Slightly off topic.

@KWSN
I've just started running the 2.2 C2 SSSE3 app. I used to run the 1.4 C2 SSSE3 app and stopped because it was missing signals and was only weakly similar when validated. I then switched over to the 2.0 P4 SSE3 so that my WUs would validate better. My question now is whether the 2.2 C2 SSSE3 app misses signals like the first version did or will the finished WUs be more similar to the 2.2 P4 SSE3 version?

It may be far to early to know this. In any case I'll watch my validated WUs closely at least for a while.

Do you have an estimate of how much faster the 2.2 C2 SSSE3 is than the 2.2 P4 SSE3 version?

Hi folks,

bet that your patched apps can't match the unpatched newly released Rev-2.2 ones ;o)

A new patchfest may ensue, but hey, it's all in the name of quicker crunching.

Regards,
Simon.

Thanks again Simon! trying the SSE2 now. Is it really illegal to discuss patching an app?

---

Official Abuser of Boinc Buttons...
And no good credit hound!

Thanks again Simon! trying the SSE2 now. Is it really illegal to discuss patching an app?

I hope NOT!!!...

(You don't have to accept all the EULA's restrictions. You can argue against it. I suspect that some of the EULA 'requirements' are in themselves unreasonable, or illegal or unenforcable.)

[edit] There is also the question of whether whichever EULA conditions can be deemed to be "anti-competitive", which is illegal in this country at least. [/edit]

Happy crunchin',
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Thanks again Simon! trying the SSE2 now. Is it really illegal to discuss patching an app?

I hope NOT!!!...

(You don't have to accept all the EULA's restrictions. You can argue against it. I suspect that some of the EULA 'requirements' are in themselves unreasonable, or illegal or unenforcable.)

[edit] There is also the question of whether whichever EULA conditions can be deemed to be "anti-competitive", which is illegal in this country at least. [/edit]

Happy crunchin',
Martin

I was curious when I got this message from Crunch3r...
"P.S. OFF TOPIC

so while you're listening/reading the crap i post :-)... please do NOT post any of the "AMD patching"
stuff here again ... I had a talk to Intel Premier Support and... guess what ... they said it's illegal ;)

So please don't post thing like that again...

Thank you.







____________
regards
Crunch3r"

---

Official Abuser of Boinc Buttons...
And no good credit hound!

I guess that according Intel even the release of the original Athlon would be illegal.

Of course crippling a compiler to harm your competitor is absolutely legal. Wonder why they didn't start bombing AMD fabs, that must be legal for them, too ;)

---

Hi folks,

bet that your patched apps can't match the unpatched newly released Rev-2.2 ones ;o)

A new patchfest may ensue, but hey, it's all in the name of quicker crunching.

Regards,
Simon.

You've won your bet! I've been under 8000 seconds for the first time ever on a WU with a 0.425850 AR. I'm very impressed! Keep up the good work!

---

I was curious when I got this message from Crunch3r...
"P.S. OFF TOPIC

so while you're listening/reading the crap i post :-)... please do NOT post any of the "AMD patching"
stuff here again ... I had a talk to Intel Premier Support and... guess what ... they said it's illegal ;)

So please don't post thing like that again...

Thank you.

Oh for gods sake, are you THAT naive? Of COURSE Intel said it was illegal, they would wouldn't they???

Why does anyone ever bother with AMD chips anyway???? they are the worst crap ever invented in the history of computers.

---

LOL OK I hereby pronounce the word "Patch" shall be changed to "Pabst" (a crappy American beer, for our Eurofriends).

there....that's fixed.

Slightly off topic.

@KWSN
I've just started running the 2.2 C2 SSSE3 app. I used to run the 1.4 C2 SSSE3 app and stopped because it was missing signals and was only weakly similar when validated. I then switched over to the 2.0 P4 SSE3 so that my WUs would validate better. My question now is whether the 2.2 C2 SSSE3 app misses signals like the first version did or will the finished WUs be more similar to the 2.2 P4 SSE3 version?

