consistency-in-usb-30-flash-drives)

Those wouldn't be much of a concern still, NVME USB drives do exist and even those don't have any problems with sustaining speeds with good heatsinking. While I guess you can watercool it there wouldn't be a need since you can simply add a couple of heatsinks and call it a day.

http://www.silverstonetek.com/product.php?pid=612

Updated Modding FAQ

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author
1 minute ago, W-L said:

Those wouldn't be much of a concern still, NVME USB drives do exist and even those don't have any problems with sustaining speeds with good heatsinking. While I guess you can watercool it there wouldn't be a need since you can simply add a couple of heatsinks and call it a day.

http://www.silverstonetek.com/product.php?pid=612

Yeah, never said it would be very practical. ;) But if it's possible, there's gotta be someone crazy enough to attempt it.

Hi there!

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
38 minutes ago, GuitarBro said:

More ridiculous stuff has been attempted in the past, so why not try to water-cool a USB flash drive to see if it makes it go any faster under sustained load?

 

Maybe someone with too much free time can test this for science?

I don't think it will be a good idea making Water Cool won't increase its Read/Write Speed...

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

So I use a 128 GB SanDisk UltraFit for a lot of stuff and felt like the speeds were slower when it was a lot hotter. This drive gets not able to be held hot after being in my laptop for a while. 

 

Decided to test it using this method: http://www.binarytides.com/linux-test-drive-speed/ 

Results (of my crude test) were... Not definite. It seems to favor being cool for writing but reads faster while hot. 

Full test included at bottom of post. 

 

This is relevant because I've been looking to cool down my drive (since using/saving "seemed" slower, though I'm unsure after now testing). I can't put something like a heat sink on as it sticks out less than a cm. Haven't given it much thought, but I did stumble upon this: http://thanko.jp/shopdetail/000000002619/ , which would work well for this. Basically, having a water bottle with small fish tank pump going through a tube (which could wrap around the drive). Cheap water cooling with not really any risk.  

 

 

Full test: (Note, started 1st test while too hot to touch. I took out the drive and put it on a metal surface to cool each "------" break. The "sync" lines are to clear the caches to make sure it's actually doing a transfer, not using the cache) 

$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 325.801 s, 3.3 MB/s
$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 137.217 s, 7.8 MB/s
$ dd if=/dev/zero of=./largefile2 bs=1M count=4096
4096+0 records in
4096+0 records out
4294967296 bytes (4.3 GB) copied, 316.554 s, 13.6 MB/s
$ sudo sh -c "sync && echo 3 > /proc/sys/vm/drop_caches"
$ dd if=./largefile of=/dev/null bs=4k
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 9.94723 s, 108 MB/s
$ dd if=./largefile2 of=/dev/null bs=4k
1048576+0 records in
1048576+0 records out
4294967296 bytes (4.3 GB) copied, 38.7523 s, 111 MB/s
----------------------------------------------------------- 
Avg write: 8.23 MB/s         |         Avg read: 109.5 MB/s
----------------------------------------------------------- 
$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 162.251 s, 6.6 MB/s
$ sudo sh -c "sync && echo 3 > /proc/sys/vm/drop_caches"
$ dd if=./largefile of=/dev/null bs=4k
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 17.7766 s, 60.4 MB/s
----------------------------------------------------------- 
Avg write: 6.6 MB/s         |         Avg read: 60.4 MB/s
----------------------------------------------------------- 
$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 55.6096 s, 19.3 MB/s
$ sudo sh -c "sync && echo 3 > /proc/sys/vm/drop_caches"
$ dd if=./largefile of=/dev/null bs=4k
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 17.5442 s, 61.2 MB/s
----------------------------------------------------------- 
Avg write: 19.3 MB/s         |         Avg read: 61.2 MB/s
----------------------------------------------------------- 
$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 24.4734 s, 43.9 MB/s
$ sudo sh -c "sync && echo 3 > /proc/sys/vm/drop_caches"
----------------------------------------------------------- 
Avg write: 43.9 MB/s         |         Avg read: ---- MB/s
----------------------------------------------------------- 
$ dd if=./largefile of=/dev/null bs=4k
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 20.6621 s, 52.0 MB/s
----------------------------------------------------------- 
Avg write: ---- MB/s         |         Avg read: 52.0 MB/s
-----------------------------------------------------------

 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author
11 hours ago, umer936 said:

So I use a 128 GB SanDisk UltraFit for a lot of stuff and felt like the speeds were slower when it was a lot hotter. This drive gets not able to be held hot after being in my laptop for a while. 

