How lame is this?

[ElNono]

Intel charging $50 to unlock extra processor features (http://www.engadget.com/2010/09/18/intel-wants-to-charge-50-to-unlock-stuff-your-cpu-can-already-d/)

[leemajors]

I just read that, incredibly lame. I read it on boing boing, which had disturbing quotes:


I remember the first time someone from the studios put this position to me. It was a rep from the MPAA at a DRM standards meeting, and that was just the example he used. He said: "When you buy a movie to watch in your living room, we're only selling you the right to see it in your living room. Sending the same show upstairs to watch in your bedroom has value, and if it has value, we should be able to charge money for it."

http://www.boingboing.net/2010/09/19/intel-drm-a-crippled.html

AMD should take this and run the other way with it.

[CubanMustGo]

Whenever someone goes around singing the praises of the DCMA, hold on to your wallet.

[Wild Cobra]

Nothing new here except the ability to unlock the features.

This has been the commonplace in processors and memory for years.

When it comes to memory, you don't make 800mhz memory and 1066 mhz memory. It's all the same. Some of the determining factors get complicated, but often, due to expected demand for different price and performance, they downgrade a rating. You might be buying 800 mhz memory that works fine at 1066mhz. That is, if there isn't a sense line that tells the system what speed to use. I don't know all the technical details, just that more often than not, far more faster memory is made than can be sold at the higher end price.

now when it comes to CPU's, Intel has been doing this for as long as I know of. I used to be professionally involved with Intel from 1994 to 1998. They would make a full feature processor then test them. The best ones would get marked with the highest speeds and full features. Once they had the ratio of top qualifiers, the others would be downgraded. CPU would be marked with slower clock speeds. Coprocessor units and cache would be snipped out by blowing a fusible link on the chip. All they did today, was made them deactivated in firmware and capable of reactivating it.

Cry all you want. If you want the full features, you simply pay more. Nothing new or unethical here, except they made the surgery reversible.

[Wild Cobra]

I just read that, incredibly lame. I read it on boing boing, which had disturbing quotes:



http://www.boingboing.net/2010/09/19/intel-drm-a-crippled.html

AMD should take this and run the other way with it.

AMD probably disables chips also. Their chips might not be reversible though.

[Wild Cobra]

If you have two different cache sizes, it's simple. In the end, it's cheaper to print on silicon all the same chips, then customize them during the electrical testing.

The more features/performance, etc. the more they charge.

You still get what you pay for. Why cry about it?

[Dex]

http://www.boingboing.net/2010/09/19/intel-drm-a-crippled.html


Wow, that's got to be one of the most retarded, capitalistic examples I have ever read. When I pay good money for a movie, I expect to be able to watch it where I want, when I want, and however many times I want to. It's bad enough you have to shell our thirteen bucks to see it once in the theatre, and another eighteen if you want your own copy. Now they want to charge you for watching it on different TVs?

[resistanze]

I read this on Engadget. LMAO Intel.

Hackers gonna hack.

[Wild Cobra]

I read this on Engadget. LMAO Intel.

Hackers gonna hack.
That may not be so easy.

I'll bet each CPU has it's own key that needs to be activated. Not like one key will fit all.

From what I know of Intel, they wouldn't allow an externally activated feature to exist if it wasn't hacker-proof.

Did you know each CPU has a unique serial number embedded on the chip? I'll bet the process requires the system be online to activate it. the Intel server likely reads the CPU serial number, has a list of keys, and then unlocks it. I'll bet no two keys are the same.

[leemajors]

Nothing new here except the ability to unlock the features.

This has been the commonplace in processors and memory for years.

When it comes to memory, you don't make 800mhz memory and 1066 mhz memory. It's all the same. Some of the determining factors get complicated, but often, due to expected demand for different price and performance, they downgrade a rating. You might be buying 800 mhz memory that works fine at 1066mhz. That is, if there isn't a sense line that tells the system what speed to use. I don't know all the technical details, just that more often than not, far more faster memory is made than can be sold at the higher end price.

now when it comes to CPU's, Intel has been doing this for as long as I know of. I used to be professionally involved with Intel from 1994 to 1998. They would make a full feature processor then test them. The best ones would get marked with the highest speeds and full features. Once they had the ratio of top qualifiers, the others would be downgraded. CPU would be marked with slower clock speeds. Coprocessor units and cache would be snipped out by blowing a fusible link on the chip. All they did today, was made them deactivated in firmware and capable of reactivating it.

