Intel tri-gate finFETs

How do you know that's a "magic byte"?

In one of the projects I lead, the programmer used an 80-byte end-of-file terminator. The chances of finding that exact terminator in a fixed-length record is very small, no? He thought so. He was wrong. It only took a few months to happen and it brought the system down. The dumbass forgot that his source code had the EOF record in it. There wasn't usually any reason to ship his code through the system but sooner or later he did and it got aligned just "right".

Reply to
krw
Loading thread data ...

AlwaysWrong things so. AlwaysWrong believes that Windows7 is perfect. AlwaysWrong is *always* wrong.

Reply to
krw

formatting link

More...

formatting link

John

Reply to
John Larkin

I "thing so"? Hahaha... you spell like a little pants past the asscrack playground punk.

I never made any such statement whatsoever, you retarded, lying punk piece of shit. Not ever once. So, you are a goddamned liar, but we already knew that.

List Windows 7 exploits and failure modes:

Damned short list. You lose... again.

Reply to
FatBytestard

Spell? Hardly. Why don't you learn to quote, AlwaysWrong.

AlwaysWrong *always* defends M$. AlwaysWrong is wrong, always.

Exactly. AlwaysWrong is *always* wrong.

Reply to
krw

"thing" IS how you spelled it, boy. Fuck off and die, you 4-F fucktard.

Reply to
FatBytestard

Well, f*****ad? Where is your list, little bitch?

Reply to
FatBytestard

The good news is that this probably gave Motorola another ten good years

- maybe more.

-- Les Cargill

Reply to
Les Cargill

Fetish again

and again...

Not ever once. So, you are a goddamned liar, but we

Google windows 7 vulnerabilities

It's not a short list.

John

Reply to
John Larkin

Linux is what?

Reply to
John KD5YI

I thought it was Reduced Instruction Set Computer (RISC not RCIS).

Reply to
John KD5YI

formatting link

Reply to
John KD5YI

Of course DimBulb is too dim to recognize a simple tupo.

If you learned to use all caps you'd as just as smart as Roy, AlwaysWrong.

Reply to
krw

Reduced Instruction Set Computing

Reply to
FatBytestard

I corrected it later; Reduced Instruction Set Complexity.

Reply to
krw

The g key is more than a typo away from the k key.

Just like you are more than a few chromosomes away from humankind.

Reply to
FatBytestard

The main reason as far as I am concerned to have more CPU power is because you need it to complete a task quicker. If you cannot harness that power then you may as well not pay for it in the first place.

If I thought Larkins suggestion of a CPU for every task would make things appreciably more secure then I would support it. But there are other ways to achieve the same goals and OS/2 came very close to doing it right nearly two decades ago.

Windows is a mess but that is more to do with the industry's ship it and be damned mentality than anything else. Bugfixes can go out later.

That seems to be what happens in a lot of multicore portables now, as well as throttling back the clock when the machine is idle. The whole multicore thing is a marketting gimmick with almost no consumer benefit. Only a handful of consumer apps really use true multiCPU operation to obtain higher performance and some of them do it badly.

Regards, Martin Brown

Reply to
Martin Brown

Relatively secure.

Thanks, Rich

Reply to
Rich Grise

It's called the CELL and can of course be found in a games console, the only platforms for which seriously powerful hardware is designed (like 256 fp core GPU's, COPPER).

Reply to
Jaded Hobo

Let each CPU have some local cache. It can spend 100% of its time computing, if you need to do that. No task switching or interrupt overhead, no compute time wasted on ethernet traffic.

A multicore chip could also have a small number of killer high-speed processors, with vector floating point and such, which could be fired up as needed. Most cores could be low power, slower.

Actually, except for gamers maybe, few people need a lot of CPU power. What people do need is security, reliability, and, increasingly, battery life.

Again, it's not about performance. It's about reliability and security. Since programmers don't seem to be very good at this, let engineers do it.

John

Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.