Clearly one thread with "caw-caw-caw, CAW! I [STRIKE]laid an egg[/STRIKE] found a factor! Everyone, look!" vanity factors was not enough. Let's do that in every thread. "My factor is bigger than your factor!" -- "I've got more credit" -- "I've got a faster rate of getting credit" :edit:

...Kindergarten, kindergarten, everywhere!

[edit: my reply was for Kracker, didn't see Batalov's crosspost]

Well, hehe, I still consider my record being the 103 bits which I got it by "honest hard work"...

OTOH, the situation today is like:
kracker's aliens: 312 G. with 9 factors
chalsall's aliens: 652 G. with 8 factors
total: 964 G.
times 8: 7712 G.
me (TF): 5950 G. with 6 factors

difference: 1762 G.

This has to be recovered over the weekend.. :razz:
The other potential contributors retracted their computers before having time to report any work. No big deal, I may have some difficulties right now to keep the pace... I also added one more CPU (i7-2600k) to P-1 marathon. Technically you can retreat your computers after the weekend, my "top" is safe now. I will continue to TF for a while, I am struggling right now to add a third card here (the 7970), I put it into an old rag, but having trouble with opencl.dll and windows xp sp2 (!) I am already wasting more than 4 hours since I came from job, to "upgrade" this freaking winxp32 to something more recent. Misfit also doesn't work, because the dotnet is old, and I didn't make it so fast to update dotnet yet.

312G? That's all? Damn, going to step it up.

BTW, what is "xp"? experience points? :razz:

Up and running, I had to downgrade all the way down the AMD drivers, and the APP SDK from 2.9, 2.8, 2.7, then 2.5 :smile: The idiots cut the winxp support for opencl.dll in sdk 2.6, and luckily I found some info on bitcoin forums... I had no idea why mfakto says that opencl.dll is not found, when I put it there with my own hand, then, after convincing it to "see" the files, he started shouting at me that:

[CODE]Microsoft Windows XP [Version 5.1.2600]
(C) Copyright 1985-2001 Microsoft Corp.

_\prime\mfakto\MFO_01>mfakto_x32
mfakto 0.13-Win (32bit build)

Runtime options
Inifile mfakto.ini
Verbosity 1
SieveOnGPU yes
GPUSievePrimes 82486
GPUSieveSize 64Mi bits
GPUSieveProcessSize 16Ki bits
WorkFile worktodo.txt
ResultsFile results.txt
Checkpoints enabled
CheckpointDelay 300s
Stages enabled
StopAfterFactor class
PrintMode compact
V5UserID none
ComputerID none
TimeStampInResults no
VectorSize 2
GPUType GCN
SmallExp no
Compiletime options
MORE_CLASSES enabled
Select device - Error -1001 (Unkown errorcode): clGetPlatformsIDs(num)
init_CL(1, 0) failed

_\prime\mfakto\MFO_01>[/CODE]Pretty freaking confusing... With SDK 2.5 it worked... More confusing was that [URL="http://developer.amd.com/tools-and-sdks/heterogeneous-computing/amd-accelerated-parallel-processing-app-sdk/downloads/download-archive/"]here[/URL], there is no 2.5 anymore, and none to be found on their pages. Some clever bitminers made copies :D

You could just cut all the confusion and go Linux.

AMD cut the ropes on XP a while ago, so did nVidia from their main fork of drivers.

Grrr, didn't know that, and anyhow I talked too soon...

[CODE]Microsoft Windows XP [Version 5.1.2600]
(C) Copyright 1985-2001 Microsoft Corp.

_\prime\mfakto\MFO_01>mfakto_x32
mfakto 0.13-Win (32bit build)

Runtime options
Inifile mfakto.ini
Verbosity 1
SieveOnGPU yes
GPUSievePrimes 82486
GPUSieveSize 64Mi bits
GPUSieveProcessSize 16Ki bits
WorkFile worktodo.txt
ResultsFile results.txt
Checkpoints enabled
CheckpointDelay 300s
Stages enabled
StopAfterFactor class
PrintMode compact
V5UserID none
ComputerID none
TimeStampInResults no
VectorSize 2
GPUType GCN
SmallExp no
Compiletime options
MORE_CLASSES enabled
Select device - GPU not found, fallback to CPU.
Get device info - Compiling kernels.