 

Decided to test it using this method: http://www.binarytides.com/linux-test-drive-speed/ 

Results (of my crude test) were... Not definite. It seems to favor being cool for writing but reads faster while hot. 

Full test included at bottom of post. 

 

This is relevant because I've been looking to cool down my drive (since using/saving "seemed" slower, though I'm unsure after now testing). I can't put something like a heat sink on as it sticks out less than a cm. Haven't given it much thought, but I did stumble upon this: http://thanko.jp/shopdetail/000000002619/ , which would work well for this. Basically, having a water bottle with small fish tank pump going through a tube (which could wrap around the drive). Cheap water cooling with not really any risk.  

 

 

Full test: (Note, started 1st test while too hot to touch. I took out the drive and put it on a metal surface to cool each "------" break. The "sync" lines are to clear the caches to make sure it's actually doing a transfer, not using the cache) 


$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 325.801 s, 3.3 MB/s
$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 137.217 s, 7.8 MB/s
$ dd if=/dev/zero of=./largefile2 bs=1M count=4096
4096+0 records in
4096+0 records out
4294967296 bytes (4.3 GB) copied, 316.554 s, 13.6 MB/s
$ sudo sh -c "sync && echo 3 > /proc/sys/vm/drop_caches"
$ dd if=./largefile of=/dev/null bs=4k
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 9.94723 s, 108 MB/s
$ dd if=./largefile2 of=/dev/null bs=4k
1048576+0 records in
1048576+0 records out
4294967296 bytes (4.3 GB) copied, 38.7523 s, 111 MB/s
----------------------------------------------------------- 
Avg write: 8.23 MB/s         |         Avg read: 109.5 MB/s
----------------------------------------------------------- 
$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 162.251 s, 6.6 MB/s
$ sudo sh -c "sync && echo 3 > /proc/sys/vm/drop_caches"
$ dd if=./largefile of=/dev/null bs=4k
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 17.7766 s, 60.4 MB/s
----------------------------------------------------------- 
Avg write: 6.6 MB/s         |         Avg read: 60.4 MB/s
----------------------------------------------------------- 
$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 55.6096 s, 19.3 MB/s
$ sudo sh -c "sync && echo 3 > /proc/sys/vm/drop_caches"
$ dd if=./largefile of=/dev/null bs=4k
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 17.5442 s, 61.2 MB/s
----------------------------------------------------------- 
Avg write: 19.3 MB/s         |         Avg read: 61.2 MB/s
----------------------------------------------------------- 
$ dd if=/dev/zero of=./largefile bs=1M count=1024
1024+0 records in
1024+0 records out
1073741824 bytes (1.1 GB) copied, 24.4734 s, 43.9 MB/s
$ sudo sh -c "sync && echo 3 > /proc/sys/vm/drop_caches"
----------------------------------------------------------- 
Avg write: 43.9 MB/s         |         Avg read: ---- MB/s
----------------------------------------------------------- 
$ dd if=./largefile of=/dev/null bs=4k
262144+0 records in
262144+0 records out
1073741824 bytes (1.1 GB) copied, 20.6621 s, 52.0 MB/s
----------------------------------------------------------- 
Avg write: ---- MB/s         |         Avg read: 52.0 MB/s
-----------------------------------------------------------

 

Huh, interesting results. Perhaps cooling may yield some benefit?

Hi there!

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
16 hours ago, GuitarBro said:

More ridiculous stuff has been attempted in the past, so why not try to water-cool a USB flash drive to see if it makes it go any faster under sustained load?

