apparently something is still pointing people towards it
anyone with a dual 6276?
thats 2 memory channels per socket? http://www.cpu-world.com/CPUs/Bulldozer/AMD-Opteron%206276.html if that's the case i'd guess 800kH/s .. 900 kH/s, but am more than welcome to be corrected
4 channels i think
but i may be wrong
The number of controllers: 2 Memory channels per controller: 2
@Rotonen if memory is "reserved" can we get more hasharte?
or about the same
that'd only be relevant with very asymmetric NUMA nodes and very large snowfields, but yes, but not yet
think of something like a 2TB field on a multisocket AMD EPYC, that and up
i rented a 6276 x2
i have some sketches, i welcome all updates to them https://snowplough.kekku.li/wizard/ A quickstart wizard to help you get started with mining Snowblossom.
just some hours ago
click on links in the table to navigate around topics
Im aware of your pool / website
but as you have 4 memory channels, i'd estimate 800kH/s
so 1.6 mh?
i'm not sure how i'm reading that processor spec, 0,8 or 1,6, but yes
oh that's a multipackage thingy like the core2quad was, kk, so 1,6MH/s is plausible
maybe if i use pre cache field
for using cpus cache
i get more?
no, memfield
as these cpus have a lot of cache
no, that'll happen below the operating system
completely automatic, not something you can fiddle with unless you write your own operating system
ok tell me another think
128gb is not enough
correct?
yep, but you can still probably get 1MH/s off it
just don't precache, the filesystem cache will eventually have enough of it in memory
i tried a vps with 300 gb
alternatively you can precache and tweak the abort settings for how manieth can be a memory cache miss for it to give up or not
was getting 1.6mh
but its shared memory
with virtual machines it is very hard to judge how many memory channels your seeks get mapped over
exactly
but it's reasonable to assume they pack 1 .. 2 socket blades into 1U units as their backing hardware, and then overprovision those for their margin
the vendors probably have conference talks on that kinda stuff on youtube if someone is actually interested in trying to reverse engineer how to max that out
Soon they will make snowblossom optimized hardware
aka file server
i'm pretty sure pixar and other similar houses would have just the right stuff for snowblossom
Probably
when is the private miner released?
there is a private miner?
ofc there is one
such big hashrates
do you think its gpu?
that's not a software problem, though
i do think its gpu
along system ram
how'd that even work?
as you have 16GB per card on high end datacentre gear, there'd still be ram in the mix, limiting the seek speeds?
gpu memory access is fast, but you need 128GB
ref. how that works for not enough ram vs. disk
The V100 is also in 32GB, but at a very high price
well, those cost along the lines of 15k usd a pop
So .. 4xV100 32GB + NVLink, maybe possible
nvlink is not that fast
couldnt the gpu act
as cpu
you still need 128GB gpu memory
i'll rather believe someone is throwing a mainframe at this before believing someone solved the interconnect problem, which is a pretty hard HPC problem
i said act as cpu
meaning
you have x cuda cores
the number crunching is not limiting the hash rate
plus system memory
even CPUs could do more hashes, you just cannot get the data to them fast enough
I experimented with this quite a lot. The point is - the gpu needs a lot of data from memory when mining. So much its saturating the PCIe-Bus
While the memory is fully loaded
mjay may you share your flags in private?
as we are still early miners
:wink:
and even if you have an omnipath fabric or equivalent, that'll still be about as fast as ram is, or slower
what flags?
for mining
its more about cheap hardware
ram its not cheap
DDR3 Reg is
hmm how much do you think 1 snow is worth
if we started now
im looking at market and hashrates
and i dont get how ones get 5+ mh
@dk3 but if you're interested in figuring this stuff out, start reading into HPC interconnects, they're not magic (also limiting AI research currently, so there is money being thrown at solving those issues) https://www.intel.com/content/www/us/en/high-performance-computing-fabrics/omni-path-driving-exascale-computing.html Find the latest benefits on Intel® Omni-Path Architecture and how Intel is clearing the path to Exascale computing, addressing tomorrow’s HPC issues.
@Rotonen maybe we just buy an movidius acelarator
do note the pools just sum the hash rates of multiple miners if they mine to the same address
instead of a cpu
:wink:
i know how it works @Rotonen just trying to figure this out as a miner
I also tried a xeon Phi, which is actually slower than a cpu
i think its more about cores
*threads
heh, was not expecting anyone to have actually tried the knights landing stuff out
once the memory channels are fully loaded more threads are useless
(and descedants of the mad 'let us superglue 100 pentiums together' project)
@mjay say, would you have any IA-64 laying around?
