Error running beagle_GPU Transposed

Atila Iamarino

unread,

Mar 14, 2012, 4:32:34 PM3/14/12

to beagl...@googlegroups.com

Dear Rambaut,

Here is the output of 'nvidiasmi -q':

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

Timestamp : Wed Mar 14 17:28:39 2012

Driver Version : 295.20

Attached GPUs : 1

GPU 0000:01:00.0

Product Name : GeForce GTX 480

Display Mode : N/A

Persistence Mode : Disabled

Driver Model

Current : N/A

Pending : N/A

Serial Number : N/A

GPU UUID : N/A

VBIOS Version : 70.00.35.00.00

Inforom Version

OEM Object : N/A

ECC Object : N/A

Power Management Object : N/A

PCI

Bus : 0x01

Device : 0x00

Domain : 0x0000

Device Id : 0x06C010DE

Bus Id : 0000:01:00.0

Sub System Id : 0x22801462

GPU Link Info

PCIe Generation

Max : N/A

Current : N/A

Link Width

Max : N/A

Current : N/A

Fan Speed : 40 %

Performance State : N/A

Memory Usage

Total : 1535 MB

Used : 276 MB

Free : 1259 MB

Compute Mode : Default

Utilization

Gpu : N/A

Memory : N/A

Ecc Mode

Current : N/A

Pending : N/A

ECC Errors

Volatile

Single Bit

Device Memory : N/A

Register File : N/A

L1 Cache : N/A

L2 Cache : N/A

Total : N/A

Double Bit

Device Memory : N/A

Register File : N/A

L1 Cache : N/A

L2 Cache : N/A

Total : N/A

Aggregate

Single Bit

Device Memory : N/A

Register File : N/A

L1 Cache : N/A

L2 Cache : N/A

Total : N/A

Double Bit

Device Memory : N/A

Register File : N/A

L1 Cache : N/A

L2 Cache : N/A

Total : N/A

Temperature

Gpu : 35 C

Power Readings

Power Management : N/A

Power Draw : N/A

Power Limit : N/A

Clocks

Graphics : N/A

SM : N/A

Memory : N/A

Max Clocks

Graphics : N/A

SM : N/A

Memory : N/A

Compute Processes : Not Supported

Thank you,

Atila

Em quarta-feira, 14 de março de 2012 06h08min18s UTC-3, rambaut escreveu:

Dear Atila,
This sounds like one of two things - either an issue with the CUDA driver you have installed or a physical problem with your GPU board. If you have installed CUDA tools, try running 'nvidia-smi -q' to get a detailed probe of your GPU board.
Andrew
On 13 Mar 2012, at 15:09, Atila Iamarino wrote:
> Dear Rambaut,
>
> Yes, it works with -beagle_cpu and -beagle_sse. Also, my GPU is recognized, but I checked BEAGLE installation and it fails the tests. I haev forwarded the error message to beagle-dev group.
>
> Thank you,
> Atila
>
> Em terça-feira, 13 de março de 2012 07h43min51s UTC-3, rambaut escreveu:
> Dear Atila,
> To answer your last question, it looks like it is a problem in BEAGLE. Firstly, does it work if you use -beagle_cpu or -beagle_sse? If so, it may be an issue with CUDA. Have you installed the latest CUDA drivers? If you run the 'nvidia-smi' utility does it recognize your GPU?
>
> Andrew
>
> On 12 Mar 2012, at 21:34, Atila wrote:
>
> > Hi,
> >
> > I'm trying to run BEAST with CUDA. I have tried ../../bin/beast -
> > beagle_GPU -overwrite benchmark2.xml both with BEAST1.6.2 and 1.7.0
> > and get the following error:
> >
> >
> > Using BEAGLE TreeLikelihood
> > Branch rate model used: strictClockBranchRates
> > Using BEAGLE resource 1: GeForce GTX 480
> > Global memory (MB): 1536
> > Clock speed (Ghz): 1.50
> > Number of cores: 480
> > with instance flags: PRECISION_DOUBLE COMPUTATION_SYNCH
> > EIGEN_REAL SCALING_MANUAL SCALERS_RAW VECTOR_NONE THREADING_NONE
> > PROCESSOR_GPU
> > Ignoring ambiguities in tree likelihood.
> > With 5565 unique site patterns.
> > Using rescaling scheme : dynamic
> > Creating the MCMC chain:
> > chainLength=100000
> > autoOptimize=true
> > autoOptimize delayed for 1000 steps
> > full evaluation test off
> > #
> > # A fatal error has been detected by the Java Runtime Environment:
> > #
> > # SIGSEGV (0xb) at pc=0x00007fb9f020d83d, pid=23711,
> > tid=140436735981312
> > #
> > # JRE version: 7.0_03-b04
> > # Java VM: Java HotSpot(TM) 64-Bit Server VM (22.1-b02 mixed mode
> > linux-amd64 compressed oops)
> > # Problematic frame:
> > # C [libhmsbeagle-cuda.so.1+0x1283d]
> > beagle::gpu::BeagleGPUImpl<double>::updateTransitionMatrices(int, int
> > const*, int const*, int const*, double const*, int)+0xed
> > #
> > # Failed to write core dump. Core dumps have been disabled. To enable
> > core dumping, try "ulimit -c unlimited" before starting Java again
> > #
> > # An error report file with more information is saved as:
> > # /home/atila/applications/BEASTv1.7.pre/examples/Benchmarks/
> > hs_err_pid23711.log
> > #
> > # If you would like to submit a bug report, please visit:
> > # http://bugreport.sun.com/bugreport/crash.jsp
> >
> > I tried to use open-jdk and the error only changes to
> >
> > CUDA error: "Invalid value" from file <GPUInterfaceCUDA.cpp>, line
> > 485.
> >
> > and then:
> >
> > # A fatal error has been detected by the Java Runtime Environment:
> > #
> > # SIGSEGV (0xb) at pc=0x00007f5871d38b50, pid=23950,
> > tid=140018069825280
> > #
> > # JRE version: 6.0_23-b23
> > # Java VM: OpenJDK 64-Bit Server VM (20.0-b11 mixed mode linux-amd64
> > compressed oops)
> > # Derivative: IcedTea6 1.11pre
> > # Distribution: Ubuntu 11.10, package 6b23~pre11-0ubuntu1.11.10.2
> > # Problematic frame:
> > # C [libhmsbeagle-cuda.so.1+0x11b50]
> > beagle::gpu::BeagleGPUImpl<float>::removeScaleFactors(int const*, int,
> > int)+0x30
> >
> > Is this a problem from BEAST ou beagle-lib?
> >
> > Thank you,
> > Atila Iamarino
> >
> > --
> > You received this message because you are subscribed to the Google Groups "beast-users" group.
> > To post to this group, send email to beast...@googlegroups.com.
> > To unsubscribe from this group, send email to beast-users+unsubscribe@googlegroups.com.
> > For more options, visit this group at http://groups.google.com/group/beast-users?hl=en.
> >
>
>
>
>
>
>
>
> --
> You received this message because you are subscribed to the Google Groups "beast-users" group.
> To view this discussion on the web visit https://groups.google.com/d/msg/beast-users/-/IvmQlmZ7qxEJ.
> To post to this group, send email to beast...@googlegroups.com.
> To unsubscribe from this group, send email to beast-users+unsubscribe@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/beast-users?hl=en.
___________________________________________________________________
Andrew Rambaut
Institute of Evolutionary Biology University of Edinburgh
Ashworth Laboratories Edinburgh EH9 3JT
EMAIL - a.ra...@ed.ac.uk TEL - +44 131 6508624
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

