R Mode memory allocation question on Instinct MI25 and MI60. Is the memory allocation acting as expected, leaving ~10 / 26GiB free on the MI25s and MI60? TRM 0.9.4.7 Linux 5.10.0-hiveos #83.hiveos.211201 SMP Wed Dec 1 11:51:04 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux AMD Driver 20.40 (5.11.1001) 0 04:00.0 Radeon Pro V320 16368 MB · 113-D0515300-101 1 08:00.0 Radeon Pro V320 16368 MB · 113-D0515300-101 2 0c:00.0 Radeon Pro VII 32752 MB · 113-D1630600-107 Team Red Miner version 0.9.4.7
[2022-05-13 21:16:27] Linux kernel vm mode parameters are already correct.
[2022-05-13 21:16:27] Auto-detected AMD OpenCL platform 0
[2022-05-13 21:16:28] Initializing GPU 0.
[2022-05-13 21:16:28] Initializing GPU 1.
[2022-05-13 21:16:28] Initializing GPU 2.
[2022-05-13 21:16:28] GPU 0 Vega boost applied.
[2022-05-13 21:16:28] GPU 0 Starting tuning.
[2022-05-13 21:16:28] GPU 1 Vega boost applied.
[2022-05-13 21:16:28] GPU 1 Starting tuning.
[2022-05-13 21:16:28] GPU 2 Radeon VII boost applied.
[2022-05-13 21:16:28] GPU 2 Starting tuning.
[2022-05-13 21:16:28] Successfully initialized GPU 0: Vega with 64 CU (PCIe 04:00.0)
[2022-05-13 21:16:28] Successfully initialized GPU 1: Vega with 64 CU (PCIe 08:00.0)
[2022-05-13 21:16:28] Successfully initialized GPU 2: Vega2 with 64 CU (PCIe 0c:00.0)
...
[2022-05-13 21:16:33] GPU 0 DAG allocation trying 16368.0 - 9712.0 = 6656.0 MiB.
[2022-05-13 21:16:33] GPU 2 DAG allocation trying 32752.0 - 26096.0 = 6656.0 MiB.
[2022-05-13 21:16:33] GPU 1 DAG allocation trying 16368.0 - 9712.0 = 6656.0 MiB.
...
[2022-05-13 21:16:36] GPU 0 Starting DAG generation for epoch 492, 4.84 GB (single buffer mode).
[2022-05-13 21:16:37] GPU 1 Starting DAG generation for epoch 492, 4.84 GB (single buffer mode).
[2022-05-13 21:16:37] GPU 2 Starting DAG generation for epoch 492, 4.84 GB (single buffer mode).
[2022-05-13 21:16:41] GPU 2 DAG generation finished in 3.498 seconds.
[2022-05-13 21:16:41] GPU 0 DAG generation finished in 5.149 seconds.
[2022-05-13 21:16:42] GPU 1 DAG generation finished in 5.150 seconds.
Thanks for trm! |
|
|
|
0.8.0 on Radeon VII -- ! [2021-01-19 12:01:56] ------------------------------------------------------------------------------------------------
[2021-01-19 12:01:56] Mining ethash with 1 GPU workers
[2021-01-19 12:01:56] GPU PCIe CUs CoreMHz SocMHz MemMHz TEdge TJct TMem FanPct FanRpm VDDC ETH Cfg
[2021-01-19 12:01:56] 0 04:00.0 60 1940 971 1200 53C 85C 69C 82.75% 3216 1025 mV C320
[2021-01-19 12:01:56]
[2021-01-19 12:01:56] Stats Uptime: 0 days, 00:27:30
[2021-01-19 12:01:56] ----------------------------------------- GPU Status -------------------------------------------
[2021-01-19 12:01:56] GPU 0 [53C, fan 82%] ethash: 120.0Mh/s, avg 119.4Mh/s, pool 121.0Mh/s a:131 r:0 hw:0
[2021-01-19 12:01:56] Total ethash: 120.0Mh/s, avg 119.4Mh/s, pool 121.0Mh/s a:131 r:0 hw:0
[2021-01-19 12:01:56] ----------------------------------------- Pool Status ------------------------------------------
[2021-01-19 12:01:56] daggerhashimoto.usa.nicehash.com ethash: 119.0Mh/s, avg 118.2Mh/s, pool 121.0Mh/s a:131 r:0
[2021-01-19 12:01:56] ------------------------------------------------------------------------------------------------ That's at 285W, so 421MH/s/kW. |
|
|
|
i am actually interested on how loud this miner can be. i did not find any info on the matter on their site. only that the fans push 206 cfm of air.
