Are there any publicly available custom memory strap timings for the rx 400 series? (samsung, elpida, or hynix)
I posted a basic mod for Elpida And Hynix a few pages back.. |
|
|
|
Kinda OT since this is a RAM Timings thread, but ill ask this anyways :
For Sapphire RX 470 4GB (Ref) cards GPU Core Volt offset is at A992 correct?
Now for the cards with Hynix memory, I find the default 04 which is 4 X 6.25 or +25mv, which seems legit.
But for the cards with Samsung memory, the value at A992 is FF which is -1 X 6.25 or -6.25mv, so something looks off.
Do different memory versions of the card have different default offset values? Or is the location different?
Any help/guidance would be appriciated.
Seems like the samsung one doesn't have global offset. Stock roms with global offset have either '03' +18.75mV value or '04' +25mV as VDDC offset. I've never seen something else, or rather negative offset.. and i've opened up a lot of them ;-) 'But correct me if i'm wrong..' Greetings. |
|
|
|
Original Samsung 4G ( your particular GPU ) 1625 555000000000000022CC1C00CE596B44D0570F1531CB2409004AE700 [ 0B03 | 1420 ] 7A8900A003000000170F2E36922A3217 --> MC_SEQ_MISC1
-- MR0
WL = 3, CL = 22, TM = 0, WR = 23, BA0 = 0, BA1 = 0, BA2 = 0, BA3 = 0
-- MR1
DS = 0, DT = 1, ADR = 1, CAL = 0, PLL = 0, RDBI = 0, WDBI = 0, ABI = 0,
RES = 0, BA0 = 0, BA1 = 1, BA2 = 0, BA3 = 0
Original Samsung 4G 1750 777000000000000022CC1C0010626C49D0571016B50BD509004AE700 [ 1405 | 1420 ] 7A8900A003000000191131399D2C3617 --> MC_SEQ_MISC1
-- MR0
WL = 4, CL = 23, TM = 0, WR = 25, BA0 = 0, BA1 = 0, BA2 = 0, BA3 = 0
-- MR1
DS = 0, DT = 1, ADR = 1, CAL = 0, PLL = 0, RDBI = 0, WDBI = 0, ABI = 0,
RES = 0, BA0 = 0, BA1 = 1, BA2 = 0, BA3 = 0
Bingo!  |
|
|
|
Just wanted to ask did anybody managed to get over 31+ mhs, but i got my answer before asking  That's rather easy, but keeping it there and not setting insanely high clocks, that's the interesting part. + i would take a perfectly fine tuned setup with a very low power consumption over high speed any day. I almost hit 1000h on a 480 8G (XMR) but i prefer 850 stable and very low power consumption. Greetings! |
|
|
|
Since everyone is sharing now I suppose i'll put what I've come up with out here. Running -125mv 470 Nitro Sapphire 8GB with Samsung memory with ETH hitting between 28.5MH/s to 29.2MH/s @1140 cor and @2100 mem pulling around 920watts at the wall with 6 GPU per rig. On XMR hitting 785h/s to 795h/s @1170 cor and @2100 mem pulling around 660 watts at the wall with 6 GPU per rig. Also running ethOS 1.2.0. Im here to learn more about the mistakes I made on the mod and see what others in the community have come up with. Here is the strap I've put together: 777000000000000022CC1C00106A5B47C0570E16B08C05090068C70014051420FA8900A00300000 0190D2F399D2D2E17 .... Looking forward to others input! Cleaned it up for you. --> HEX strap: 777000000000000022CC1C00AD695D47C0570E16B08C05090048C70014051420FA8900A003000000190D2F399D2D2E17
--> MC_SEQ_WR_CTL_D0
DAT_DLY = 7, DQS_DLY = 7, DQS_XTR = 0, DAT_2Y_DLY = 0, ADR_2Y_DLY = 0, CMD_2Y_DLY = 0, OEN_DLY = 7, OEN_EXT = 0
--> MC_SEQ_WR_CTL_D1
DAT_DLY = 0, DQS_DLY = 0, DQS_XTR = 0, DAT_2Y_DLY = 0, ADR_2Y_DLY = 0, CMD_2Y_DLY = 0, OEN_DLY = 0, OEN_EXT = 0