@Rotonen Knights corner, I cheaped out on this one :wink:
No never worked with IA-64
afaik at least huawei is still selling them
i think fujitsu finally gave up
do they have advantages when it comes to mining snow?
they have a set of caching choices i've not had access to, probably not, but i'm curious
and they updated the itanium line with the 9700 with just last year
disabling caching would be the best choice?
well, they do clever stuff to avoid the cycle miss
or, so they say on paper
Is there no equivalent on x86?
iirc they do hypertheading in some weird registry magic way instead of just cache interleaving
not finding the actual chart i remember, but https://en.wikipedia.org/wiki/Montecito_(processor) Montecito is the code-name of a major release of Intel's Itanium 2 Processor Family (IPF), which implements the Intel Itanium architecture on a dual-core processor. It was officially launched by Intel on July 18, 2006 as the "Dual-Core Intel Itanium 2 processor". According to Intel, Montecito doubles performance versus the previous, single-core Itanium 2 processor, and reduces power consumption by about 20%.[1] It also adds multi-threading capabilities (two threads per core), a greatly expanded cache subsystem (12 MB per core), and silicon support for virtualization.
When I look at the kernel source of madvise, MADV_RANDOM would be the flag to set
when we reach 256 snowfield
will we see an hash srop?
average diff 41, probably 4G nethash (minus diff creep per block)
ok my question is
I see special code for alpha, mips, parsic and xtensa
or should be can snowfields go back?
and yes, that's been customary so far, but big miners usually have the next field anyway
snowfields never go back, the current one activated at diff 39
ok got it
which sorta drove away a lot of the early bigger miners as memory on the cloud is expensive
the best thing i see happening is people mining at multiples of 100kH/s on NVMe, nethash distribution is good and IMO 100kH/s is longterm viable
the nice thing is, that memory is only 10x quicker than NVMe
what is that
modern m.2 ssd
more gbs more allocation
more speed*?
and a quick one 512GB .. 1TB costs 100 .. 300 usd, vs. thousands in ram
or has to do with with
and they take very little electricity
nvms speed
well, like with ram, nand channels
but as a good guideline, read IOPS / 6 -> hashrate
so 500k read IOPS -> ~85kH/s + whatever fscache / preallocate gives you on top -> ~150kH/s .. 200kH/s
how many blocks are per day btw?
it varies?
i recommend the 1TB sandisk for a reason
well, it's random, but the network is trying to adjust the diff so it hits 144 per day
aka one every 10min
so i assume there are big drops
every day
was funny when block 5000 was way behind schedule and people had a panic about a potential bug :smile:
yeah, some people, especially early on, were gaming the adjustment algorithm by pumping the diff
that gets harder by nethash distribution
having smaller nvme ssd´s add using raid 0 would be a cheaper option? You could add more if the field grows
raid 1, until the field grows
the issue is the smaller SSDs have less NAND channels
currently 1TB drives seem to be the sweet spot
and one should get a low power one, as they have thermal throttling issues under sustained loads
I was looking at the 960 EVO 250 GB
330k IOPS read
that's a 10W device, i'm recommending 0,5W devices
oh understood
@Rotonen can you have more than 1 nvm
look at the 1TB sandisk 3d extreme
does it makes sense?
in the same machine
or 1 nvm per machine?
depends on your motherboard block diagram, but there are also pci-e risers, but you'll still eventually hit some limit as data passes through ram, and the bandwidth saturates eventually
the pci-e risers are very various in quality so cannot recommend to get into those yet before someone bites the bullet and figures out a good one
but if one is so inclined to do, there's one from asus which could actually work
https://www.asus.com/Motherboard-Accessory/HYPER-M-2-X16-CARD/ ASUS Global: HYPER M.2 X16 CARD | Motherboard Accessory | ASUS Global
Random Read/Write Speed Up to 150,000/110,000 IOPS
mining motherboards do not work well for that, but any multi-gpu SLI motherboard would (if one wants more than one of those)
is tihs any good?
150 000 / 6 -> 25kH/s
really that simple to ballpark, just divide the read iops by six
SanDisk Extreme PRO 500 GB, costs half of the 1TB version, 410k IOPS
looks like a better option to me
dunno, 300usd is not *that* bad, and i'm not expecting anyone to go beyond 500k per m.2
but sure, definitely in the correct ballpark
150usd for the smaller one, 2 of those will be faster I guess
correct
and it'll take 2 new fields to force raid 1 -> raid 0, and 3 new fields to obsolete
3 new fields to force me to buy another 2 :wink:
striping reads has a management overhead, dunno if that's relevant, though, never done that on such a request density
350,000 IOPS
this one is the best
i think
if the total size aligns well with the total devices size, and you have various device sizes, JBOD works well too
ok its not
90 eur
500 000 to 550 000 IOPS is good, but beyond that, the power consumption should be low as well, so it does not cook itself
where are the big ones?
in iops?
