[Dev] NeoScrypt Hardware Comparison Site
-
I attempted to add some R9 270x settings, but once I finished submitting the form the settings do not seem to appear in the main table. I don’t know if i did something wrong or if the submissions are moderated and need to be approved.
I have to approve the submissions first (stops people spamming silly entries)
Edit: also, where did you get your power value from? Is this measured at the wall or via gpu-z? As 80w consumption from a 180w R9 270x is a little odd :P i’ll assume you meant 180 and it was a typo unless you can prove otherwise :)
Edit 2: I’m not sure on many of the specs, cores, power, datarate dont match to what an R9 270x has, can you use GPU-z (get it here: http://www.techpowerup.com/downloads/2418/techpowerup-gpu-z-v0-8-0/) to get your information and re-enter it?
Thanks
-
I have to approve the submissions first (stops people spamming silly entries)
Edit: also, where did you get your power value from? Is this measured at the wall or via gpu-z? As 80w consumption from a 180w R9 270x is a little odd :P i’ll assume you meant 180 and it was a typo unless you can prove otherwise :)
Edit 2: I’m not sure on many of the specs, cores, power, datarate dont match to what an R9 270x has, can you use GPU-z (get it here: http://www.techpowerup.com/downloads/2418/techpowerup-gpu-z-v0-8-0/) to get your information and re-enter it?
Thanks
Yeah, the bus width and stuff I had placeholder info, I was going to look it up and update later. I misunderstood how the submission worked, thought it would be real time and not big a big deal to update it right away.
The power values are correct using under-volted cards. I measured a 4x 270x rig drawing 370 watts total from the wall, subtract the 50 watts idle for the system and divide by 4 to get the per card wattage.
Edit: Resubmitted with correct values from GPUz.
Also, I tried adding (Elpida) on the memory entry, so “GDDR5 (Elpida)” but there were not enough allowed characters in the field, perhaps you could extend the field length to allow a few more extra?
-
Actually looking at the other entries on the chart all the power values seem to be on the high side. I know some of those were applicable when using scrypt settings and overclocking, but for instance the 280x cards are showing 250 watts each. I have some 4x 280x rigs as well and they are only drawing about 150 watts each, also undervolted, and mildly overclocked.
The rig I am using for reference (4X 280x) draws 660 watts total from the wall, I first check the idle usage, which draws about 60 watts, then check after it has been mining for about 5 minutes, which is the 660 watt figure. Again subtracting the 60 system watts and dividing the remaining 600 watts by 4 gives me an average of 150 watts per card. These are giving me ~315 kHash each with a mild 1100 core and 1500 memory overclock. Since they are undervolted they are not stable much higher, but I only get maybe 10-20 more khash each for quite a large increase in power draw, so not worth it in my opinion.
Edit: Actually on further investigation, if I ran a 280x giving only 315 kHash at a 250 watt power draw (let’s assume this includes the pro-rated share of the system draw as well; i.e. 230 watt GPU 20 watts system on a multi-GPU rig), this would work out to just breaking even if you pay $0.08 per kw/hour. You would not be in profit unless you were paying $0.07 or less per kw/hour, which unfortunately in my case is closer to $0.10.
-
Those power entries for the 280X etc, are for the rated TDP (where the actual power could not be measured, e.g no meter)
Seems pretty good that you undervolted and it only hit 80w ^^, and for the nvidia card that is actual power draw, as you can see they arent very power efficient at scrypt lol
As for GDDR5 (Elpida), please can you add that in the more info box, alomg with the info of what voltages and miner parameters, thats what that box is for :)
Also adding a submission is a 1 time thing, however if youve submitted one, but would like it updating, in the notes you would send the link to the individual card page (use the more info link in the main tables), and say that it needs updating.
-
############### Update ###############
Styles are now updated for all base pages, ACP still under construction.
Had a massive panic attack when i deleted new submission form page by accident, luckily i had a backup, so panic over, and its all alive :D
ACP Pages are a pita, since i have to do checks to make sure user is logged in and in a legit session (-_-) also designing them all to interface together is a bigger pita (-_-)
The new style should be a bit cleaner to use, and a bit easier on the eyes.