--> MC_SEQ_PMG_TIMING
TCKSRE = 2, Pad0 = 0, TCKSRX = 2, Pad1 = 0, TCKE_PULSE = 12, TCKE = 12, SEQ_IDLE = 7, Pad2 = 0, TCKE_PULSE_MSB = 0, SEQ_IDLE_SS = 0
--> MC_SEQ_RAS_TIMING
TRCDW = 13, TRCDWA = 13, TRCDR = 26, TRCDRA = 26, TRRD = 5, TRC = 71, Pad0 = 0
--> MC_SEQ_CAS_TIMING
TNOPW = 0, TNOPR = 0, TR2W = 28, TCCLD = 3, TR2R = 5, Pad0 = 0, TW2R = 14, TCL = 22, Pad1 = 0
--> MC_SEQ_MISC_TIMING
TRP_WRA = 48, Pad0 = 2, TRP_RDA = 12, TRP = 22, TRFC = 144
--> MC_SEQ_MISC_TIMING2
PA2RDATA = 0, Pad0 = 0, PA2WDATA = 0, Pad1 = 0, FAW = 8, TREDC = 2, TWEDC = 7, T32AW = 6, Pad2 = 0, TWDATATR = 0
--> MC_SEQ_MISC1
-- MR0
WL = 4, CL = 23, TM = 0, WR = 25, BA0 = 0, BA1 = 0, BA2 = 0, BA3 = 0
-- MR1
DS = 0, DT = 1, ADR = 1, CAL = 0, PLL = 0, RDBI = 0, WDBI = 0, ABI = 0,
RES = 0, BA0 = 0, BA1 = 1, BA2 = 0, BA3 = 0
--> MC_SEQ_MISC3
-- MR4
EDCHP = 10, CRC WL = 7, CRC RL = 3, RD CRC = 0, WR CRC = 0, EDCHPi = 1, BA0 = 0, BA1 = 0, BA2 = 0, BA3 = 1
-- MR5
LP1 = 0, LP2 = 0, LP3 = 0, PLL/DLL BW = 0, RAS = 0, BA0 = 0, BA1 = 1, BA2 = 0, BA3 = 1
--> MC_SEQ_MISC8
-- MR8
CLEHF = 1, WREHF = 1, RFU = 0, BA0 = 0, BA1 = 0, BA2 = 0, BA3 = 0
-- MR7
PLL Stby = 0, PLL Fclk = 0, PLL DelC = 0, LF Mode = 0, Auto Sync = 0, DQ PreA = 0, Temp Sensor = 0, HVFRED = 0,
VDD Range = 0, RFU = 0, BA0 = 0, BA1 = 0, BA2 = 0, BA3 = 0
--> MC_ARB_DRAM_TIMING
ACTRD = 25, ACTWR = 13, RASMACTRD = 47, RASMACTWR = 57
--> MC_ARB_DRAM_TIMING2
RAS2RAS = 157, RP = 45, WRPLUSRP = 46, BUS_TURN = 23
Lots of options.. lots of things to fine tune.. |
|
|
|
WOW - thats a bold statement -- I would suggest doing a little more research before making wild statements like that, The Stilt is a well known and respected guy in the AMD world.  In regards to the other comments, welcome to the club, several of us (myself, wolf0 etc) have put in countless hours Respect is irrelevant. I said 'Out of historic context it is hard to tell who copied who.' It is not bold, it just states what either vendors used Stilt work or other way around. Who done it is irrelevant (at least for me), changed values are not. Respect maybe irrelevant to you but not to me. I respect those that put in the work. Ultimately what matters on this forum is if the work yields a result. I am hitting 530+ mining ZEC with a Fury X and a modified BIOS way under stock voltages. There isn't a GDDR5 based card that can touch the numbers HBM cards can put up. Obviously it isn't only the HBM but it certainly doesn't hurt Yeah lol, now that a lot is out of the box, this thread is back to being super active.. Funny how "certain" people come back to live when free things are given out... It's one thing to have the tools, it's another thing to know what you're doing... (where are them mode registers  ) |
|
|
|
There is some Elpida mem out there, that even when modded with the 1500 strap, that seems only stable at 26-26.5 mhs or so
I experienced this with a 480 4gb nitro (not nitro+)
They do 30Mh as well.. (those nitro's) Got 2 Elpida 4G's, one nitro and one non nitro, the nitro one does 30+ and the non nitro one does 29.5 stable. But this is with custom timings. Greetings |
|
|
|
A shitload of potential here, lets hope they make something out of it..
