v37.4

readme.md

@@ -30,7 +30,7 @@ GPU Miner for `ETH`, `RVN`,  `GRIN`, `BEAM`, `CFX`, `ZIL`, `ERGO`, `AE`, `SERO`
 | kawpow           |   RVN   |   10.3M    |   17.5M    |   13.3M    |  22.5M   |  25.8M   |     11M      |
 | beamv3           |  BEAM   |    12.5    |    19.6    |    19.5    |    26    |   30.5   |      X       |
 | octopus          |   CFX   |    5.5M    |    8.5M    |    9.8M    |  14.8M   |  48.5M   |     X     |
-| ergo | ERGO | 41M | 67M | 52M | 63M | 73M | X |
+| ergo | ERGO | 41M | 67M | 52M | 63M | 73M | 83M(eth bios) |
 
 ## Features
 
@@ -162,7 +162,8 @@ GPU Miner for `ETH`, `RVN`,  `GRIN`, `BEAM`, `CFX`, `ZIL`, `ERGO`, `AE`, `SERO`
 * --proxy    Socks5 proxy used to eastablish connection with pool, E.g. 127.0.0.1:1080
 * --cuckoo-intensity \<intensity>    Set intensity of cuckoo, cuckaroo, cuckatoo, [1, 12]. Smaller value means higher CPU usage to gain more hashrate. Set to 0 means autumatically adapt. Default: 0.
 * --cuckatoo-power-optimize    Set this option to reduce the range of power consumed by rig when minining with algo cuckatoo. This feature can reduce the chance of power supply shutdown caused by overpowered. Warning: Setting this option may cause drop on minining performance.
-* --temperature-limit \<temp-limit>    Set temperature limit of GPU in Celsius. If it exceeds, stop the GPU for 10 seconds and continue.
+* --temperature-limit, --tl \<temp-limit>    Set temperature limit of GPU, if exceeds, stop GPU.
+* --temperature-start, --ts \<temp-start>    Set cool-down temperature target if GPU is stopped by `temperature-limit`, default to \<temp-limit> - 5.
 * --log    Generate log file named `logs/log_<timestamp>.txt`.
 * --log-file \<filename>    Generate custom log file. Note: This option will override `--log`.
 * --no-nvml    Do not query cuda device health status.
@@ -273,6 +274,13 @@ GET http://api_host:port/api/v1/status
 
 ## Change Log
 
+#### v37.4(2021-05-20)
+
+- `new algo`: `ergo` for AMD GPU, can be faster with ETH mining timings. typical hashrate
+  - `580 8G`: 83MH/s
+  - `5700`: 125MH/s
+- `feature`: use `--temperature-limit` & `--temperature-start` to protect GPU from overheat, detail in readme.md
+
 #### v37.3(2021-05-06)
 
 - `feature`: add option `--enable-dag-cache` to allow an extra `DAG` for different `epoch` cached in GPU memory, useful for `ETH+ZIL` mining and mining on `NiceHash`.

readme_zh.md

@@ -24,7 +24,7 @@ NVIDIA、AMD显卡的`ETH`, `RVN`,  `GRIN`, `BEAM`, `CFX`, `ZIL`, `ERGO`, `AE`,
 | kawpow           |   RVN   |   10.3M    |   17.5M    |   13.3M    |  22.5M   |  25.8M   |     11M      |
 | beamv3           |  BEAM   |    12.5    |    19.6    |    19.5    |    26    |   30.5   |      X       |
 | octopus          |   CFX   |    5.5M    |    8.5M    |    9.8M    |  14.8M   |  48.5M   |     X     |
-| ergo | ERGO | 41M | 67M | 52M | 63M | 73M | X |
+| ergo | ERGO | 41M | 67M | 52M | 63M | 73M | 83M(eth bios) |
 
 ## 功能特点
 
@@ -154,7 +154,8 @@ NVIDIA、AMD显卡的`ETH`, `RVN`,  `GRIN`, `BEAM`, `CFX`, `ZIL`, `ERGO`, `AE`,
 - --proxy    使用Socks5代理连接矿池，例如: 127.0.0.1:1080
 - --cuckoo-intensity \<intensity>    设置挖Grin时的CPU负载，取值范围[1,12]，值越小挖矿算力越高，相对应的CPU负载也会更高。设置为0软件从1开始自适应调整。默认为0
 - --cuckatoo-power-optimize    减小多卡矿机挖Grin31的总功耗波动，避免电源过载关机（设置该选项可能导致算力略微降低，请测试后谨慎使用）
-- --temperature-limit \<temp-limit>    设置显卡温度上限，一旦超过，停止挖矿10s后继续
+- --temperature-limit, --tl \<temp-limit>    设置显卡温度上限，一旦超过，停止挖矿
+- --temperature-start, --ts \<temp-start>    设置显卡由于达到设定的温度上限停止后重新启动挖矿的温度条件，默认为\<temp-limit> - 5 度.
 - --log    生成日志文件，文件名为 `log_<时间戳>.txt`.
 - --log-file \<filename>    自定义日志文件名，该选项将覆盖`--log`.
 - --no-nvml    不周期性地查询GPU的温度功耗等状态.
@@ -265,6 +266,13 @@ GET http://api_host:port/api/v1/status
 
 ## 修改记录
 
+#### v37.4(2021-05-20)
+
+- `新增算法`: `ergo` 在AMD显卡上的挖矿，配合ETH挖矿时序使用速度更快。参考典型算力：
+  - `580 8G`: 83MH/s
+  - `5700`: 125MH/s
+- `功能`: 调整温控策略，通过设置 `-tl`限制温度上限以停止挖矿，`-ts`设置重新启动挖矿的温度条件，详见readme_zh.md
+
 #### v37.3(2021-05-06)
 
 - `功能`: 新增`--enable-dag-cache`选项，可额外在GPU显存中缓存另一个`epoch`的`DAG`文件，可用于`ETH+ZIL`挖矿以及在`NiceHash`挖矿的`DAG`快速切换