Errors found during quantizing PrimeIntellect-INTELLECT-3-FP16

[blackcat1402]

I can successfully quantize GLM4.5 Air. however, when I quantized one of its variant PrimeIntellect-INTELLECT-3-FP16, it reported errors as below and failed:

Microsoft Windows [version 10.0.19045.6466]
(c) Microsoft Corporation

C:\Users\blackcat1402>cd exllamav3

C:\Users\blackcat1402\exllamav3>python convert.py ^
More?   -i C:\Users\blackcat1402\PrimeIntellect-INTELLECT-3-FP16 ^
More?   -o C:\Users\blackcat1402\PrimeIntellect-INTELLECT-3-exl3-5.76bpw ^
More?   -w C:\Users\blackcat1402\exl3_working ^
More?   -b 5.76 ^
More?   -d 2
Detected Windows operating system. Triton does not have an official Windows release, thus FLA will not be adapted for Windows, and any potential errors will not be fixed. Please consider using a Linux environment for compatibility.
 -- Creating new job
    Input directory: C:\Users\blackcat1402\PrimeIntellect-INTELLECT-3-FP16
    Output directory: C:\Users\blackcat1402\PrimeIntellect-INTELLECT-3-exl3-5.76bpw
    Working directory: C:\Users\blackcat1402\exl3_working
    Calibration size: 250 rows, 2048 columns
    Target bitrate: 5.76 (decoder), 6 (head)
    Output scales: auto
    Codebook: mcg
 -- Loaded model config
    Architecture: Glm4MoeForCausalLM
 -- Created model instance:
     - Glm4MoeModel
         - Embedding
         - TransformerBlock
             - RMSNorm
             - Attention
                 - [4x] Linear
             - RMSNorm
             - GatedMLP
                 - [3x] Linear
         - [45x] TransformerBlock
             - RMSNorm
             - Attention
                 - [4x] Linear
             - RMSNorm
             - BlockSparseMLP
                 - [385x] Linear
                 - GatedMLP
                     - [3x] Linear
         - RMSNorm
         - Linear
 -- Loaded tokenizer
    Vocab size: 151367
 -- Preparing input state
 -- Loading unquantized module: model.embed_tokens
 -- Quantized: model.embed_tokens                                           bpw: 16.00  rfn: 0.000000  cos: 0.000000  sqnr: 0.000000  [6.28 s]
 -- Loading unquantized module: model.layers.0
 -- Captured: model.layers.0
 -- Quantized: model.layers.0.self_attn.q_proj                              bpw:  6.00  proxy_err: 0.000025  .  g_sc: 0.405454  [9.13 s]
 -- Quantized: model.layers.0.self_attn.k_proj                              bpw:  7.00  proxy_err: 0.000006  .  g_sc: 0.467450  [1.35 s]
 -- Quantized: model.layers.0.self_attn.v_proj                              bpw:  7.00  proxy_err: 0.000008  .  g_sc: 0.629755  [1.32 s]
 -- Quantized: model.layers.0.self_attn.o_proj                              bpw:  6.00  proxy_err: 0.000024  o  g_sc: 0.691751  [10.49 s]
 -- Quantized: model.layers.0.mlp.up_proj                                   bpw:  5.00  proxy_err: 0.000160  o  g_sc: 0.859647  [8.23 s]
 -- Quantized: model.layers.0.mlp.gate_proj                                 bpw:  5.00  proxy_err: 0.000123  o  g_sc: 0.863102  [8.14 s]
