Koolance also specifically states not to use silver with nickel. not just EK

see here: http://koolance.com/processor-water-block-amd-cpu-400a

must be some basis for it

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

Well, I don't understand what the problem would be, as Aquacomputer makes CPU waterblocks with the silver base and nickel plated top housings.
In fact, that is what I am running with right now.
Meow.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Koolance also specifically states not to use silver with nickel. not just EK

see here: http://koolance.com/processor-water-block-amd-cpu-400a

must be some basis for it

Again, just like EK for blaming silver for their crappy thin nickel plating. I refer you to this page on metallurgical galvanic and corrosion compatibility.
galvanic_compatibility

In it it states that for "harsh" environments, (which a cooling loop can be considered somewhat), it says to keep the difference in the "Anodic index" no greater than 0.15V and 0.15V is acceptable.
The Anodic Index for solid silver is 0.15. The Anodic Index for nickel plating is 0.30. So (0.30-0.15=0.15V) and is acceptable. The Anodic Index for copper is 0.35. So how come no copper block manufacturers have screamed bloody murder NOT to use silver in the loop for their products. The difference is greater for copper-silver than it is for nickel-silver.

Again, this is just manufacturers needing to blame the consumer for a product failure on the RMA request so they don't have to make good on any warranty. This was the consensus of all the forum posts after the first RMA denial was brought to light.

Nickel-Silver galvanic corrosion is a "red herring".

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

You got it in your first line - poor plating will always give problems.
Also use of lower quality materials, particularly in those in contact with the water where inclusions can lead to premature corrosion through to the base metal.

One thing to consider is the relative price of copper vs sliver - copper is about $6700 per ton, while silver is about $500 per kg (or $500,000 per ton). Given that there is going to be a fair bit of "overhead" in a cooler block, I'd still expect to see a solid silver block costing at least double (or maybe even ten times) an equivalent solid copper block., as with that raw material price difference there is bound to be an impact on the selling price.

Hmm, maybe my price difference figures are a bit off - I've just looked at the weight of a sample GTX1080 block - just over 1kg, so about $7 worth of copper, or about $500 worth of silver....

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

The entire block is not made of silver though. The coldplate is, but the larger portion of the weight is the upper portion with the threaded fittings. The coldplate and the microfins that transfer the heat are machined from solid silver.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

The AC nickel-nickel cpu block is $100. The AC copper-copper block is $85. The AC silver-nickel block is $220. So about twice the cost of silver over copper so follows your rule of thumb formula.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Yes, only the coldplate is 0.925 sterling siliver. Which is not pure silver. Pure silver is too soft for machining. Sterling silver is an alloy of copper and silver with the 0.075 portion content being copper.

I am getting the PVD version for the upper part of the block. That is copper plated by titanium via Physical Vapor Deposition. That makes it a scratch free surface and also a nice satin black color.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Even though PC shipments grew for the first time in 6 years, WD are shutting down one of their HDD factories, and boosting production of SSDs.

Looks like SSDs are finally mainstream products, although basic systems still come with a HDD due to the smallest capacities incredibly cheap prices. I doubt it will be much longer till even the cheapest of systems comes with a SSD & only high capacity HDDs are produced.
Western Digital to Shut Down HDD Manufacturing Facility, Increase Production of SSDs.

---

Grant
Darwin NT

Saw that coming a few years ago when SSD's first arrived at the scene. Traditional hard drive technologies had succumbed to their own version of 'Moore's Law' and the improvements in areal density had become increasingly stretched out. Only way to make profitable drives to pay for increased R&D costs that would produce minimal improvements over the previous technology was to build multi-platter enterprise level hard drives for the datacenter customers who could pay the high costs for huge amounts of products.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

HDDs will still be around for a LONG time. even if only for large data storage projects. both in personal storage solutions and enterprise solutions.

large capacity SSDs are prohibitively expensive. I run my personal Server/NAS on an 8-disk Raid10 array (4x mirrors). I have about 40TB of disks (20TB total storage) that i was about to get for less than $700. that kind of pricing wont come to SSDs for a good while.

for small storage, or OS drives though, yeah, no real need to use a HDD anymore, unless you want to repurpose an old drive for free or something.

---

Seti@Home classic workunits: 29,492 CPU time: 134,419 hours

That's been my path. I have no need for large storage but I have moved my old hard drives into Bulk storage purposes for my SSD OS host drive systems. Move files from the download directory after installation to the Bulk storage hard drive for archival purposes if I need to reinstall an app or something.

I guess that people with lots of games or music or videos will always find hard drives useful until or when SSD pricing at much higher densities matches hard drive costs. With the increasing 3D layer counts and single chip density going up, that might happen in 5 years.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Heh, it's not often that you get a blc task that isn't a .vlar task.
But here's one though, that is a shortie, with WU true angle range is : 1.810926

blc01_2bit_guppi_58227_21810_HIP67105_0060.19135.409.22.45.73_1

I got a few more of them waiting to run.

I just crunched one.
Arecibo shortie run time on my GTX 1070 is around 2min 32sec.
This particular GBT shortie task took 1min 50sec.

Edit- just crunched another non VLAR GBT WU, this one a non-shortie. 3min 41sec.

---

Grant
Darwin NT

I was seeing quite a few of them earlier today. This one crunched in 24 seconds on my 1080Ti with an AR of 1.81 and it wasn't an overflow.
blc01_2bit_guppi_58227_21810_HIP67105_0060.15103.409.21.44.49_0
That's approaching Petri speed.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Does that mean that estimated run time could be calculated in advance, for only that of Angle Range (AR),
or could it also be that of a short running time, for that of possible RFI or the like, if not any aborted or broken task either?