Cry all you want. If you want the full features, you simply pay more. Nothing new or unethical here, except they made the surgery reversible.

Going back and disabling hyperthreading on new models when it has been available in the past is lame.

[Wild Cobra]

Going back and disabling hyperthreading on new models when it has been available in the past is lame.
Look.

It's easier for them to make one mold instead of several. If they were all the same, they couldn't have different price structures.

[mouse]

This is like God starting to charge us for sperm we waste during masturbation.

[Wild Cobra]

This is like God starting to charge us for sperm we waste during masturbation.
Would you rather they charge top dollar for all CPU's because they were all configured fully?

You see, not everyone will pay top prices and they would have less sales. You sell a lesser product for less, and the people who are limited to that cost will buy them. there is nothing wrong with this strategy, except for those of you with the entitlement mentality. I'll bet your take is the top of the line processors should sell for the price of the cheapest ones, right?

Shouldn't you get what you pay for?

[mouse]

Shouldn't you get what you pay for?

I am all for someone getting paid for what they know or what they can provide. I charge people 25-50 dollars for a Logo and they could do it themsleves it cost me only 20 minutes of my time the software does all the work.

The chip and CPU people can make all the money they want and do as they wish but it's kinda like the AC repairman fixing your air conditioner and he knows if he turns a certain screw on the compressor you can get cooler air but he wants 50 dollars to do it, to me Its borderline chicken shit but I am not against him making money I just point out how it looks.

[Wild Cobra]

it's kinda like the AC repairman fixing your air conditioner and he knows if he turns a certain screw on the compressor you can get cooler air but he wants 50 dollars to do it, to me Its borderline chicken shit but I am not against him making money I just point out how it looks.
Whatever.

If this marketing strategy didn't work, they wouldn't do it. nobody's being cheated. they don't have a monopoly on the market. AMD gives them good competition to keep prices down.

Other than this marketing style insulting those of you with the entitlement mentality, what is wrong with it?

Are you not getting what you pay for?

You know...

The Lancer Evoution is much the same. You buy it and get a car that pushes maybe 330 HP. If you go in their with a chip programmer, you can get I think around 450 HP without making any other changes.

[BlairForceDejuan]

The ones that would want hardcore performance are already spending hundreds of dollars on watercooling setups to overclock hardware anyways.

[ElNono]

Nothing new here except the ability to unlock the features.

This has been the commonplace in processors and memory for years.

When it comes to memory, you don't make 800mhz memory and 1066 mhz memory. It's all the same. Some of the determining factors get complicated, but often, due to expected demand for different price and performance, they downgrade a rating. You might be buying 800 mhz memory that works fine at 1066mhz. That is, if there isn't a sense line that tells the system what speed to use. I don't know all the technical details, just that more often than not, far more faster memory is made than can be sold at the higher end price.

now when it comes to CPU's, Intel has been doing this for as long as I know of. I used to be professionally involved with Intel from 1994 to 1998. They would make a full feature processor then test them. The best ones would get marked with the highest speeds and full features. Once they had the ratio of top qualifiers, the others would be downgraded. CPU would be marked with slower clock speeds. Coprocessor units and cache would be snipped out by blowing a fusible link on the chip. All they did today, was made them deactivated in firmware and capable of reactivating it.

Cry all you want. If you want the full features, you simply pay more. Nothing new or unethical here, except they made the surgery reversible.

This is not binning...This is purposeful locking of features. Very different.

[ElNono]

Would you rather they charge top dollar for all CPU's because they were all configured fully?

You see, not everyone will pay top prices and they would have less sales. You sell a lesser product for less, and the people who are limited to that cost will buy them. there is nothing wrong with this strategy, except for those of you with the entitlement mentality. I'll bet your take is the top of the line processors should sell for the price of the cheapest ones, right?

Shouldn't you get what you pay for?

Your comments are retarded (nothing new there).

They already have multiple-tier pricing for different processors. You already can pay different prices for different features.

This is basically purposely crippling a processor by software in order to double-dip. Lame and silly.

And BTW, this is entirely hackable. The question is merely how much are they going to be charging for these artificial 'upgrades'. Once it makes it worth using a SEM, it will be hacked.

[ElNono]

The ones that would want hardcore performance are already spending hundreds of dollars on watercooling setups to overclock hardware anyways.

The cache and hyperthreading upgrades are basically negligible. Probably why they charge such a minuscule amount. Fact is, this is a processor that should have had those features enabled from the get go.