 -- Quantized: model.layers.0.mlp.down_proj                                 bpw:  6.00  proxy_err: 0.000032  o  g_sc: 0.806696  [8.69 s]
 -- Quantized: model.layers.0                                               bpw:  5.67  rfn: 0.005075  cos: 0.000013  sqnr: 46.547755  [80.01 s]
 -- Estimated remaining time: 1 hour, 4 minutes
 -- Loading unquantized module: model.layers.1
 -- Captured: model.layers.1
 !! Warning: block.mlp.0.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.1.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.10.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.100.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.101.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.102.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 -- Quantized: model.layers.1.self_attn.q_proj                              bpw:  6.00  proxy_err: 0.000017  .  g_sc: 0.376184  [8.21 s]
 -- Quantized: model.layers.1.self_attn.k_proj                              bpw:  7.00  proxy_err: 0.000002  .  g_sc: 0.525990  [1.31 s]
 -- Quantized: model.layers.1.self_attn.v_proj                              bpw:  7.00  proxy_err: 0.000007  .  g_sc: 0.644391  [1.34 s]
 -- Quantized: model.layers.1.self_attn.o_proj                              bpw:  6.00  proxy_err: 0.000031  o  g_sc: 0.786471  [10.38 s]
 -- Quantized: model.layers.1.mlp.experts.0.up_proj                         bpw:  6.00  proxy_err: (OoM)     o  g_sc: 1.895478  [1.48 s]
 -- Quantized: model.layers.1.mlp.experts.0.gate_proj                       bpw:  5.00  proxy_err: 0.000178  o  g_sc: 0.859647  [1.38 s]
Traceback (most recent call last):
  File "C:\Users\blackcat1402\exllamav3\convert.py", line 11, in <module>
    main(_in_args, _job_state)
  File "C:\Users\blackcat1402\AppData\Local\Programs\Python\Python312\Lib\site-packages\torch\utils\_contextlib.py", line 120, in decorate_context
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\conversion\convert_model.py", line 520, in main
    proxy_err = linear.convert_exl3(
                ^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\linear.py", line 299, in convert_exl3
    weight_q, proxy_err, out_tensors = quantize_exl3(
                                       ^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\quant\exl3_lib\quantize.py", line 793, in quantize_exl3
    H, L, su, H_diag = finalize_capture_H(H_data, quant_args, verbose)
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\quant\exl3_lib\quantize.py", line 488, in finalize_capture_H
    L, H = block_ldl(H, 16, verbose)
           ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\quant\exl3_lib\quantize.py", line 291, in block_ldl
    raise e
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\quant\exl3_lib\quantize.py", line 278, in block_ldl
    L = torch.linalg.cholesky(H)
        ^^^^^^^^^^^^^^^^^^^^^^^^
torch._C._LinAlgError: linalg.cholesky: The factorization could not be completed because the input is not positive-definite (the leading minor of order 1 is not positive-definite).