BUILD OUTPUT
.\common.cl(31): warning: unrecognized #pragma
#pragma "Enabling printf"
^

.\common.cl(41): warning: unrecognized #pragma
#pragma "Replacing popcount"
^

.\common.cl(91): warning: unrecognized #pragma
#pragma "Emulating amd_max3"
^

.\gpusieve.cl(95): warning: integer conversion resulted in a change of sign
1<<16, 1<<17, 1<<18, 1<<19, 1<<20, 1<<21, 1<
<22, 1<<23, 1<<24, 1<<25, 1<<26, 1<<27, 1<<28, 1<<29, 1<<30, 1<<31,

^

.\gpusieve.cl(249): warning: unknown attribute "work_group_size_hint"
__kernel void __attribute__((work_group_size_hint(256, 1, 1))) SegSieve (__glo
bal uchar *big_bit_array_dev, __global uchar *pinfo_dev, uint maxp)
^

.\gpusieve.cl(287): warning: integer conversion resulted in a change of sign
uint mask, mask2, mask3, mask4, i11=(-1), i13, i17, i19, i23, i29, i31
, i37, i41, i43, i47, i53, i59, i61;
^

.\gpusieve.cl(1323): warning: unknown attribute "work_group_size_hint"
__kernel void __attribute__((work_group_size_hint(256, 1, 1))) CalcModularInve
rses (uint exponent, __global int *calc_info)
^

.\gpusieve.cl(1364): warning: unknown attribute "work_group_size_hint"
__kernel void __attribute__((work_group_size_hint(256, 1, 1))) CalcBitToClear
(uint exponent, ulong k_base, __global int *calc_info, __global uchar *pinfo_dev
)
^

.\barrett15.cl(2938): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d0 = mad24(NUM_CLASSES, k_delta & 0x7FFF, k_base.d0); //
k_delta can exceed 2^24
^

.\barrett15.cl(2939): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d1 = (k.d0 >> 15) + mad24(NUM_CLASSES, k_delta >> 15, k_base.d1); // k
is limited to 2^64 -1
^

.\barrett15.cl(3063): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d0 = mad24(NUM_CLASSES, k_delta & 0x7FFF, k_base.d0); //
k_delta can exceed 2^24
^

.\barrett15.cl(3064): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d1 = (k.d0 >> 15) + mad24(NUM_CLASSES, k_delta >> 15, k_base.d1); // k
is limited to 2^64 -1
^

.\barrett15.cl(3188): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d0 = mad24(NUM_CLASSES, k_delta & 0x7FFF, k_base.d0); //
k_delta can exceed 2^24
^

.\barrett15.cl(3189): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d1 = (k.d0 >> 15) + mad24(NUM_CLASSES, k_delta >> 15, k_base.d1); // k
is limited to 2^64 -1
^

.\barrett15.cl(3313): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d0 = mad24(NUM_CLASSES, k_delta & 0x7FFF, k_base.d0); //
k_delta can exceed 2^24
^

.\barrett15.cl(3314): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d1 = (k.d0 >> 15) + mad24(NUM_CLASSES, k_delta >> 15, k_base.d1); // k
is limited to 2^64 -1
^

.\barrett15.cl(3448): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d0 = mad24(NUM_CLASSES, k_delta & 0x7FFF, k_base.d0); //
k_delta can exceed 2^24
^

.\barrett15.cl(3449): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d1 = (k.d0 >> 15) + mad24(NUM_CLASSES, k_delta >> 15, k_base.d1); // k
is limited to 2^64 -1
^

.\barrett15.cl(3577): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d0 = mad24(NUM_CLASSES, k_delta & 0x7FFF, k_base.d0); //
k_delta can exceed 2^24
^

.\barrett15.cl(3578): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d1 = (k.d0 >> 15) + mad24(NUM_CLASSES, k_delta >> 15, k_base.d1); // k
is limited to 2^64 -1
^

.\barrett15.cl(3706): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d0 = mad24(NUM_CLASSES, k_delta & 0x7FFF, k_base.d0); //
k_delta can exceed 2^24
^

.\barrett15.cl(3707): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
k.d1 = (k.d0 >> 15) + mad24(NUM_CLASSES, k_delta >> 15, k_base.d1); // k
is limited to 2^64 -1
^

.\barrett.cl(1876): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
my_k_base.d1 = k_base.d1 + mul_hi(NUM_CLASSES, k_delta) - AS_UINT_V(k_base
.d0 > my_k_base.d0); /* k is limited to 2^64 -1 so there is no need for k.d2
*/
^