Hair-dryer loud in efficient mode, vacuum cleaner loud during regular operation. -ck, the following in cgminer readme is tantalizing: --T1fantarget <arg> Throttle T1 frequency to keep fan less than target fan speed (default: 100)
Is there any chance of exposing this parameter via the web interface or API? This would be so helpful in using T1s as space heaters where desired power output varies with ambient temperature. Thanks in any case. |
|
|
|
Thanks for the great work on this firmware series -ck. I'm loving the fact that in efficiency mode I can actually work in the same room without hearing protection  Do we need to re-run tuning to take advantage of the changes in 20180425_055323? Did the fan controller change from 20180421? It's overshooting/hunting more in my environment (efficiency mode / 25C room / ~35% fan speed). Lastly, a feature request for consideration - is it feasible to store the tuning results for the various modes so the tuning process isn't necessary on mode switch? |
|
|
|
So Halong are people going to have to wait for their miners? Ba dum tsckkk...
I guess is best to wait for people in the thread to receive and put reviews with pics and videos first. It is probably a few days more only. My shipment now has a delivery day estimate of April 3rd. It has departed the sort facility in Hong Kong. As long as there are no Duty/Customs delays I'll put pics up on Tuesday. |
|
|
|
Just wanted to update the thread that I got my DHL tracking information today from Halong.
Unfortunately it just shows "Shipment information received" so far, but I'll update again if I get an actual shipment notification from them.
I just logged in to post the same thing. Just received the following from DHL. My order status on halongmining.com is still "Processing". DHL EXPRESS SHIPMENT ADVISORY The following 6 piece(s) have been sent by Tina Ana from Shipme Limited via DHL Express on 28-Mar-2018 via AWB# [redacted] If you wish to track this(these) shipment(s) please contact your local DHL customer service office or visit the DHL Web Site at http://www.dhl.com.hk If you have a Web-enabled mail reader, click the link below to view shipment tracking details: http://www.dhl.com/cgi-bin/tracking.pl?AWB=[redacted] SHIPMENT TO: [redacted] SHIPMENT FROM: Shipme Limited Tina Ana NO 8 TAI CHUNG ROAD TCL Tower NEW TERRITORIES Hong Kong SHIPMENT CONTENTS: DragonMint Bitcoin Miner and Switching Power Supply SHIPPER REFERENCE: Halong [redacted order number] ADDITIONAL MESSAGE FROM SHIPPER: RECAP AWB:[redacted] WEIGHT:26.7 PIECES:6 CONTENTS:DragonMint Bitcoin Miner and Switching Power Supply REF:Halong [redacted order number] Thank you for requesting DHL Express for your delivery needs |
|
|
|
|
My batch 1 order (Order # 4xx) still shows status "Processing", as it has since I placed the order. I'll post here if that changes.
Moderator's note: This post was edited by frodocooper to remove a quote from a deleted post.
|
|
|
|
Looking at the logs, there was only one time that the command included a hashed password (the last time) - so it would seem the password field was always blank (except the last time)
Do you mean in your miner configuration? Just use 'x' or '123' or similar. At the risk of sound really really stupid...what do I enter as the password for my BTC address? Blank obviously doesn't work, and coinbase does not issue me a password for my BCT address. Where do I get a password for a BTC address? Thanks in advance for your reply...eager to try out KanoPool! |
|
|
|
|
A few more notes on the youtube video. These notes are not an endorsement.
The driver returning results is BA1.
Results come from three different BA1 devices - 0, 1, and 2, presumably the 3 boards.
Chip numbers as high as 62 are reported, so at minimum there are 3*62 (or 63) = 186 or 189 chips per tube.
(Claimed DM8575 hashrate is 85GH/s/chip; 16TH/85GH = 189 chips)
The elapsed time displayed on the watt meter is 11h18m45s; the slush hashrate graph shows activity from about 22:35 to the end at 9:05, about 10h30m.