C:\Users\blackcat1402\exllamav3>
C:\Users\blackcat1402\exllamav3>
C:\Users\blackcat1402\exllamav3>
C:\Users\blackcat1402\exllamav3>python convert.py ^
More?   -i C:\Users\blackcat1402\PrimeIntellect-INTELLECT-3-FP16 ^
More?   -o C:\Users\blackcat1402\PrimeIntellect-INTELLECT-3-exl3-5.76bpw ^
More?   -w C:\Users\blackcat1402\exl3_working ^
More?   -b 5.76 ^
More?   -d 1
Detected Windows operating system. Triton does not have an official Windows release, thus FLA will not be adapted for Windows, and any potential errors will not be fixed. Please consider using a Linux environment for compatibility.
 -- Creating new job
    Input directory: C:\Users\blackcat1402\PrimeIntellect-INTELLECT-3-FP16
    Output directory: C:\Users\blackcat1402\PrimeIntellect-INTELLECT-3-exl3-5.76bpw
    Working directory: C:\Users\blackcat1402\exl3_working
    Calibration size: 250 rows, 2048 columns
    Target bitrate: 5.76 (decoder), 6 (head)
    Output scales: auto
    Codebook: mcg
 -- Loaded model config
    Architecture: Glm4MoeForCausalLM
 -- Created model instance:
     - Glm4MoeModel
         - Embedding
         - TransformerBlock
             - RMSNorm
             - Attention
                 - [4x] Linear
             - RMSNorm
             - GatedMLP
                 - [3x] Linear
         - [45x] TransformerBlock
             - RMSNorm
             - Attention
                 - [4x] Linear
             - RMSNorm
             - BlockSparseMLP
                 - [385x] Linear
                 - GatedMLP
                     - [3x] Linear
         - RMSNorm
         - Linear
 -- Loaded tokenizer
    Vocab size: 151367
 -- Preparing input state
 -- Loading unquantized module: model.embed_tokens
 -- Quantized: model.embed_tokens                                           bpw: 16.00  rfn: 0.000000  cos: 0.000000  sqnr: 0.000000  [4.49 s]
 -- Loading unquantized module: model.layers.0
 -- Captured: model.layers.0
 -- Quantized: model.layers.0.self_attn.q_proj                              bpw:  6.00  proxy_err: 0.000025  .  g_sc: 0.405454  [8.48 s]
 -- Quantized: model.layers.0.self_attn.k_proj                              bpw:  7.00  proxy_err: 0.000006  .  g_sc: 0.467450  [1.36 s]
 -- Quantized: model.layers.0.self_attn.v_proj                              bpw:  7.00  proxy_err: 0.000008  .  g_sc: 0.629755  [1.33 s]
 -- Quantized: model.layers.0.self_attn.o_proj                              bpw:  6.00  proxy_err: 0.000024  o  g_sc: 0.691751  [10.13 s]
 -- Quantized: model.layers.0.mlp.up_proj                                   bpw:  5.00  proxy_err: 0.000160  o  g_sc: 0.859647  [8.20 s]
 -- Quantized: model.layers.0.mlp.gate_proj                                 bpw:  5.00  proxy_err: 0.000123  o  g_sc: 0.863102  [8.10 s]
 -- Quantized: model.layers.0.mlp.down_proj                                 bpw:  6.00  proxy_err: 0.000032  o  g_sc: 0.806696  [8.64 s]
 -- Quantized: model.layers.0                                               bpw:  5.67  rfn: 0.005075  cos: 0.000013  sqnr: 46.547755  [78.07 s]
 -- Estimated remaining time: 1 hour, 2 minutes
 -- Loading unquantized module: model.layers.1
 -- Captured: model.layers.1
 !! Warning: block.mlp.0.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.1.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.10.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.100.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.101.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 !! Warning: block.mlp.102.down state has 0 inf values and 720,896,000 NaN values (out of 720,896,000)
 -- Quantized: model.layers.1.self_attn.q_proj                              bpw:  6.00  proxy_err: 0.000017  .  g_sc: 0.376184  [8.21 s]
 -- Quantized: model.layers.1.self_attn.k_proj                              bpw:  7.00  proxy_err: 0.000002  .  g_sc: 0.525990  [1.33 s]
 -- Quantized: model.layers.1.self_attn.v_proj                              bpw:  7.00  proxy_err: 0.000007  .  g_sc: 0.644391  [1.30 s]
 -- Quantized: model.layers.1.self_attn.o_proj                              bpw:  6.00  proxy_err: 0.000031  o  g_sc: 0.786471  [10.21 s]
 -- Quantized: model.layers.1.mlp.experts.0.up_proj                         bpw:  6.00  proxy_err: (OoM)     o  g_sc: 1.895478  [1.46 s]
 -- Quantized: model.layers.1.mlp.experts.0.gate_proj                       bpw:  5.00  proxy_err: 0.000178  o  g_sc: 0.859647  [1.38 s]
Traceback (most recent call last):
  File "C:\Users\blackcat1402\exllamav3\convert.py", line 11, in <module>
    main(_in_args, _job_state)
  File "C:\Users\blackcat1402\AppData\Local\Programs\Python\Python312\Lib\site-packages\torch\utils\_contextlib.py", line 120, in decorate_context
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\conversion\convert_model.py", line 520, in main
    proxy_err = linear.convert_exl3(
                ^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\linear.py", line 299, in convert_exl3
    weight_q, proxy_err, out_tensors = quantize_exl3(
                                       ^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\quant\exl3_lib\quantize.py", line 793, in quantize_exl3
    H, L, su, H_diag = finalize_capture_H(H_data, quant_args, verbose)
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\quant\exl3_lib\quantize.py", line 488, in finalize_capture_H
    L, H = block_ldl(H, 16, verbose)
           ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\quant\exl3_lib\quantize.py", line 291, in block_ldl
    raise e
  File "C:\Users\blackcat1402\exllamav3\exllamav3\modules\quant\exl3_lib\quantize.py", line 278, in block_ldl
    L = torch.linalg.cholesky(H)
        ^^^^^^^^^^^^^^^^^^^^^^^^
torch._C._LinAlgError: linalg.cholesky: The factorization could not be completed because the input is not positive-definite (the leading minor of order 1 is not positive-definite).   