As always, any suggestions are welcome!
############### Thanks ###############
Sydney
-
Type: GPU
Hashrate: 21.60 Kh/s
Vendor: AMD HD 6870
Est. Power: 151w
Miner: cgminer 3.7.8
Proof: -I 13 -w 128
Type: GPU
Hashrate: 57.34 Kh/s
Vendor: AMD HD 7870
Est. Power: 185w
Miner: cgminer 3.7.8
Proof: -I 14 -w 32
-
Type: GPU
Hashrate: 20.38 Kh/s
Vendor: AMD HD 6850
Est. Power: 127w
Miner: cgminer 3.7.8
Proof: -I 12 -w 32
Type: GPU
Hashrate: 97.22 Kh/s
Vendor: AMD HD 7950
Est. Power: 200w
Miner: cgminer 3.7.8
Proof: -I 14 -w 128
-
Type: GPU
Hashrate: 10.34 Kh/s
Vendor: AMD HD 6770
Est. Power: 108w
Miner: cgminer 3.7.8
Proof: -I 12 -w 64
Type: GPU
Hashrate: 56.68 Kh/s
Vendor: AMD HD 7850
Est. Power: 130w
Miner: cgminer 3.7.8
Proof: -I 13 -w 32
-
Thanks for those Lizhi, ill add those once i’m back on the pc, and i’ve replied to your message about the html/css page link thing,
Thanks,
Sydney
-
Type: GPU
Hashrate: 20.38 Kh/s
Vendor: AMD HD 6850
Est. Power: 127w
Miner: cgminer 3.7.8
Proof: -I 12 -w 32I’m looking to help someone setup a 6850 GPU, and ran across this.
The info shows -I12, but the image shows -I14.
Also, isn’t 98c a little high temp?
-
I agree, lizhi, check out that card and re-fit the cooler! you’re gonna burn that poor thing lol
98c is too high in my opinion, even for an AMD card (i remember those first batches of 512mb 4870/4890’s idling at 80c+ in load 120c+ lol)
speaking of 4870’s, i have one lying round somewhere, was a good one aswell, overclocked to 800/1000, undervolted a lot and nice and stable, idled at 37c, load 55c
doubt it would push out many hashes, as the general trend with neoscrypt seems to be the newer/faster the hardware the better, and seems less dependant on compute performance, 660Ti vs HD 5670 are equal in SHA256 (both about 100mh/s), but in neoscrypt, the newer bigger 660Ti leaves it in the dust (34kh/s vs 7kh/s).
Either way lizhi check that card before it sets fire to something!
-
############### Update ###############
API page working!
All entries info: http://hw.neoscrypt…gethardwarelist
Individual entries: http://hw.neoscrypt…ualhardware&id= where is the entry_id found within gethardwarelist
Search entries: http://hw.neoscrypt…rdware&qstring= where is your query (e.g “HD 5870”), the search is fulltext, and will pull any entries containing any of the search terms, and will order them with highest relevancy first.
Currently the output is formatted in pretty json mode with
tag.This will be changed back to block text json once i have written a guide page for it.
############### Thanks ###############
Sydney
-
Thanks, i will add this entry now, just got home from athletics competition, did shit, as ive just started getting back from illness, but its a start :)
It seems theres a HUGE difference between wolf0 and normal kernel for amd cards with sgminer, with nvidia cards, it seems its pot luck whether sgminer actually works, and if it does work, its hard to get it working at any speed comparable to cgminer.
seems sgminer + wolf0’s kernel is best for AMD,
and cgminer + wolf0’s kernel is best for nVidia,
i did get my i7’s HD 4000 running with cgminer pretty well, and it actually did produce an alright hashrate (11.68kh/s @ 1700MHz) and gave about same h/w as a 750Ti, i did not test with sgminer unfortunately, which perhaps i should have tried, ohwell maybe another time
i hope a cudaminer for neoscrypt is released, as it would probably boost performance and efficiency on nvidia cards by alot, we all know nvidias opencl performance is pretty damn poor, but cuda actually is not that bad, cuda with scrypt gave me 350kh/s on a 660Ti, whereas using cgminer i got around 65kh/s, big difference!
anyway thanks for the submission
Sydney
-
Hey Spartan. No idea about power + screen shots are a bit impractical but here’s all the relevant info.