1. The validator allows up to a 1% difference between results for reported signal levels. The 1.41 code was usually better than that, but perhaps not at high Chirp rates and toward the end of the 107.34 second duration of a WU. The 2.2 code is much better than that for all cases.

2. The decision on whether a signal is good enough to report is made in the application, and cannot have a tolerance. For signals near the threshold, the 1.41 code may report a signal where the stock app doesn't, or vice versa. The 2.2 code can do the same, but it is much less likely.

3. The result file includes "best of" signal reports which are also checked by the validator. Again, minor differences could cause 1.41 to judge differently which of two signals is best. The 2.2 code can do the same, but it is much less likely.

It may be far to early to know this. In any case I'll watch my validated WUs closely at least for a while.

Do you have an estimate of how much faster the 2.2 C2 SSSE3 is than the 2.2 P4 SSE3 version?

Plus or minus 2% is my estimate. Either might be the best for a particular Core 2 system, and perhaps even one of the other 2.2 builds would exceed both. The differences are whatever the Intel compiler adjusts for the various targets, and though that compiler does very well the S@H app is not exactly the typical kind of usage for which the tuning is done.

                                                             Joe

Why does anyone ever bother with AMD chips anyway???? they are the worst crap ever invented in the history of computers.
[/quote]
Think about it...Were it not for AMD, the chips we use today would not be as advanced or as cheap. AMD helps keep intel on track, we need the competition...Can you imagine the way Windows would be today had Apple decided to port to Intel and competed with Microsoft all these years?

On Subject...If you install the wrong app on the wrong machine it errors out real quick...LOL

---

Official Abuser of Boinc Buttons...
And no good credit hound!

Why does anyone ever bother with AMD chips anyway???? they are the worst crap ever invented in the history of computers.

Think about it...Were it not for AMD, the chips we use today would not be as advanced or as cheap. AMD helps keep intel on track, we need the competition...

So what of the very expensive Intel P4 "Netburst" fiasco?

That took a roll-back to the P3 design by an Israeli design team, and a two year clear lead by AMD to sort out!

Intel is still strangled by their FSB restrictions that has meant much more expensive RAM is required and additional expense (and heat) on the motherboard Northbridge.

Meanwhile, AMD has only recently needed to move up to DDR2, very nicely at the same time as the sweet spot for pricing for that RAM.

So who's Marketing crap do you swallow hook line and sinker?

Or do you just like those Intel boyz prancing about in their spacesuits adverts?


Happy crunchin',
Martin

ps: Both Intel and AMD have their strengths and weaknesses. Its all down to what you want to do, how, and for how much.

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Okay folks,

the AMD vs. Intel debate doesn't really belong here. Both are great companies with great products. 'nuff said.

As far as pabsting (good one Tony!) goes, I really believe talking about it and even telling people how to do it is not even close to illegal. After all, debugging can hardly be called unfair use (which is how this method of unlocking ICC-compiled apps for non-Intel CPUs was realised).

The actual act may be different, but then that's why I don't offer patched apps.

People can make up their own mind and make an informed choice, I'd prefer that to blindly following propaganda any day.

Enough of that now, too.

As Joe posted, one of the bigger improvements on Core 2 systems was the accuracy on VHAR WUs. That the apps are quicker than before is a nice side effect, of course ;o)

Like I said earlier in this thread, the unpatched R-2.2 apps are quicker than the patched 2.0 ones. It's all good in my book.

Regards,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

And a big thanks to you and your dev team!

---

me@rescam.org

And a big thanks to you and your dev team!

I gotta admit- this time it wasn't Misfit's fault. It was-

- Ben Herndon
- Joseph Segur
- Alex Kan
- Simon Zadra

Sometimes known as the "Chicken Crew"?

Way to go guys!

Brock

---

Okay folks,
Like I said earlier in this thread, the unpatched R-2.2 apps are quicker than the patched 2.0 ones. It's all good in my book.

Regards,
Simon.

Any chance you'll compile the 2.2 flavor w/o graphics for Linux too?