[turboderp]

This is weird. It's quantizing fine here. Can you verify that you downloaded it correctly and there isn't a corrupt .safetensors file somewhere? It could be a Windows issue of course... what hardware and CUDA/driver version are you getting this on?

[blackcat1402]

This is weird. It's quantizing fine here. Can you verify that you downloaded it correctly and there isn't a corrupt .safetensors file somewhere? It could be a Windows issue of course... what hardware and CUDA/driver version are you getting this on?

Thanks for your comments. I will try to re-download the model files first.

[blackcat1402]

This is weird. It's quantizing fine here. Can you verify that you downloaded it correctly and there isn't a corrupt .safetensors file somewhere? It could be a Windows issue of course... what hardware and CUDA/driver version are you getting this on? I am using CUDA 12.8 with driver version as 551.78

[turboderp]

Any update on this? I made a 5.76bpw quant of the same model and it seems to be working fine. Can't quite match the CUDA/driver version, but the difference could be so in many other places. The safetensors library has been especially difficult in the past on Windows.

What GPU are you using? I noticed that it OoMs on up_proj in the second layer, which is non-critical in itself but could suggest it's running out of memory on the first layer too while somehow not catching the error.

[blackcat1402]

Any update on this? I made a 5.76bpw quant of the same model and it seems to be working fine. Can't quite match the CUDA/driver version, but the difference could be so in many other places. The safetensors library has been especially difficult in the past on Windows.

What GPU are you using? I noticed that it OoMs on up_proj in the second layer, which is non-critical in itself but could suggest it's running out of memory on the first layer too while somehow not catching the error.

Sorry for late reply. I solved it by changing my downloader from gopeed to modelscope. The root cause is some of the safetensors were currupted. However, I cannot locate which was currupted. I tried several times and finally went through quantization by a version of downloaded via modelscope downloader. I found this model is very RL style with "over-thinking" like QWQ. Although if it can output code which would be high quality, but output efficiency is rather low. I have no idea how to shorten the thinking process or turn it off completely. BTW, the tool calls of intellect-3 is not as good as orinigal instruct versio of GLM4.5 Air, which inflenced my agentic coding experience as well. Finally, thanks to your reminding of the curruption issue of model files.

[blackcat1402]

What GPU are you using? I noticed that it OoMs on up_proj in the second layer, which is non-critical in itself but could suggest it's running out of memory on the first layer too while somehow not catching the error.

For my prior failure, I noticed that it OoMs. When I tried with another complete download the error positions changed. This helped me identify it was a file corruption error. This is the first time I downloaded currupted files with gopeed. I use it because the download speed is rather fast for my network, which could be faster than 200MB/s and even faster than my hard drive copying.

[blackcat1402]

Any update on this? I made a 5.76bpw quant of the same model and it seems to be working fine. Can't quite match the CUDA/driver version, but the difference could be so in many other places. The safetensors library has been especially difficult in the past on Windows.

What GPU are you using? I noticed that it OoMs on up_proj in the second layer, which is non-critical in itself but could suggest it's running out of memory on the first layer too while somehow not catching the error.

Mon Dec  1 15:01:47 2025
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 551.78                 Driver Version: 551.78         CUDA Version: 12.4     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                     TCC/WDDM  | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA DRIVE-PG199-PROD      TCC   |   00000000:04:00.0 Off |                    0 |
| N/A   31C    P0             44W /  N/A  |   26298MiB /  32768MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   1  NVIDIA DRIVE-PG199-PROD      TCC   |   00000000:05:00.0 Off |                    0 |
| N/A   41C    P0             42W /  N/A  |   20448MiB /  32768MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   2  NVIDIA DRIVE-PG199-PROD      TCC   |   00000000:81:00.0 Off |                    0 |
| N/A   43C    P0             39W /  N/A  |   26298MiB /  32768MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   3  NVIDIA DRIVE-PG199-PROD      TCC   |   00000000:82:00.0 Off |                    0 |
| N/A   31C    P0             46W /  N/A  |   26934MiB /  32768MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    0   N/A  N/A     15548      C   ...rograms\Python\Python312\python.exe      26288MiB |
|    1   N/A  N/A     15548      C   ...rograms\Python\Python312\python.exe      20438MiB |
|    2   N/A  N/A     15548      C   ...rograms\Python\Python312\python.exe      26288MiB |
|    3   N/A  N/A     15548      C   ...rograms\Python\Python312\python.exe      26924MiB |  ```