i use this for scouting if new interesting drives show up on the market, lessen any of the criteria to widen the search https://geizhals.eu/?cat=hdssd&sort=r&xf=2235_500~2384_0.5~4832_3 Preisvergleich und Bewertungen für Solid State Drives (SSD) mit Schnittstelle: M.2 (PCIe), IOPS 4K lesen: ab 500k IOPS, Leistungsaufnahme Betrieb: bis 0.5W
even though it's a german speaking service, the tech lingo is mostly english, so one should be able to manage mucking around with the filters
above 500k there are only expensive pcie-drives, not worth it
if you want to go beyond 500k, you need to get u.2 or the intel ruler format, but then the price is about 5k .. 10k per drive
can you link me your wizard again?
what i'm speculating there, is that a completely fanless low power system is somewhat optimal for snowblossom
but i've not figured out how to metrify suitability of motherboards yet
i predict snowblossom more than 100
Thats a little too much I think
its not
when snowfield 2
kicks in
actually the next one *
the next one is 8
thats why i corrected
is anyone selling?
at how much?
Not at these prices
I doubt anyone is actually selling now
I wait for the exchange, makes it a lot easier for me
bch exchange is cool idea
qtrade is going to add it
can u link me
they are still working on it, its not listed yet
any new coins like snowblossom and amoveo
in early stages?
none that I know of
lets say i pay 500 $ sv for snow
and i get 10 coins per day
1 snow
has to be more than 1 $ for sure
what hashrate did you calculate with?
just wondering
how much $ month
do you think its worth right now
to rent servers
and get y hashrate @Rotonen
no renting servers is not worth it
1.6 mh
how much would you pay
for it?
month?
Currently you get 1,05 snow per day for 100kh/s
1 snow 100kh?
no way
50 coins / (2^38,6 / 100kh/s / 3600 / 24)
ok i get it
but how much would you pay
for 1.6 mh
~510 snow
you're asking us to speculate on future valuations, we're probably not the crowd for that
@Rotonen dont get me wrong im just seeing things as a "miner" now
i throw idle capacity into mining snow since i genuinely enjoy the PoW as a technical challenge
Heh. I am working on a distributed content network that leverages snowblossom
renting a big-ass server is just gambling right now
better gamble on snow
have a look at used DL580 servers
Ram is cheap for them, they have plenty of cores, and they heat up your house!
4xE7-4870 => 2.6mh/s
they are also rather loud
how much 1 of those should cost?
depends on the deal you find, that is enterprise gear, a couple of generations old
with ram below 1k usd if you are patient
You need generation 7 btw
And yes, they are loud :wink:
keep in mind that you are not as flexible if the snowfield grows
funny how the perf per watt is about the same as getting a modern small low power thing with about the same money, but of course they're ~10x faster (which is also funny how well that ram being 10x nvme holds through many examples)
yeah, the absolute performance will probably drop to the same as the low power nvme miners once the next field hits?
but you can just resell old servers, as you bought them, but meh, effort
400 $ for 1.6 mh is it any good?
thats .. per month?
as a single purchase, yes it is
what do you mean by as a single purchase?
its per monthg
you can get 800kH/s 256GB machines from the hetzner auctions for about 100 / month -> half the price per hash
https://www.ebay.de/itm/HP-ProLiant-DL580-G7-Server-19-Zoll-gebraucht-4x-Xeon-E7-4850-10-Core-256GB-RAM/401475362520?hash=item5d79cbded8:g:udQAAOSwRE5aWOqy HP ProLiant DL580 G7 Server 19 Zoll gebraucht 4x Xeon E7-4850 10-Core 256GB RAM | Computer, Tablets & Netzwerk, Firmennetzwerke & Server, Server, Clients & Terminals | eBay!
They are sold in large quantities, 256GB ram already
but still wouldnt be 1 cpu only better?
than 4x
only memory channel counts matter
these have 4?
16
four per socket
6276 have 4 per socket ?
or 2 ?
but they have 2 controllers
or, well, depends how you count sockets
i count numa nodes
the dual package thing keeps throwing off my count on those
yeah, they have 4 per sockets as well
so how much per 1 cpu?
can you estimate?
depends on what you count as a cpu :stuck_out_tongue:
the only ram mining estimate i give is 200kH/s per memory channel
but realistically across a wider range of hardware that's 150 .. 250, but 200 is a reasonable baseline
this guy is running 4x
i will have less 2-x
can you tell by the image how many threads is he running?
no idea
you can see
start with hyperthread count, see what you get, try -50% and + 50% to see if either makes a dent, if so, explore in that direction
well, htop is saying 711 threads, but dunno how many of them are active
i also tried the new javas
the parallel GC for old objects can have an impact on some architectures, so try that out (there's a pointer on the wiki somewhere)
i tried that on vps
i will share my results tomorrow
when server is ready
thank you, always interested in more information
but reserving memory
is any better
i know you answered this
but i dont remember
-XXaggressive:memory
what about this one?
@Rotonen maybe you can answer me what is "garbadge" lol
Planning for a several million node distributed hash table is fun
What is this hash table for?