Type: GPU
Hashrate: 254.1 Kh/s
Vendor: AMD HD 7950
Model Specific: XFX Double D FX-795A-TDJC
Est. Power: n/a
Miner: sgminer-5-dev-neoscrypt
-rI 4584 -w 64 77cExtra info: Shaders: 1972, Core Clock: 800, Mem Size: 3072, Mem Clock: 1250, Mem Type: GDDR5, Bus Width: 384 bit, Mem Datarates: 240 GB/s
Notes: Raw intensity gives me the best results by far. -rI 4584 is my sweet spot. Any deviation above or below that gives reduced hashrates.
Important! Using **Wolf0’**s kernel is the biggest single factor in getting optimum hashrates.
Thanks for the heads up on RL. Changing from I 15 to your RL 4584 added 30Kh/s to my XFX 7950!
Hashrate: 216.91 Kh/s (12-hr Avg)
Model: XFX Core FX-795A-TNFC
Est. Power: 200-215w
Miner: sgminer-5-dev-neoscrypt
-rI 4584 -w 64 79cShaders: 1792,
Core Clock: 980,
Mem Size: 3072,
Mem Clock: 1375,
Mem Type: GDDR5,
Bus Width: 384 bit+1 for Wolfs kernel. Took this card from 78kh/s to 200+ Kh/s.
-
+1 for Wolfs kernel. Took this card from 78kh/s to 200+ Kh/s.
Wolf’s kernel took my 290’s from 70 to 330, and my 290x’s from 70 to 302… (%^@$^ Elpida memory)
-
Thanks for the heads up on RL. Changing from I 15 to your RL 4584 added 30Kh/s to my XFX 7950!
Hashrate: 216.91 Kh/s (12-hr Avg)
Model: XFX Core FX-795A-TNFC
Est. Power: 200-215w
Miner: sgminer-5-dev-neoscrypt
-rI 4584 -w 64 79cShaders: 1792,
Core Clock: 980,
Mem Size: 3072,
Mem Clock: 1375,
Mem Type: GDDR5,
Bus Width: 384 bit
+1 for Wolfs kernel. Took this card from 78kh/s to 200+ Kh/s.Will get this added asap, having a few delays with stuff, usual festive stuff with family and being unable to access a fully working pc to actually do things on
Recently tested an R7 240 (its pointless dont try)
Got the core clock to 1300 MHz (default 780), and the memory from 800 to 1300 also (tried upping the voltage from 1.15v all the way up to 1.4v but still couldnt push past that 1300mhz), huge boost in hashrate, from around 12kh/s to about 20 kh/s, this didnt last long sadly as the cards VRMs blew after just 3 days, poor thing, ohwell was a free gpu :D
-
Will get this added asap, having a few delays with stuff, usual festive stuff with family and being unable to access a fully working pc to actually do things on
Recently tested an R7 240 (its pointless dont try)
Got the core clock to 1300 MHz (default 780), and the memory from 800 to 1300 also (tried upping the voltage from 1.15v all the way up to 1.4v but still couldnt push past that 1300mhz), huge boost in hashrate, from around 12kh/s to about 20 kh/s, this didnt last long sadly as the cards VRMs blew after just 3 days, poor thing, ohwell was a free gpu :D
If someone hands you a 7990, please don’t do the same overclock till burnout test… lol
-
You can bump the sapphire R9 290 Tri-x OC up from 333 to 340… The GPU’s came back from warranty replacement, and can get the juice cranked up a little with fans that aren’t failing…
-
If someone hands you a 7990, please don’t do the same overclock till burnout test… lol
Lol i wouldnt do it with a 7990, and the overclock was stable, just the vrms weren’t up to the job lol
If i had a 7990 id probably look into watercooling :P
The R7 240 temps stayed pretty low even at 1.4v, only just above 70c, its designed for 35w TDP, i unlocked it to 75 in the bios, and overclocked the clappers off it, it *seemed* alright xD
-
NeoScrypt CPUminer updated to v2.4.1, up to 50% performance increase.