[blackcat1402]

The GPUs I used is Nvidia Drive PG199 32GB HBM2e, which is more cost effective than RTX3090 with bigger VRAM: https://forums.servethehome.com/index.php?threads/automotive-a100-sxm2-for-fsd-nvidia-drive-a100.43196/

==============NVSMI LOG==============

Timestamp                                 : Mon Dec  1 15:07:05 2025
Driver Version                            : 551.78
CUDA Version                              : 12.4

Attached GPUs                             : 4
GPU 00000000:04:00.0
    Product Name                          : NVIDIA DRIVE-PG199-PROD
    Product Brand                         : NVIDIA
    Product Architecture                  : Ampere
    Display Mode                          : Disabled
    Display Active                        : Disabled
    Persistence Mode                      : N/A
    Addressing Mode                       : N/A
    MIG Mode
        Current                           : Disabled
        Pending                           : Disabled
    Accounting Mode                       : Disabled
    Accounting Mode Buffer Size           : 4000
    Driver Model
        Current                           : TCC
        Pending                           : TCC
    Serial Number                         : 1655322003424
    GPU UUID                              : GPU-304c0c55-3965-e98f-8f90-33caf67136a4
    Minor Number                          : N/A
    VBIOS Version                         : 92.00.A0.00.01
    MultiGPU Board                        : No
    Board ID                              : 0x400
    Board Part Number                     : 900-6G199-0000-200
    GPU Part Number                       : 20BB-550-A1
    FRU Part Number                       : N/A
    Module ID                             : 1
    Inforom Version
        Image Version                     : G199.0600.00.02
        OEM Object                        : 2.0
        ECC Object                        : 6.16
        Power Management Object           : N/A
    Inforom BBX Object Flush
        Latest Timestamp                  : N/A
        Latest Duration                   : N/A
    GPU Operation Mode
        Current                           : N/A
        Pending                           : N/A
    GPU C2C Mode                          : N/A
    GPU Virtualization Mode
        Virtualization Mode               : None
        Host VGPU Mode                    : N/A
        vGPU Heterogeneous Mode           : N/A
    GPU Reset Status
        Reset Required                    : No
        Drain and Reset Recommended       : No
    GSP Firmware Version                  : N/A
    IBMNPU
        Relaxed Ordering Mode             : N/A
    PCI
        Bus                               : 0x04
        Device                            : 0x00
        Domain                            : 0x0000
        Device Id                         : 0x20BB10DE
        Bus Id                            : 00000000:04:00.0
        Sub System Id                     : 0x14A110DE
        GPU Link Info
            PCIe Generation
                Max                       : 3
                Current                   : 3
                Device Current            : 3
                Device Max                : 4
                Host Max                  : 3
            Link Width
                Max                       : 16x
                Current                   : 8x
        Bridge Chip
            Type                          : N/A
            Firmware                      : N/A
        Replays Since Reset               : 0
        Replay Number Rollovers           : 0
        Tx Throughput                     : 0 KB/s
        Rx Throughput                     : 0 KB/s
        Atomic Caps Inbound               : N/A
        Atomic Caps Outbound              : N/A
    Fan Speed                             : N/A
    Performance State                     : P0
    Clocks Event Reasons
        Idle                              : Not Active
        Applications Clocks Setting       : Not Active
        SW Power Cap                      : Not Active
        HW Slowdown                       : Not Active
            HW Thermal Slowdown           : Not Active
            HW Power Brake Slowdown       : Not Active
        Sync Boost                        : Not Active
        SW Thermal Slowdown               : Not Active
        Display Clock Setting             : Not Active
    Sparse Operation Mode                 : N/A
    FB Memory Usage
        Total                             : 32768 MiB
        Reserved                          : 270 MiB
        Used                              : 26298 MiB
        Free                              : 6199 MiB
    BAR1 Memory Usage
        Total                             : 16384 MiB
        Used                              : 4 MiB
        Free                              : 16380 MiB
    Conf Compute Protected Memory Usage
        Total                             : N/A
        Used                              : N/A
        Free                              : N/A
    Compute Mode                          : Default