.\barrett.cl(2000): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
my_k_base.d1 = k_base.d1 + mul_hi(NUM_CLASSES, k_delta) - AS_UINT_V(k_base
.d0 > my_k_base.d0); /* k is limited to 2^64 -1 so there is no need for k.d2
*/
^

.\barrett.cl(2124): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
my_k_base.d1 = k_base.d1 + mul_hi(NUM_CLASSES, k_delta) - AS_UINT_V(k_base
.d0 > my_k_base.d0); /* k is limited to 2^64 -1 so there is no need for k.d2
*/
^

.\barrett.cl(2248): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
my_k_base.d1 = k_base.d1 + mul_hi(NUM_CLASSES, k_delta) - AS_UINT_V(k_base
.d0 > my_k_base.d0); /* k is limited to 2^64 -1 so there is no need for k.d2
*/
^

.\barrett.cl(2372): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
my_k_base.d1 = k_base.d1 + mul_hi(NUM_CLASSES, k_delta) - AS_UINT_V(k_base
.d0 > my_k_base.d0); /* k is limited to 2^64 -1 so there is no need for k.d2
*/
^

.\barrett.cl(2496): error: bad argument type to opencl builtin function:
expected type "uint", actual type "uint2"
my_k_base.d1 = k_base.d1 + mul_hi(NUM_CLASSES, k_delta) - AS_UINT_V(k_base
.d0 > my_k_base.d0); /* k is limited to 2^64 -1 so there is no need for k.d2
*/
^

20 errors detected in the compilation of "C:\DOCUME~1\tati\LOCALS~1\Temp\OCL6.tm
p.cl".

END OF BUILD OUTPUT
Error -11 (Build program failure): clBuildProgram
init_CL(1, 0) failed

_\prime\mfakto\MFO_01>[/CODE]

I thought this is from the driver, but I reinstalled back the last driver (after the SDK adventures) and still doesn't run. What I am missing? I may look quite stupid here, but now at 2 o'clock in the night, I really am stuck and have no idea! I go to sleep...

For mfakto, you do not need the APP SDK for 13.* drivers. What catalyst driver do you have now? 13.1 seems to support xp.

[URL="http://www.guru3d.com/files_details/amd_catalyst_13_1_legacy_download.html"]http://www.guru3d.com/files_details/amd_catalyst_13_1_legacy_download.html[/URL]

EDIT: I wouldn't run XP though... as I said earlier linux should work fine

[QUOTE=kjaget;361983]I do my best to be unremarkable at everything I attempt.[/QUOTE]

Remarkable

See attached.

I never know if I'm reading these things correctly and if I am reading enough of them to get the full picture. Let's take a crack at it.


During the period of December 8 to December 15, in the 60M range, 1923 exponents were taken from 71 to something, and 2359 exponents were taken from something to 74. In particular, 1412 exponents were taken to 74 in the 64M range.

194 factors were found in the 60M range, and 1017 LL's were completed in the 60M range.


From this it would seem factoring is massively ahead of LL. There have been varying opinions on this statement. Was this just a weird week or is there a third graph somewhere I need to look at before drawing any conclusions?

Thanks

You are making one logical mistake, you forget about the lower ranges and growth of GIMPS.
Any one completing an assignment in a range below 50M needs a new assignment from the 60-69M range so you need to look at the whole project, not only the 60-69M range. New members also gets assignments without clearing any old ones. Same about people switching from DC and other work to LL.

There were 2127 LL assignment cleared between 30M and 100M, only 1214 were in the 60M-70M range. This is still lacking growth and switched assignments but it hints about a decent lead again with 2359 factors LLTF to 74 in the same period, about a 10% lead.

/Göran

[QUOTE=Axelsson;362079]This is still lacking growth and switched assignments but it hints about a decent lead again with 2359 factors LLTF to 74 in the same period, about a 10% lead.[/QUOTE]

Göran has it (mostly) correct.

I prefer to look at the monthly deltas, rather than weekly, because some of our participants submit their results infrequently. So, for the last month (which, as most know, was rather unusual because of Jerry's "blow-out")...

9,816 LL candidates were either LLed once, or factored.
10,067 LL candidates were TFed to at least 74 "bits".

Before the "incident" we were at about a 10% faster rate in TFing to at least 74 or above vs. LL completion. It's too early to tell what we're at now, but it's probably back to at least 8%.

Lastly, keep in mind that LLing gets more "expensive" the higher the candidate, while TFing gets "cheaper".