The timestamps in the log are consistent with the end time on the slush hashrate graph.
|
|
|
|
And some questions have arisen about the validity of the performance of the machine, if you watch the frames starting at 0:31s and ending 0:33s in the video you can see that the hash rate is 4TH but then on frame at 0:32s that line is edited out of the frame, and then it switches to slush "screenshot" at 0:33s.
Share Submission Rate is Plausible The log lines we see cover the interval [2017-11-22 09:08:10.720] to [2017-11-22 09:08:38.904], or 18.2 seconds. We can see 15 share accepted messages during that interval, or about 50 shares per minute. The pool difficulty is 4948. This means each 25 shares per minute corresponds to 4948*1.79GH/s (see -ck's post https://bitcointalk.org/index.php?topic=274023.0). So we have 50/25*4948*1.79 = 17.7TH/s. Since we observed 15 events the std dev is about 3.8 and we can say that the hash rate likely lies between 13TH/s and 22TH/s. |
|
|
|
The specs state 1205W DC consumption and 1432W at the wall. That's 84%, but their power supply specs say 92%. Either they don't have their PSUs on hand yet, they only hit 92% efficiency under very different load conditions, or 1205W doesn't include the cooling fans.
If 1205W is accurate, a DPS-2980AB will run about 2550W at the wall for two of them, which is 0.079J/GHash.
I don't think 1205W does include the fans. The specs list 100W consumption for the fans separately. For comparison the S9 fans are 50W at 100%. I should add that all the posters in this thread pointing out scam warning flags are absolutely right and you shouldn't place an order for these miners if losing the money would be a disaster for you. |
|
|
|
Based on my order number they took in a minimum of ~360BTC assuming 100% payment rate. |
|
|
|
I Pulled the trigger on an ASIC pre-order for the first time in about four years. Also ordered two of their power supplies for tear-down and efficiency measurement. Anyone gonna run an order tracking thread on this, or am I stuck in 2013 thinking that's a thing? Notes and thoughts - The controller board and hashboards don't match S9 or T9. There are a lot of similarities, of course, but they are easily distinguished beyond simple mirroring. The hashboard ribbon cables have the same number of connectors, however. The specs state 1205W DC consumption and 1432W at the wall. That's 84%, but their power supply specs say 92%. Either they don't have their PSUs on hand yet, they only hit 92% efficiency under very different load conditions, or 1205W doesn't include the cooling fans. If 1205W is accurate, a DPS-2980AB will run about 2550W at the wall for two of them, which is 0.079J/GHash. |
|
|
|
Greetings,
I'm using psu from bladecenter and I wonder if theres a chance to manage eg. fans in this unit. After turning them on, they're spinning at max speed (after while) even when nothing is connected to the PSU. Air blowing from the back is cold. I have few questions about that:
1. Is there any way to manage fan speed (reduce it)?
2. PSU only connected to AC draws about 70W (fans turned off) from the wall - is it normal?
1. No straightforward way to reduce fan speed. You can talk PMBus to the power supply and convince it it is in a healthy chassis or splice into the fan supply lines and control them that way, but no easy method that I'm aware of. 2. None of the IBM bladecenter supplies I have experience with are happy being plugged in but not running. Their standby consumption is high enough that internal components can easily overheat without the fans running. I've lost DPS-2000BB and DPS-2980AB supplies to this. |
|
|
|
Thanks so much for this - working great here. I suggest -dcri 225 for DaggerSia on 1080Ti right now btw   One issue: Update version: 1.7.5.15-modChangelog: ... - Fix Lyra2REv2 Algorithm returning crazy ass numbers (i.e: 2000 USD/day)
... I had to multiply my Lyra2REv2 speed by 1000 to get sensible numbers out of this build (and sorting is broken for profit > $9.99): Profits for GPU#1 ASUS GeForce GTX 1080 Ti:
PROFIT = $10.32 (SPEED = 754 | NHSMA = 4847.135) [ewbf_Equihash]
PROFIT = $10.98 (SPEED = 3300000000 | NHSMA = 0.00117783) [ccminer_Sia]
PROFIT = $11.78 (SPEED = 28300000/2114000000 | NHSMA = 0.059396/0.00117783) [ClaymoreAMD_DaggerSia]
PROFIT = $2.04 (SPEED = 433830 | NHSMA = 1.662647) [experimental_NeoScrypt]
PROFIT = $3.77 (SPEED = 11510000 | NHSMA = 0.11591) [ccminer_X11Gost]
PROFIT = $5.56 (SPEED = 33164250 | NHSMA = 0.059396) [ethminer_DaggerHashimoto]
PROFIT = $6.68 (SPEED = 33000000/330000000 | NHSMA = 0.059396/0.00122464) [ClaymoreAMD_DaggerPascal]
PROFIT = $6.68 (SPEED = 5266660000 | NHSMA = 0.00044877) [ccminer_Decred]
PROFIT = $6.70 (SPEED = 1937340000 | NHSMA = 0.00122464) [excavator_Pascal]
PROFIT = $7.53 (SPEED = 23000000 | NHSMA = 0.11591) [ccminer_alexis_X11Gost]
PROFIT = $8.13 (SPEED = 27000000/2837000000 | NHSMA = 0.059396/0.00044877) [ClaymoreAMD_DaggerDecred]
PROFIT = $9.10 (SPEED = 664.733373 | NHSMA = 4847.135) [excavator_Equihash]
PROFIT = $9.44 (SPEED = 73550000000 | NHSMA = 0.045437) [ccminer_Lyra2REv2]
PROFIT = $9.61 (SPEED = 541990000 | NHSMA = 0.00627472) [ccminer_Lbry]
PROFIT = $9.88 (SPEED = 27600000/296000000 | NHSMA = 0.059396/0.00627472) [ClaymoreAMD_DaggerLbry]
MOST PROFITABLE ALGO: ClaymoreAMD_DaggerSia, PROFIT: $11.78
|
|
|
|
|
Stamp wall finally coming under pressure - heads up!
|
|
|
|
From the Technology Review article: The structure of Bitcoin means that if a small number of users hold the majority of the currency, as is the case with these Chinese companies, then they are able to veto changes to the underlying technology.
/eyeroll |
|
|
|
for those of you wandering if this is the auto tune version. This pic seems to be a stock 14TH model. so it will be set to hash at 650M frequency by the firmware. Has anyone tried the firmware from 11-19? I'm giving it a shot on one machine right now. Curious how it is working for others?
Fan speed is much better handled for me - slight variations in speed, gently ramped, instead of the bang-bang controller in the previous autotune FWs. Seeing much lower reported temps (~10C lower) but I haven't dug into whether that is just a cosmetic change or a change in setpoint for target temp. Hashrate is good - 13.904avg over 3025 hours on a batch 1 set to 650MHz (theoretical speed 14.004): Nice of you to jump in, and still be following the thread. but please, can we discuss the elephant in the room? wtf is with these fans They idle so low one minute, and the next minute my neighbors can hear them. I have no issue with noisy miners, but this seems like it's just unrequired noise. One minute they blast at 110% the next they are cruising at 25%. than boom back up again. If they idled Higher, at 60% maybe? they would need to spike as often, and this would help with the noise greatly. takagari - This is what I posted about re: the 19 Nov firmware. For me, this firmware stops the fan cycling and replaces it with much subtler changes in speed. Higher average speed over all, but barely perceptable speed changes. It's a big improvement on my S9, which is a Batch 1 / 14TH/s as BitmainWarranty suggested. |
|
|
|
Has anyone tried the firmware from 11-19? I'm giving it a shot on one machine right now. Curious how it is working for others?
Fan speed is much better handled for me - slight variations in speed, gently ramped, instead of the bang-bang controller in the previous autotune FWs. Seeing much lower reported temps (~10C lower) but I haven't dug into whether that is just a cosmetic change or a change in setpoint for target temp. Hashrate is good - 13.904avg over 3025 hours on a batch 1 set to 650MHz (theoretical speed 14.004):  That's awesome, so does it allow fan tweaking or tweaking of the Hastrate? Nope :\ At least not through the web UI and I haven't the time to poke through the config files and see where that's being handled now. |
|
|
|
|
Show Posts