The speedup I see between 2.0 vs 2.2 on an Athlon XP is almost rediculous.
And I found the speedup between vanilla and 2.0 already mindblowing XD


Btw, the plain 2.0 SSE client, how come it runs SSE correctly on Socket A Athlons? (Or doesn't it and is the speedup I got from going from Vanilla to 2.0 SSE just from the plain optimizations?)

From what I read about the ICC hack, the naughty code Intel put in the compiler that checked for the GenuineIntel string impacted SSE aswell, or have they removed the naughty bit by now in the new versions?

After all, debugging can hardly be called unfair use (which is how this method of unlocking ICC-compiled apps for non-Intel CPUs was realised).

I'm not quite sure if ya'll get this right an how serious that is ...

It's not debugging it's patching/disassembling and that's what is beeing done here.

Parts of the intel libraries linked into the code of the binary (ipp & icc ) that do check for intel cpus (may it be good or bad ) are patched and that is illegal.

It's like someone post a howto for cracking Windows Vistas activation code and how to prevent it from stopping to work after 30 days.

---

Join BOINC United now!

Cruncher: actualy what Intel is doing is borderlining (if not plain) illegal.

They license out SSE and SSE2 to AMD, who pay rather massive fees for it and subsequently make SSE and SSE2 not work on CPU's that aren't from Intel.

After all, debugging can hardly be called unfair use (which is how this method of unlocking ICC-compiled apps for non-Intel CPUs was realised).

I'm not quite sure if ya'll get this right an how serious that is ...

It's not debugging it's patching/disassembling and that's what is beeing done here.

Parts of the intel libraries linked into the code of the binary (ipp & icc ) that do check for intel cpus (may it be good or bad ) are patched and that is illegal.

It's like someone post a howto for cracking Windows Vistas activation code and how to prevent it from stopping to work after 30 days.

Crunch3r, that is not true as such.

There are two ways described on the site how to do it - one is to patch your libs and compile using patched libs. This is what you're describing and definitely a BAD idea, I'd agree.

BUT this is not what people are doing, and not what I've suggested, either ;o)

Instead, here, there is a script that patches EXECUTABLES.

And yes, this method was developped using debugging, not reverse-engineering.

As far as the Vista analogy - er, please. The DMCA does not apply here. That's really a bad and invalid comparison.

Using fear to coerce people to see things your way really upsets me. Not cool.

Regards,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Cruncher: actualy what Intel is doing is borderlining (if not plain) illegal.

They license out SSE and SSE2 to AMD, who pay rather massive fees for it and subsequently make SSE and SSE2 not work on CPU's that aren't from Intel.

Wow!

That is even more nasty.

Sure shows why Intel and Microsoft are such good "bedfellows"...

Microsoft is effectiviely doing similar tricks to make Windows Vista deliberately not work on "non authorised" hardware. The sting is that for such hardware to be "authorised" by Microsoft, the hardware must be made for Windows only and no other operating system. The likely hope is that Apple OS X, Linux, BSD, and all the other OSes that use PC hardware will get starved of cheap new hardware and so then will enable Microsoft to increase its prices at everyone else's expense. (All just my personal conjecture ofcourse.)

See: Windows Vista Content Protection for various incredulous gory bits. No wonder Vista has taken a long time in development! That has taken quite some effort to coerce the hardware manufacturers to swallow that one!

Ouch!
Martin

[edit]
Reference added.

Also see: Vista Upgrade Decision Chart
[/edit]

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Using fear to coerce people to see things your way really upsets me. Not cool.

I very strongly agree on that one.


[rant]
Similarly, I've got a strong distaste for the whole Microsoft Viruses, Worms, Trojans, malware expensive fiasco and the "Marketing" tricks of fear used by both Microsoft and various "Internet Security" companies whom then foist their expensive software and updates on computer users.

Somehow, the source problem of the underlaying system inadequacies never get mentioned by those same Marketing types. I guess putting lots of duct tape over all the holes to patch over the symptoms is far more profitable...

All a very expensive waste of time and money that is holding the whole world back.

Just sheer blind stupidity.
[/rant]

Regards,
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)