 

Maybe someone with too much free time can test this for science?

Until USB / SD Card manufacturers actually start improving write speeds on their devices, water cooling is completely pointless, save for doing it just because you can. (Which would still be cool same as how RGB all the things is apparently all the rage now.)

Desktop: KiRaShi-Intel-2022 (i5-12600K, RTX2060) Mobile: OnePlus 5T | Koodo - 75GB Data + Data Rollover for $45/month
Laptop: Dell XPS 15 9560 (the real 15" MacBook Pro that Apple didn't make) Tablet: iPad Mini 5 | Lenovo IdeaPad Duet 10.1
Camera: Canon M6 Mark II | Canon Rebel T1i (500D) | Canon SX280 | Panasonic TS20D Music: Spotify Premium (CIRCA '08)

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author
46 minutes ago, kirashi said:

Until USB / SD Card manufacturers actually start improving write speeds on their devices, water cooling is completely pointless, save for doing it just because you can. (Which would still be cool same as how RGB all the things is apparently all the rage now.)

Anyone wanna volunteer to do it just because they can? ;)

Hi there!

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
17 hours ago, GuitarBro said:

Well, suppose you just really don't like warm flash drives when they're ejected and you have to touch it then.

Water cooling won't change that, and in fact, might transfer heat to your fingers faster, making it feel warmer.

 

If somehow you can push a low profile drive to saturate USB3, you may have some purpose to experiment with better cooling systems.

My eyes see the past…

My camera lens sees the present…

Link to comment
Share on other sites
Link to post
Share on other sites
Posted

Going back to the root of the problem, you don't need watercooling to cool something, in principle it would be enough to glue some heatsinks on the controller chip and maybe the nand memory and then put a tiny fan over the heatsinks which could be powered from 5v.

Video cards used to have those fans but you can get some that are even smaller, for example this one at 12x12mm and 3 mm height : https://www.digikey.com/short/q75whb

UF3C3-700.jpg

or you could use a blower style fan at the end of the stick to push air through the fins towards the front of the fan... you make the stick longer but still slim and only as wide as the fan (the one in picture below is 15mm wide ) : https://www.digikey.com/short/q75w93

 

 

MFG_BSB01503HA3-00CGE.jpg
Link to comment
Share on other sites
Link to post
Share on other sites
Posted

Passive heatsinks would be more than adequate.  By the time you add tubing and pump and shit it's way too bulky and probably wouldn't have any performance benefit.

Workstation:  13700k @ 5.5Ghz || Gigabyte Z790 Ultra || MSI Gaming Trio 4090 Shunt || TeamGroup DDR5-7800 @ 7000 || Corsair AX1500i@240V || whole-house loop.

LANRig/GuestGamingBox: 9900nonK || Gigabyte Z390 Master || ASUS TUF 3090 650W shunt || Corsair SF600 || CPU+GPU watercooled 280 rad pull only || whole-house loop.

Server Router (Untangle): 13600k @ Stock || ASRock Z690 ITX || All 10Gbe || 2x8GB 3200 || PicoPSU 150W 24pin + AX1200i on CPU|| whole-house loop

Server Compute/Storage: 10850K @ 5.1Ghz || Gigabyte Z490 Ultra || EVGA FTW3 3090 1000W || LSI 9280i-24 port || 4TB Samsung 860 Evo, 5x10TB Seagate Enterprise Raid 6, 4x8TB Seagate Archive Backup ||  whole-house loop.

Laptop: HP Elitebook 840 G8 (Intel 1185G7) + 3080Ti Thunderbolt Dock, Razer Blade Stealth 13" 2017 (Intel 8550U)

Link to comment
Share on other sites
Link to post
Share on other sites
  • 1 year later...
Posted

Yeah i am creating a big file on my usb  and its gets slower probably cause of heat. I may remove the case and put a little raspiberry pi heatsink on it but i am afraid to broke it while trying to remove case. Linus make it pls

Link to comment
Share on other sites
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share
Go to topic listing Custom Loop and Exotic Cooling

×