    Utilization
        Gpu                               : 0 %
        Memory                            : 0 %
        Encoder                           : 0 %
        Decoder                           : 0 %
        JPEG                              : 0 %
        OFA                               : 0 %
    Encoder Stats
        Active Sessions                   : 0
        Average FPS                       : 0
        Average Latency                   : 0
    FBC Stats
        Active Sessions                   : 0
        Average FPS                       : 0
        Average Latency                   : 0
    ECC Mode
        Current                           : Enabled
        Pending                           : Enabled
    ECC Errors
        Volatile
            SRAM Correctable              : 0
            SRAM Uncorrectable Parity     : 0
            SRAM Uncorrectable SEC-DED    : 0
            DRAM Correctable              : 0
            DRAM Uncorrectable            : 0
        Aggregate
            SRAM Correctable              : 0
            SRAM Uncorrectable Parity     : 0
            SRAM Uncorrectable SEC-DED    : 0
            DRAM Correctable              : 0
            DRAM Uncorrectable            : 0
            SRAM Threshold Exceeded       : No
        Aggregate Uncorrectable SRAM Sources
            SRAM L2                       : 0
            SRAM SM                       : 0
            SRAM Microcontroller          : 0
            SRAM PCIE                     : 0
            SRAM Other                    : 0
    Retired Pages
        Single Bit ECC                    : N/A
        Double Bit ECC                    : N/A
        Pending Page Blacklist            : N/A
    Remapped Rows
        Correctable Error                 : 0
        Uncorrectable Error               : 0
        Pending                           : No
        Remapping Failure Occurred        : No
        Bank Remap Availability Histogram
            Max                           : 512 bank(s)
            High                          : 0 bank(s)
            Partial                       : 0 bank(s)
            Low                           : 0 bank(s)
            None                          : 0 bank(s)
    Temperature
        GPU Current Temp                  : 30 C
        GPU T.Limit Temp                  : N/A
        GPU Shutdown Temp                 : 110 C
        GPU Slowdown Temp                 : N/A
        GPU Max Operating Temp            : N/A
        GPU Target Temperature            : N/A
        Memory Current Temp               : 42 C
        Memory Max Operating Temp         : N/A
    GPU Power Readings
        Power Draw                        : 44.07 W
        Current Power Limit               : N/A
        Requested Power Limit             : N/A
        Default Power Limit               : N/A
        Min Power Limit                   : N/A
        Max Power Limit                   : N/A
    GPU Memory Power Readings
        Power Draw                        : N/A
    Module Power Readings
        Power Draw                        : N/A
        Current Power Limit               : N/A
        Requested Power Limit             : N/A
        Default Power Limit               : N/A
        Min Power Limit                   : N/A
        Max Power Limit                   : N/A
    Clocks
        Graphics                          : 1260 MHz
        SM                                : 1260 MHz
        Memory                            : 1404 MHz
        Video                             : 1140 MHz
    Applications Clocks
        Graphics                          : 1260 MHz
        Memory                            : 1404 MHz
    Default Applications Clocks
        Graphics                          : 1260 MHz
        Memory                            : 1404 MHz
    Deferred Clocks
        Memory                            : N/A
    Max Clocks
        Graphics                          : 1275 MHz
        SM                                : 1275 MHz
        Memory                            : 1404 MHz
        Video                             : 1155 MHz
    Max Customer Boost Clocks
        Graphics                          : 1260 MHz
    Clock Policy
        Auto Boost                        : N/A
        Auto Boost Default                : N/A
    Voltage
        Graphics                          : 825.000 mV
    Fabric
        State                             : N/A
        Status                            : N/A
        CliqueId                          : N/A
        ClusterUUID                       : N/A
        Health
            Bandwidth                     : N/A
    Processes
        GPU instance ID                   : N/A
        Compute instance ID               : N/A
        Process ID                        : 15548
            Type                          : C
            Name                          : C:\Users\blackcat1402\AppData\Local\Programs\Python\Python312\python.exe
            Used GPU Memory               : 26288 MiB 

[blackcat1402]

Intellect-3 5.0bpw catbench result: