Pinched a thread title from the Café as it seems appropriate for this particular post, and probably others people might come across in their wanderings.
This one isn't really appropriate for the Café, but is certainly appropriate for Number Crunching.

A very interesting video.

CPU utilization is wrong
Everyone uses %CPU to measure performance, but everyone is wrong, says Netflix's Brendan Gregg in his UpSCALE Lightning Talk.

It is well worth watching this video through to the end- 5 minutes is all it takes.

---

Grant
Darwin NT

Yes. I was talking to a number of BOINC project (and hence server) administrators around the time the punchline to that video was written. I won't spoil the ending, but their figures matched the ones in the examples, and the causal analysis reached the same conclusion. I think the effect is probably a lot more pronounced for servers than for our desktop machines, not least because they're running a different mix of software.

It would be interesting to see a similar video dissecting GPUs. I suspect that the headline might be the same, but that the analysis would find a different cause.

It would be interesting to see a similar video dissecting GPUs. I suspect that the headline might be the same, but that the analysis would find a different cause.

I was thinking along similar lines.
According to GPUz, GPU load for longer running Arecibo WUs & Arecibo VLAR WUs is quite similar, yet it takes considerably longer to process the VLARs and they have a much greater impact on system responsiveness when processing them, compared to the effects of other WUs.

---

Grant
Darwin NT

New Noctua fans capable of both high static pressure & high air flow.
With a bit of luck their price should improve after a while.

---

Grant
Darwin NT

Interesting news about the new Noctua fans. Also that there doesn't seem to be a 140mm product. I assume that the larger fan blades of a 140mm fan wouldn't be able to control the expansion and maintain the small 0.5mm clearances. So you are stuck with using adapters.

---

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

A proud member of the OFA (Old Farts Association)

The problem with getting as many storage devices in a Rack Unit is the physical size of those devices.
5.25", then 3.5", then 2.5" and even 1.8" but the problem with the really small form factors is it reduces just how much capacity can actually fit in that form.

So Intel came up with yet another form factor specifically for Rack Unit systems- The Ruler. And Supermicro have released the first system that makes use of the new form factor.
32 hot swappable SSDs in a single 1U server.

Supermicro Launches 1U Ruler NVMe Server

---

Grant
Darwin NT

R+Python=...

Last month, McKinney announced the founding of Ursa Labs, an innovation group intended to improve data-science tools. McKinney will partner with RStudio—Wickham’s employer, which maintains the most popular user interface for R—on the project. The main goals of Ursa Labs are to make it easier for data scientists working in different programming languages to collaborate, and avoid redundant work by developers across languages. In addition to improving R and Python, the group hopes its work will also improve the user experience in other open-source programming languages like Java and Julia.

---

Grant
Darwin NT

GTX 1180 details sighted.

Possibly the most interesting spec is the 1180’s rated peak floating point performance of around 13 teraflops, putting it ahead of the 1080Ti (11.4 TFLOPS) and even the Titan Xp (12.1 TFLOPS).

---

Grant
Darwin NT

"I just prefer people to be, uh... you know... intellectually honest." :)

Why CUDA "Cores" Aren't Actually Cores, ft. David Kanter
https://www.youtube.com/watch?v=x-N6pjBbyY0

Comment section is hilarious too (especially considering the subject matter)

The perfect crunching machine.

The new servers feature dual Intel Xeon Scalable processors with up to 28 core CPUs and the highest CPU TDP requirement of 205W for each socket.
The G481-S80 is able to accommodate up to eight SXM2 form factor GPUs from the NVIDIA Volta V100 or Pascal P100 lines,... meanwhile the G481-HA0 is able to support a total of 10 PCIe GPU cards

GIGABYTE Launches Two 4U NVIDIA Tesla GPU Servers: High Density for Deep Learning

---

Grant
Darwin NT

The thing with Gigabyte (and MSi for that matter) is that they'd have to pay me to use their gear as I certainly wouldn't be paying them a cent.

Cheers.

The thing with Gigabyte (and MSi for that matter) is that they'd have to pay me to use their gear as I certainly wouldn't be paying them a cent.

Cheers.

I'm all EVGA and nVidia here. With one exception.
I have a Galax Katana coming, because it's the only single slot GTX1070 ever made.
For an upgrade from a nVidia K4000 I am running now in the single slot space between the 1080ti cards in this rig.

---

"Time is simply the mechanism that keeps everything from happening all at once."

So where did you find it Mark? It isn't listed in their U.S. Store.

---

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

A proud member of the OFA (Old Farts Association)

So where did you find it Mark? It isn't listed in their U.S. Store.

Ebay....and there are not any more there presently. Supposedly used for 2 months.
Rather hard to come by.
I was also very surprised, but I also found a water block for that obscure card.
There is one of those left on ebay at the moment.

---

"Time is simply the mechanism that keeps everything from happening all at once."

Yes I had stumbled across this card a year ago when it got reviewed. I never was able to find one to purchase in the U.S. Not even in the Galax store. Still the same. I think they must have retailed it only outside the U.S. and then very briefly. It disappeared from their product listings 3 months after product release. So it never was popular or it was pulled for other reasons.

---

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

A proud member of the OFA (Old Farts Association)

A repair shop guy with 388k subscribers on YouTube

Louis Rossmann

(right to repair advocate with a foul mouth)

Recommended :)

Hi, (thank You for this thread)

To get a grasp of what you can get with the layest HW you ccould check out this example.
http://setiathome.berkeley.edu/workunit.php?wuid=2976251655

This is a random pick. A TITAN V and a 1080 Ti. Different software. No way of knowing if any over clocking or under voltaging is applied.
I do not know if the other one is running one at a time or not.

My task is a plain vanilla one at a time with the latest software and with the overclock of the GPU suitable for Finnish summer.

p.s.
None of my GPUs has never been to a sauna but I just came from there so I'm probably in the need of a cooler more than my GPUs. Luckily I'm liquid cooled.

*i* sipping some liquid -- cooling

p.p.s. I'm kind of a cool guy for this age of a man anyway! Words of a nerd you might say unless you were not as nerd as I am.

p.p.p.s. Just admit it. Come out of your cooled computer closets/man caves/nerd holes/utility rooms/garages/family spaces/from under the stairs/ or move out of your mama.

Now *sipping coolant*

---

To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones

To get a grasp of what you can get with the layest HW you ccould check out this example.
http://setiathome.berkeley.edu/workunit.php?wuid=2976251655

How well does the Special app & the Titan V do with Arecibo VLARs?

---

Grant
Darwin NT

To get a grasp of what you can get with the layest HW you ccould check out this example.
http://setiathome.berkeley.edu/workunit.php?wuid=2976251655

How well does the Special app & the Titan V do with Arecibo VLARs?

I don't know, but unfortunately it's still dropping ~6 inconclusive validations an hour. The project can probably cope with that number of wasteful resends, but it's not ready for widespread use yet.

To get a grasp of what you can get with the layest HW you ccould check out this example.
http://setiathome.berkeley.edu/workunit.php?wuid=2976251655

How well does the Special app & the Titan V do with Arecibo VLARs?

Arecibo VLAR in 75 seconds!!
http://setiathome.berkeley.edu/workunit.php?wuid=2912450437

This is one that I found.

p.s. Yes there is a bug somewhere or it is the heat wave we're having here in Finland. My Titan V runs at 57C, Ti at 83C and the 1080s at 75C.

---

To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones