1)
Message boards :
Number crunching :
Rtx 30 series
(Message 143785)
Posted 907 days ago by Arjant2
I support the brand that is the fastest for my money.
So if AMD is faster i buy that but for the last few years AMD has nothing compared to NVIDIA.
So lets see what they come up with in a month or 2.
|
2)
Message boards :
Number crunching :
Rtx 30 series
(Message 143365)
Posted 920 days ago by Arjant2
According to reviews, the RTX3080 is around 20% stronger than the RTX2080Ti on INT32.
okay given the much higher power usage it's not really interesting to switch from 2080ti to a 3080.
But at least you now get 2080ti performance plus a bit against lower prices (not taking into consideration the power bill). For those who dont need to pay the powerbill it may be a better choise to buy a 3080. But running 24/7 and paying your own power bills, and only 20% increase... a 2080ti will be the better choice money wise (if only 20% increase).
|
3)
Message boards :
Number crunching :
Rtx 30 series
(Message 143357)
Posted 920 days ago by Arjant2
They claim 2x FP32 performance among other things, which sounds really good for now...
Yes, 2x FP32 since 2x FP32 compute units.
PG softwares use FP32 right?
No, all of them use INT32.
Each Pascal core is a single INT32+FP32 unit. Turing has two separate units: INT32 and FP32. Each RTX-30 Ampere core has two units: INT32+FP32 and FP32. Then RTX-30 can perform 2x FP32 but 1x INT32.
So that means no big improvement can be expected compared to the 2xxx-series, since the INT32 is used by PG?
|
4)
Message boards :
Number crunching :
New computer running slow?
(Message 134363)
Posted 1243 days ago by Arjant2
Ok didnt now that.
Will try with mt on 6 cores then.
Thanks.
|
5)
Message boards :
Number crunching :
New computer running slow?
(Message 134361)
Posted 1243 days ago by Arjant2
I just installed some new hardware at the office.
Intel i7-9700
16GB RAM (DDR 2400)
Currently testing 321 LLR, 6 tasks (so 2 cores are kept idle for other purposes), no HT. The estimated time is more than 2 days for each unit. Which concurs with the calculation done after 30 minutes (not even 1% per unit finished).
Temperatures are around 70 degrees Celcius, cores are running at 3800Mhz.
Any ideas what might be wrong? I think each unit should be done in less than 24 hours on this specs?
I know it is not the fastest memory, but that can't be responsible for twice the processing time.
|
6)
Message boards :
Number crunching :
World Maths Day Challenge
(Message 133924)
Posted 1258 days ago by Arjant2
thanks for the fun and good results!
|
7)
Message boards :
Number crunching :
World Maths Day Challenge
(Message 133684)
Posted 1263 days ago by Arjant2
I run 12 units, with smt switched off, 1 core per task with an average time of 3283 seconds per task. That seems a big improvement over a Ryzen 7 Pro 1700X which does 8 tasks in 6.593. No clue what a recent Intel does.
I have to find out what kind of effect has when doing more cores per task.
This is just a guess but I'm pretty sure that the herd was always breathing the dust from the horde regardless of the cpu :)
Cheers
AtP :)
Hehe are you trying to get a reaction from the Cows? ;)
I will try to get some cows interested. But these days they are all scattered around the boinc projects but I will try
Time to round them up then, maybe ring the dinner bell and they will all come running? My own Team is scattered far and wide as well but they are beyond communication range.
Working on it! The real herd is only reachable by a front post on our website but that is impossible to arrange for just a challenge.
That is only done for our annual stampede, that would really rally all the cows.
But posting it on our community forum will at least attract the regular cows I hope.
|
8)
Message boards :
Number crunching :
World Maths Day Challenge
(Message 133660)
Posted 1264 days ago by Arjant2
I run 12 units, with smt switched off, 1 core per task with an average time of 3283 seconds per task. That seems a big improvement over a Ryzen 7 Pro 1700X which does 8 tasks in 6.593. No clue what a recent Intel does.
I have to find out what kind of effect has when doing more cores per task.
This is just a guess but I'm pretty sure that the herd was always breathing the dust from the horde regardless of the cpu :)
Cheers
AtP :)
Hehe are you trying to get a reaction from the Cows? ;)
I will try to get some cows interested. But these days they are all scattered around the boinc projects but I will try
|
9)
Message boards :
Number crunching :
World Maths Day Challenge
(Message 133590)
Posted 1266 days ago by Arjant2
That is why I said Avx512 etc because I didnt know which app exactly is working for the AMD 3900X ;)
So the client automaticly uses the right app, that is the answer I was looking for ;)
There's just one app, regardless of the type of CPU you have. (Technically, there's 6 -- a 32 and 64 bit app for each of Windows, Linux, and MacOS -- but you're asking about CPU abilities, not OS platforms. The same app handles all CPU types.)
There used to be a time when you had to manually download the different AVX types, that is why i was asking. (maybe I am wrong and it was at another project which uses those different types)
I run 12 units, with smt switched off, 1 core per task with an average time of 3283 seconds per task. That seems a big improvement over a Ryzen 7 Pro 1700X which does 8 tasks in 6.593. No clue what a recent Intel does.
I have to find out what kind of effect has when doing more cores per task.
|
10)
Message boards :
Number crunching :
World Maths Day Challenge
(Message 133588)
Posted 1266 days ago by Arjant2
That is why I said Avx512 etc because I didnt know which app exactly is working for the AMD 3900X ;)
So the client automaticly uses the right app, that is the answer I was looking for ;)
Ofcourse I know sieving on CPU and GPU is possible, i have been doing that a lot on the GPU because on the CPU is just not effective.
I will do some test runs to see if it competative. I will let you know the results.
Thanks Michael.
Ps with compatative I mean effective. The 3900X is the top of the line for AMD and if the LLR tasks are calculated way faster on a 9900K for example it means for me that it is not worth the power consumption to put my 3900X to work on LLR. I want to donate my cpu power but I do want it to be effective. By thr way Emptying my wallet is already what I have done when building this system. But not with PG primarely in my mind, that is just an extra.
|
