Hi

as you can see at this forum:

https://boinc.berkeley.edu/dev/forum_thread.php?id=11891#82080

I'm trying to run ati gpu boinc task with HD4650.

In my last post:

With the AMD Catalyst 12.1 drivers (AMD SDK 2.6) looks like all is working fine.



But the registry logs shows that computation isn't running fine.



Any idea

---

I'm about sure it is a (lack of) memory problem, also because your clinfo/GPU-Z output (posted in this post says Global Memory Size 256 MB.
Can you download https://boinc.berkeley.edu/dl/clinfo.zip, unzip it and run the clinfo.exe in it for more information?

To capture the output of clinfo.exe do this:
Save clinfo.exe in C:\temp\
Open a command line prompt (Start->All Programs->Accessories->Command Prompt)
In it type clinfo.exe >> c:\temp\clinfo.txt and hit Enter. (yes, that's two greater than characters)
When it's done and the cursor is back, navigate to C:\temp and open clinfo.txt with Notepad, CTRL+A, CTRL+C
Post that info here, CTRL+V

---

Hi

Here is the report of clinfo.exe:

Number of platforms: 1
Platform Profile: FULL_PROFILE
Platform Version: OpenCL 1.1 AMD-APP (851.4)
Platform Name: AMD Accelerated Parallel Processing
Platform Vendor: Advanced Micro Devices, Inc.
Platform Extensions: cl_khr_icd cl_amd_event_callback cl_amd_offline_devices


Platform Name: AMD Accelerated Parallel Processing
Number of devices: 2
Device Type: CL_DEVICE_TYPE_GPU
Device ID: 4098
Max compute units: 8
Max work items dimensions: 3
Max work items[0]: 128
Max work items[1]: 128
Max work items[2]: 128
Max work group size: 128
Preferred vector width char: 16
Preferred vector width short: 8
Preferred vector width int: 4
Preferred vector width long: 2
Preferred vector width float: 4
Preferred vector width double: 0
Max clock frequency: 600Mhz
Address bits: 32
Max memory allocation: 268435456
Image support: No
Max size of kernel argument: 1024
Alignment (bits) of base address: 2048
Minimum alignment (bytes) for any datatype: 128
Single precision floating point capability
Denorms: No
Quiet NaNs: Yes
Round to nearest even: Yes
Round to zero: Yes
Round to +ve and infinity: Yes
IEEE754-2008 fused multiply-add: Yes
Cache type: None
Cache line size: 0
Cache size: 0
Global memory size: 268435456
Constant buffer size: 65536
Max number of constant args: 8
Local memory type: Global
Local memory size: 16384
Error correction support: 0
Profiling timer resolution: 1
Device endianess: Little
Available: Yes
Compiler available: Yes
Execution capabilities:
Execute OpenCL kernels: Yes
Execute native function: No
Queue properties:
Out-of-Order: No
Profiling : Yes
Platform ID: 0123A4F4
Name: ATI RV730
Vendor: Advanced Micro Devices, Inc.
Driver version: CAL 1.4.1664
Profile: FULL_PROFILE
Version: OpenCL 1.0 AMD-APP (851.4)
Extensions: cl_khr_gl_sharing cl_amd_device_attribute_query


Device Type: CL_DEVICE_TYPE_CPU
Device ID: 4098
Max compute units: 1
Max work items dimensions: 3
Max work items[0]: 1024
Max work items[1]: 1024
Max work items[2]: 1024
Max work group size: 1024
Preferred vector width char: 16
Preferred vector width short: 8
Preferred vector width int: 4
Preferred vector width long: 2
Preferred vector width float: 4
Preferred vector width double: 0
Max clock frequency: 2398Mhz
Address bits: 32
Max memory allocation: 536330240
Image support: Yes
Max number of images read arguments: 128
Max number of images write arguments: 8
Max image 2D width: 8192
Max image 2D height: 8192
Max image 3D width: 2048
Max image 3D height: 2048
Max image 3D depth: 2048
Max samplers within kernel: 16
Max size of kernel argument: 4096
Alignment (bits) of base address: 1024
Minimum alignment (bytes) for any datatype: 128
Single precision floating point capability
Denorms: Yes
Quiet NaNs: Yes
Round to nearest even: Yes
Round to zero: Yes
Round to +ve and infinity: Yes
IEEE754-2008 fused multiply-add: Yes
Cache type: Read/Write
Cache line size: 0
Cache size: 0
Global memory size: 536330240
Constant buffer size: 65536
Max number of constant args: 8
Local memory type: Global
Local memory size: 32768
Error correction support: 0
Profiling timer resolution: 279
Device endianess: Little
Available: Yes
Compiler available: Yes
Execution capabilities:
Execute OpenCL kernels: Yes
Execute native function: Yes
Queue properties:
Out-of-Order: No
Profiling : Yes
Platform ID: 0123A4F4
Name: Intel(R) Pentium(R) 4 CPU 2.40GHz
Vendor: GenuineIntel
Driver version: 2.0
Profile: FULL_PROFILE
Version: OpenCL 1.1 AMD-APP (851.4)
Extensions: cl_khr_fp64 cl_amd_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_byte_addressable_store cl_khr_gl_sharing cl_ext_device_fission cl_amd_device_attribute_query cl_amd_vec3 cl_amd_printf cl_amd_media_ops cl_amd_popcnt

Max memory allocation 268435456

That says 256MB as well (268435456 (bytes), divided by 1024 (for kilobytes), divided by 1024(for megabytes), equals 256 (megabytes)). Does your GPU use shared memory from main memory, that BOINC reports it as 512MB?
But 256MB is at least not enough memory to run Breakthrough Listen tasks in on the GPU. Hence the postpone message. (And since you're using an earlier BOINC with less informative messages, because the newest BOINC wouldn't even detect your GPU's OpenCL capability, you don't know what the postpone message means)

---

No. My GPU doesn't use shared memory. I purchased It as 512MB RAM graphic card.

Yes, I'm running last BOINC version that detects it as opencl capable.

I tested some project.

Amicable Numbers
Enigma@home
LHC@home
Moo! Wraper
Seti@home

And only Moo! Wraper GPU tasks are being downloaded and running without any problems.

Thanks for all.

---

in the early 2016, i was using a HD4350 with Win XP 32bits and catalyst 11.12 ,it works well for Wu crunching but takes more than 9 hours to finish one ;)

---

Information from the readme on Lunatics applications, and knowing that the stock GPU applications are essentially the same applications as the ones in the Lunatics 0.44 Installer:

The ATI MB application will not work on ATI cards with workgroup size 128 (e.g. HD43xx). To check workgroupsize run CLinfo, downloadable (at present)
from {http://boinc.berkeley.edu/dl/clinfo.zip}

and

We offer the application in two different variants, depending on card:
plain - for most HD47xx/HD48xx and those with HD5xxx, HD6xxx and HD7xxx cards who suffer from driver incompatibilities.
HD5 - for most if not all owners of HD5xxx GPUs, HD6xxx and HD7xxx GPUs.

So minimum version these days is HD47xx, not anything below.

---