Atila Iamarino

unread,

Mar 15, 2012, 2:06:58 PM3/15/12

to beagl...@googlegroups.com

I am running memtestG80 to see if it is a cuda ou GPU issue, and it seems fine:

Test iteration 54 (GPU 0, 1024 MiB): 0 errors so far

Moving Inversions (ones and zeros): 0 errors (36 ms)

Memtest86 Walking 8-bit: 0 errors (283 ms)

True Walking zeros (8-bit): 0 errors (139 ms)

True Walking ones (8-bit): 0 errors (140 ms)

Moving Inversions (random): 0 errors (36 ms)

Memtest86 Walking zeros (32-bit): 0 errors (558 ms)

Memtest86 Walking ones (32-bit): 0 errors (557 ms)

Random blocks: 0 errors (271 ms)

Memtest86 Modulo-20: 0 errors (1830 ms)

Logic (one iteration): 0 errors (21 ms)

Logic (4 iterations): 0 errors (38 ms)

Logic (shared memory, one iteration): 0 errors (24 ms)

Logic (shared-memory, 4 iterations): 0 errors (52 ms)

Atila Iamarino

unread,

Mar 15, 2012, 4:27:27 PM3/15/12

to beagl...@googlegroups.com

Ok, after many troubles to re-install NVIDIA drivers, dev-drivers, toolkit in Ubuntu 11.10 I got the SDK to work. Yet, beagle-lib fails 'make check'

./genomictest.sh: line 2: 14892 Segmentation fault ./genomictest --states 64 --sites 100 --taxa 10

Here is the output of NVIDIA_GPU_Computing_SDK/C/bin/linux/release/deviceQuery:

Starting...

CUDA Device Query (Runtime API) version (CUDART static linking)

Found 1 CUDA Capable device(s)

Device 0: "GeForce GTX 480"

CUDA Driver Version / Runtime Version 4.1 / 4.1

CUDA Capability Major/Minor version number: 2.0

Total amount of global memory: 1536 MBytes (1610153984 bytes)

(15) Multiprocessors x (32) CUDA Cores/MP: 480 CUDA Cores

GPU Clock Speed: 1.50 GHz

Memory Clock rate: 2000.00 Mhz

Memory Bus Width: 384-bit

L2 Cache Size: 786432 bytes

Max Texture Dimension Size (x,y,z) 1D=(65536), 2D=(65536,65535), 3D=(2048,2048,2048)

Max Layered Texture Size (dim) x layers 1D=(16384) x 2048, 2D=(16384,16384) x 2048

Total amount of constant memory: 65536 bytes

Total amount of shared memory per block: 49152 bytes

Total number of registers available per block: 32768

Warp size: 32

Maximum number of threads per block: 1024

Maximum sizes of each dimension of a block: 1024 x 1024 x 64

Maximum sizes of each dimension of a grid: 65535 x 65535 x 65535

Maximum memory pitch: 2147483647 bytes

Texture alignment: 512 bytes

Concurrent copy and execution: Yes with 1 copy engine(s)

Run time limit on kernels: Yes

Integrated GPU sharing Host Memory: No

Support host page-locked memory mapping: Yes

Concurrent kernel execution: Yes

Alignment requirement for Surfaces: Yes

Device has ECC support enabled: No

Device is using TCC driver mode: No

Device supports Unified Addressing (UVA): Yes

Device PCI Bus ID / PCI location ID: 1 / 0

Compute Mode:

< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 4.1, CUDA Runtime Version = 4.1, NumDevs = 1, Device = GeForce GTX 480

[deviceQuery] test results...

PASSED

Andrew Rambaut

unread,

Mar 16, 2012, 3:29:18 AM3/16/12

to beagl...@googlegroups.com

Hi Atila,

Do the other programs in the SDK work? Try a simple one ilke 'matrixMul' or 'matrixMulDrv'.

Andrew

Reply all

Reply to author

Forward