Nvidia GT 1030 Mining 2 GB: Overclocking | Hashrate | Power Consumption – Mining on an Nvidia GT1030 video card with 2 gigabytes of memory – This article takes a closer look at the features of the Nvidia GT1030 video card in terms of its suitability for mining.
About the computational capabilities of the Nvidia GT1030 graphics card
The computational capabilities of the Nvidia GT1030 video card are severely limited by the presence of only two gigabytes of video memory operating on a 64-bit bus at 6 Gbps. This provides a ridiculous exchange rate of roughly 50 GB / s:
With such a low-performance memory bus, you should not count on a hash rate when mining on daggerhashimoto (ethash) more than 5-7 mh / s (even when overclocking and using optimized timings).
Slightly improves the situation with mining on the 1030th, support for the CUDA 6.1 version and a fairly high frequency of the video chip of the order of 1400-1500 MHz:
With 384 stream processors and GPU frequencies up to 1.5 GHz, the Nvidia GT1030 graphics card delivers 35.23 GFLOPS (FP64 performance) floating point performance. Considering the low power consumption, limited by a TDP of 30 watts, this is not too bad. For comparison, the Nvidia GT750Ti with a Kepler core has 43.4GFLOPS processing power at a TDP of 60 watts.
The budget Nvidia GT1030 is significantly inferior to the AMD Radeon RX550, which has a performance of 75.7 GFLOPS with a consumption of 50 watts.
Hardware features of the Nvidia GT1030 video card on the example of the low profile model manufactured by Palit
Palit’s Nvidia GTX1030 graphics card has a modest cooler and cooling power phases (one for the video chip and one for memory):
On the back of the Palit-made Nvidia GT1030 board, one of the significant elements is the IS25WQ040 flash memory chip:
Video chip GT1030 marked GP108-310-A1:
Two gigabytes of memory are recruited with two Micron high-density microcircuits, labeled 8LA47 D9TCB:
The only power phase of the GP108 video chip is assembled on two MOSFET transistors 4503NH (Umax = 30V, Imax = 80A) and one 4507NH (Umax = 30V, Imax = 60 A), which are controlled by a PWM controller uP1666Q:
Typical wiring diagram for uP1666Q in single-phase mode:
Nuvoton 3941S-A microcircuit (linear voltage regulator for fan):
Nuvoton 3941S-A typical wiring diagram:
The only memory power phase on the Palit Nvidia GT1030 video card is formed by the RT8128C PWM controller and two SM3319 MOSFET transistors, each of which can dissipate 12.5 W of power (Uds max = 30V, Id max = 14 A):
The electrical circuit of the memory power phase on the PWM controller (typical) RT8128C from Richtek Mach Response:
The RT8128C PWM controller operates at a fixed frequency of 300kHz with input voltages from 4.5 to 13.2 volts, and can deliver Uout from 0.6 to 5V.
Block diagram of the RT8128C PWM controller:
One phase of power supply with low-power transistors on a video card with a purely symbolic cooling is clearly not enough, despite the presence of only two GDDR5 memory modules.
Weak cooling of the power circuit plus the lack of cooling / blowing leads to serious breakdowns like this (burnout of the SM3319 field-effect transistor):
The IS25WQ040 flash memory chip is located on the rear side of the board (powered by 1.8V):
Linear voltage regulator LD1117AG, forming 3.3V, on the back side of the board (Uin = 4.75-10V, Uout = 3.3 volts, Iout = 800 mA):
The following is the hashrate information for an Nvidia GT1030 video card with 2 GB GDDR5 memory with driver version 430.64 (CUDA Version 10.1) in xubuntu 16.04 and in Windows 8.1 with driver 461.40 (CUDA version 11.2).
Hashrate of Nvidia GT1030 video card with 2 GB GDDR5 memory during mining
The Nvidia GT1030 video card is not outdated yet, therefore it supports work with many miners. The range of its application is limited by a small memory size (2 GB) and insufficient memory bus bandwidth (about 50 GB / s). In this regard, it makes sense to run the Nvidia GT1030 on algorithms that use the core more and do not require a large amount of VRAM.
An example of a script for mining UPX coins on an Nvidia GT1030 video card in the CryptoDredge program:
./CryptoDredge -a cnconceal -o stratum + tcp: //cryptonight_upx.eu.mine.zergpool.com: 4457 -u wallet -pc = BTC, worker -no-crashreport --no-color --temperature-limit 65 - -temperature-start 50 --pstate = 0 --device 0 --api-type off --intensity 5
Mining on the cn_gpu algorithm in the CryptoDredge program (hash rate 210-230 h / s at an intensity of 5, VRAM load by a miner – 559 MiB):
Mining on the cn_haven algorithm in the CryptoDredge program (hash rate 160-170 h / s at an intensity of 5, VRAM load by a miner – 1071 MiB):
On Windows 8.1 with newer drivers version 461.40, CryptoDredge 0.25.1 provides a hash rate of 170-180 h / s on cnhaven with a consumption of 30 watts:
On the verthash algorithm in the VerthashMiner 0.7.0 program, the Nvidia GT1030 outputs up to 100 kh / s:
Mining on the ethash algorithm in Phoenix Miner (DBIX coin), hash rate 5-6.5 mh / s in xubuntu 16.04:
Mining on Nvidia GT1030 on the Megabtx algorithm in the Trex miner:
Summary information on the hashrate of the Nvidia GT1030 video card with 2 GB VRAM GDDR5-type (with a connected monitor) based on the results of running miners in xubuntu 16.04 (drivers 430.64, stock settings) and in Windows 8.1 (driver 461.40, overclocking with nvidia inspector +70 in the core and +150 from memory):
- Ethash – 5-7 mh / s (Phoenix Miner 5.5c), NBMiner squeezes 6-7 Mh / s in Windows 8.1;
- Cryptonight_GPU – 210-230 h / s – xubuntu 16.04, 240-250 – Windows 8.1 (CryptoDredge 0.25.1), VRAM load in Windows 8.1 – 694 MB;
- CN_GPU – 240 h / s (xubuntu 18.04, CryptoDredge 0.25.1, intensity 6);
- CN_Haven – 160-170 h / s – xubuntu 16.04, 170-180 – Windows 8.1 (CryptoDredge 0.25.1), VRAM loading in Windows 8.1 – 1205 MB;
- phi2 – 1.2-1.3 Mh / s (xubuntu 18.04, CryptoDredge 0.25.1, intensity 6);
- CN_Conceal (UPX2) – 4.3 -4.5 kh / s (CryptoDredge 0.25.1), VRAM loading in Windows 8.1 – 284 MB;
- Argon2D-dyn – 25-26 kh / s, VRAM consumption 1022 MB (Windows 8.1, CryptoDredge 0.25.1);
- Verthash – 90-100 kh / s, (Windows 8.1 and xubuntu 18.04, VerthashMiner 0.7.0), VRAM load – 1491MB in Windows 8.1 (a monitor is connected to the video card) and 1280MB in xubuntu 18.04 (without a connected monitor);
- Equihash 144.5 – 5-5 sol / s, VRAM consumption 1965 MB (Windows 8.1, miniZ 1.7 × 3);
- Megabtx (Bitcore coin ), megamec (Megacoin) – 1.1-1.2 Mh / s, VRAM consumption 290MB (Windows 8.1, T-Rex 0.19.11).
The Nvidia GT1030 graphics card with 2 GB of GDDR5-type memory features low power consumption and shows a low hash rate on Ethash even on Linux. Nevertheless, it has good performance on algorithms of the Cryptonight, megabtx, megamec and verthash families, which allows it to be used for profitable mining in 2021.
Given the small amount of VRAM (2 GB), the GT1030 graphics card cannot be used for algorithms such as etchash, equihash 192 and 125, kawpow , as well as for most Etahsh coins, including Ethereum.
When mining on Nvidia GT1030 video cards intended for office work, it is imperative to provide thermal protection by software (temperature limit in the miner, as well as in the nvidiainspector utility or MSI Afterburner), as well as ensure good ventilation of the video card.
Depending on the hardware implementation of the Nvidia GT1030 video adapters, it is worthwhile to take care of adhesive tape of additional radiators on the power supply transistors, which, based on experience, sometimes fail.
On video cards with a passive cooling system, it is imperative to install fans or completely change it, for example, replacing it with a cooling system from MSI Aero series video cards with one fan with re-drilled holes.