Yes, it is expensive because of the sterling silver coldplate. But that is why I am willing to pay the cost. I figured the material probably would give me the gain in temps I desire over the already very good Raystorm Pro.

This is what Thermalbench says about the water block after testing :
"The .925 silver base is more for an anti microbial and unique finish than any improvement in heat transfer. I don’t really recommend it. "
"The most expensive version is going to be more for looks than performance."

. . Not sure that is entirely valid. Silver is one of the best conductors of heat.

. .Boy was I late to the party with this comment ...

Stephen

. .

While silver is one of the best conductors you can ruin its performance by placing an insulating layer, or a poor conductor, between it and where you are trying to get the heat to/from.
Both copper and silver are also biocides, silver tends to work faster on bacteria and is "safer" to mammals and other higher life forms.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

I've had lots of exposure to copper's properties after working in electronics for decades and it does conduct heat pretty darn well. But if anyone ever has stirred a cube of sugar into a cup of tea or coffee with a silver spoon knows how fast the handle of the spoon gets uncomfortable to hold and stir with. The temperature rise seems almost instantaneous. Only graphene and diamond are better thermal conductors. I like my Arctic MX-4 TIM as is it very fluid and spreads very thinly. That should reduce the interface insulation layer to a minimum. Still awaiting my block from Germany though so nothing to report yet.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

An interesting comparison.
A high-end 130W $1,000 CPU from 2010 compared to a lower end 65W $100 current CPU (and few other current CPUs).
It shows just how much performance has increased, and how much efficiency has improved, over the last 8 years.

i7-980X v Ryzen 3 2200G

---

Grant
Darwin NT

Finally a report on the AquaComputer Kryos Next silver block I ordered. Got it installed just as I lost A/C for two days and had to shut everything down. Finally back crunching again and got exactly what kittyman stated for temp reductions compared to the XSPC Raystorm Pro. Looks like 2-3° C. reduction in cpu core temps better than the Raystorm Pro. Responds to changes in cpu loading faster I noticed. Solid and heavy product, no flaws in the top of the block in the PVD coating and the silver coldplate was mirror finished. A very well designed and engineered product. Can recommend the inexpensive standard nickel or copper versions based on the apparent design and product manufacturing. The nickel and copper versions are in stock in U.S based vendors. The silver versions seem to be build to order only from the factory in Germany so expect a much longer delivery.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I'm not sure where to put this, and didn't want to start a new thread, so here goes the saga of a computer upgrade.

Well my daily driver was an off the shelf ASUS 1150 that blew up 1.5 years ago, and I upgraded the MB and CPU to a 3280/4c. It is a cheap China board. Certainly the case wasn't designed for it with only a 92mm fan in the rear, so cut a hole in the bottom for a 120mm AIO, and hole in the front for a 90mm and slotting the sides of the front panel for air inlets. It worked fine but was always a tad on the hot side, but I like the case.

So i got a CoolerMaster Pro3 mATX case, and pulled the 6c/3930k out of my top computer (10c/2690v2 now). The chip was definitely well used: dried up compound (Arctic Silver I think) which left some very dark stains on it, even after a Dremel polish, and 2 pads that don't look fully intact (but look fine when wet with alcohol). Got everything installed and nadda, no BIOS. I tried for well over an hour and gave up, reinstalled the 3820/4c. Still nothing after another hour or so. So tired of pulling RAM and swapping video cards, not to mention having to hit the input button on the monitor every damn time I powered up. WTF thought maybe, just maybe there was a standoff under the board I missed somehow, so pulled it and nothing. I gave up then and pulled the battery ... that is when I mentioned it in another post.

So next day I gave it another go (still with the good 3820/4c). Had to wait for a response from and eBay'er on a replacement board anyway. Finally after 1-2h it fired up (no drive of course). That China board is so strange for a BIOS reset, I'm still not exactly sure, but seems to be combination of jumper, and power flicks to reset it. But is there a manual available? Heck no. Just 2 photocopy sheets that say nothing about anything other than to use a magnifying glass or phone if you want to read the smallest imaginable print on the MB - they couldn't even hand draw a sketch of it, LOL.

After booting it up 5/6 times to a Mint install DVD for a test (HD got move to my 4590 so I had a Win computer) it seemed fine, so I tried the 3939k/6c again. At this point I was tied of that damn AIO and thermal grease, so disassembled the old Intel 1150 cooler to use as a 'paper weight' cooler - worked fine with no load :D This time I had it running in under an hour ... WOOHOO! Booted it up into Linux, great, it's working.

Reinstalled everything, nice and tidy cabling job, got it back under my desk - oh my back's getting so sore from bending over this thing! Sat in my chair and hit the power button ... NOTHING AGAIN! WTF. I give up. AIO shifted the CPU a bit??

So today, again another go at it (still no responses from eBay). Pulled the AIO, and 3930k. Cleaned again, installed with 'paper weight' and didn't have any luck. back to the 3820, and finally again, powered up to BIOS - only about 1.5h this time screwing around. And REALLY wishing I had some beer! So lets try that 3930 again, why not for kicks. Only took about 30m this time. I still don't understand what is needed to reset that damn BIOS, it seems different every time, but i seem to be getting better at it, LMAO, whatever 'it' is.

Well screw it at this point, lets see if the pressure of the AOI causes problem, I "HOT" installed that and it didn't crash. Put the drive back in and a quick BIOS change, it booted up. A few reboot later and it seems fine ... I still haven't actually tried a power down, I'm too scared!

It has been running for 13h now, and much cooler than before. Next is to clock it up, maybe even power it down, LOL. I figure if I don't put that final cable tie on the AIO wiring the job isn't complete, and it might just keep going!

It's been fun, fun, fun I tell ya!

I'm glad I wasn't doing this doing this simple transplant for someone else - they would faint at a 10+ hour bill.

---