Before whatever that jstenfanop is building makes them irrelevant I've seen what he's been working on. Not bad, but i prefer a 6xgpu rig over a 10x rig any day. (Non MXM GPU's) Yeah yeah, i know, more mobo's, etc.. but if i have a rig down, i don't have 10+ cards down. Plus, this (pandaminer) is interesting because of the MXM gpu's and their power efficienty, i believe what jstefanop is doing is basically a pci splitter..?.. based for Regular GPU's.. (correct me if i'm wrong) and so kind of irrelevant in comparison with this project. |
|
|
|
I really hope they will send the next batch or whatever in a decent state. I recieved my pandaminer (not b3) last week, full of dust, scratches, clearly been used for a significant period of time.. (and i know i'm not the only one....) Not very professional if you ask me. Then i got a private message from someone else who recieved his panda with 3 heatsinks completely loose.... I was very happy with my very first order, arived here in time, in perfect condition and still running smooth. But to be honest, i'm not sure if i'm going to reorder anything else after haven seen this. Lets hope they turn around and start working on their marketing skills and perhaps some decent English A shitload of potential here, lets hope they make something out of it.. |
|
|
|
@tytanick is it possible for the optiminer client to show just the total hashrate and when you hover over it to show individual hashrates? I like how you have claymore eth setup but find it a bit frustrating to have to always open console to check total hashrate for optiminer.
claymore has this just because it has clean and very nice api. I didnt check but as i renember optiminer doesnt has one so it would be very hard to do this. You could ask optiminer developer to do that. also is claymore zec slower than optiminer? Mmm. this is super easy to do. Optiminer has a log flag, just log the output, grep the hashrates out of the log, send them and done. Not really seeing the "Very hard" part in this.. Optiminer is still faster on a lot of gpu's, plus, what you see is what you get. Claymore on the other hand takes his fee from what you see. Optiminer doesn't do this. His fee is running parallel on top of what you're seeing. Which is also a shitload better because the miner doesn't have to constantly switch to devfee and back. That's of course my humble opinion. |
|
|
|
its still a good price for what you are getting. 1500 to build a 8 card 470? Where? Not with quality parts, not even close. EVGA 1300 / 8 x Sapphire 470 / MB (Good luck getting it) / cpu / memory / ssd / Metal frame (aluminum is what I make)
Your at close if not over 1900
Price is decent that's why i have a bunch coming =) PLUS a Panda shirt lols
Absolutely right. You should look at this from a large farm perspective. Try to stack "normal gpu" rigs the way you could/would stack or place these.. I bet i can stack 50 of these while only being able to place 15 normal ones.. anyhow, 8 x 470 4G at an avarage of 170 dollar per gpu = 1360$ Find a decent PSU that can feed 8 GPU's, decent quality = around 300-350$? CPU, Mem, Risers and extra risers because we all know how awesome these are. How much? 200$? Mobo that supports 8 GPU's, definitely not under 100$, more likely to be near 200$ So very quickly calculated, that's around 2100$ Panda B3 = less than that.. |
|
|
|
I couldn't sleep thinking of ETH again so I was up at 2am. I got lucky and got a few.
Remember, one does not simply pee with a boner...  |
|
|
|
The panda is pretty expensive for 8 laptop gpu's. Dual mining will probobly break it.. Bether to buy cards and build yourself..
Ok then sp_, go get yourself a custom mainboard that supports 8 GPU's that is this compact and power efficient... Curious how "cheap" you're going to get it done Seriously, going to bring up the dual mining subject? If you want to shorten the lifespan of your gpu's (even non MXM), yeah dual mine.. risk of systems dropping, risk of frying gpu's, name it.. go ahead, it's your hardware.. And btw, i have done a dual mine test on those MXM gpu's, works just fine.. they will just die a lot sooner.. |
|
|
|
How can they be out of stock already??  Not very expensive, compact, ideal for a large farm. Plus there's some guy who managed to raise hash even more and reduce power.. so yeah, you can turn it however you want, it is a very interesting product.. |
|
|
|
Specifications Fixed computing power ETH 237MH / s ± 5% ZEC 1880 Sol / s XMR 5890H / S Wall power consumption ETH 1250 W + 12% ZEC 1200 W + 12% XMR 1120 Watt Power Efficiency Great Wall 1600W power supply, efficiency of about 93%, 25 ℃ ambient temperature Rated voltage 12.4-12.6V Number of cards RX 470 * 8
Dimensions 490 mm (L) * 280 mm (W) * 170 mm (H) cool down 3 * 2 fan Operating temperature 0 ° C to 40 ° C Working humidity 5% RH-95% RH, non-condensing Internet connection Ethernet Power connection Each graphics card mining machine 7 6pin mouth need to connect, it is recommended to use an independent power supply to the machine System parameters CPU Inter (R) Celeron (R) CPU 3215U @ 1.70 GHZ Install memory 4GB System type 64-bit operating system, x64-based processor system Windows 10 ELIOVP - any chance to super rom mod this animal?Of course, as soon as it arrives Just a small disappointment that it's 4G.. but oh well.. |
|
|
|
I am going to buy a rig from them. I learned a long time ago that pushing the coins is more important then profit from the coins. (profit is still nice)
I would hope to get a better deal from them then I did on the first rig.
But I am going to buy one of these anyway.
I like to be more independent and not owe good reviews for builders of gear in turn for free gear.
But a decent discount I will take if they give it to me.
They sold out the first set of 500 units pretty fast and gpu gear is in demand so they will sell this out if the price is okay.
My suggestion to them is try to use 470's with 8gb of ram.
Not 480's with 4 gb of ram.
I am looking forward to an email with some kind of an offer from them.
470 8G is a big win, if not the biggest win.. I can push those higher than any other cards out there |
|
|
|
Eliovp I messaged you regarding license purchase. Check your PM  Oh crap haha. Uhmm, i did not forget to reply!  eliovp - i just test ran your light version Nano ZEC rom - my nanos are struggling, maybe getting worn out and abused over the years -- can it go any lighter? I am using Optiminer 1.7 with smOS. Thanks i made those when it was optiminer 1.2 or something.. 1.7 is a lot more intensive, i'll see what i can do |
|
|
|
Eliovp I messaged you regarding license purchase. Check your PM Oh crap haha. Uhmm, i did not forget to reply! |
|
|
|
Just to be completely clear here. Before others start to fly and think they can undervolt.. 0 - 7 are the DPM stages for the core clock. You don't undervolt, you just set your core differently on a specific dpm stage. As an example. Open up polarisbioseditor, you'll see that there are 8 stages for the core clock. 0 -> 7  As you can see here, on the left you have the preset core clock settings and on the right you have the voltage settings. 300Mh being the only one with a preset voltage offset. The other 7 will be auto calculated by the GPU. (By adding Global VDDC Offset in the rom, this can be manipulated). What happens here isn't really "Undervolting", the only thing that does happen is that driverwise the core clock of a specific dpm stage is being modified, therfore the voltage remains the same as it was previously on that specific dpm stage. So you are basically forcing a specific core clock to run on a specific voltage. It's like this example:  Which is 1. Definitely not stable in most cases as it's not being autocalculated anymore. 2. Not how it "should" be. It's a way, i'm pretty sure it'll work okish, but i'm also very sure, if you start messing around to much, your gpu's will fail. That's the difference between forcing voltage or adding a global vddc offset in the rom where the voltage controller starts calculating from a lower voltage offset. Just to be very clear here, it is "not" possible to really "undervolt". This can only be achieved in the rom itself. Oh and for the Fans, or better, the haters, i'm not bashing anything here, i'm just writing down facts. Greetings! |
|
|
|
|
Show Posts