[ElNono]

If this marketing strategy didn't work, they wouldn't do it.

I suspect this is actually testing the waters.

[Wild Cobra]

This is not binning...This is purposeful locking of features. Very different.
call it what you will. I don't care, and i don't have a problem with the practice. If they couldn't make it soft key upgradable, they would permanently disable the CPU's with fusible internal links and force you to buy a new one at full cost.

Get a grip and think of the realities involved. this is better than the alternative.

[ElNono]

call it what you will. I don't care, and i don't have a problem with the practice.

I can tell you're talking out of your ass and don't even know what CPU binning is. Which really tells me that your 'professional involvement' with Intel from 94-98 had nothing to do with chip design and manufacturing.


If they couldn't make it soft key upgradable, they would permanently disable the CPU's with fusible internal links and force you to buy a new one at full cost.

The only reason they would do that is if the portion they're locking out by hardware is verified defective or didn't pass their internal tests.


Get a grip and think of the realities involved. this is better than the alternative.

The alternative being paying just once for everything a CPU can do? I'll take the alternative every single time.

[Cry Havoc]

call it what you will. I don't care, and i don't have a problem with the practice.

:lmao

WC, as usual, bringing his uneducated opinion and trotting it out for everyone to lulz at.

And :lmao at WC trying to prove he knows more than ElNono on the subject. :lmao

[Wild Cobra]

I can tell you're talking out of your ass and don't even know what CPU binning is. Which really tells me that your 'professional involvement' with Intel from 94-98 had nothing to do with chip design and manufacturing.

I understand binning. You see, Intels quality control is really good. They have more top spec CPU's than they can sell at top price, so they mark and market some as a lower class than they test as. In the case of CPU's with coprocessors, they would often disable a coprocessor and bin the unit as one without a coprocessor. Same holds true with cache. In the past, they relied on burning an internal link to disable the feature permanently. Now they made it a marketable feature, unlocking it with a key.


The only reason they would do that is if the portion they're locking out by hardware is verified defective or didn't pass their internal tests.

Wrong. Supply and demand. If they held for sales all that passed the highest tests, they would have too many for that class of chip, and the price falls. They would also have a shortage of the cheaper chips. Better to shoot for the best ratings and degrade some than have a shortage of chips some consumers want, and an overage of others you cannot sell.

If their supply was solely based by testing, the supply would be inconsistent.


The alternative being paying just once for everything a CPU can do? I'll take the alternative every single time.

Yes, cheaper for consumers, less profit for Intel.

You are right that I don't know the latest practices for a certainty, but you can bank on the fact that they still use a similar practice today as what I described.

[Wild Cobra]

:lmao

WC, as usual, bringing his uneducated opinion and trotting it out for everyone to lulz at.

And :lmao at WC trying to prove he knows more than ElNono on the subject. :lmao
I did work at two different angles in the semicondictor industry. Four years with an equipment manufacturer and working with Intel people, and another four years working for a Chip maker.

ElNono is wrong, and so are you.

[MannyIsGod]

:lmao

[ElNono]

I understand binning. You see, Intels quality control is really good. They have more top spec CPU's than they can sell at top price, so they mark and market some as a lower class than they test as. In the case of CPU's with coprocessors, they would often disable a coprocessor and bin the unit as one without a coprocessor. Same holds true with cache. In the past, they relied on burning an internal link to disable the feature permanently. Now they made it a marketable feature, unlocking it with a key.

Wrong. Supply and demand. If they held for sales all that passed the highest tests, they would have too many for that class of chip, and the price falls. They would also have a shortage of the cheaper chips. Better to shoot for the best ratings and degrade some than have a shortage of chips some consumers want, and an overage of others you cannot sell.

If their supply was solely based by testing, the supply would be inconsistent.

Yes, cheaper for consumers, less profit for Intel.

It doesn't work like that at all. Nobody forces Intel to sell chips. If a batch turns out to be pretty good, they will stock it.

It costs Intel a lot of money to fab those chips. Just to get a production line running on a fab can take up to 3 years.

As it is, due to the complexities of 65nm and 45nm production and the fact that Intel refuses to use SOI on their chips, they get a sizeable amount of 'defective' batches. That is, processors that either can't max out on the speed because of parasitic capacitance or because one or more parts of the components didn't quite come out right (IE: a bad cache line). Some of those chips can be salvaged (IE: by shutting down said cache using a FIB, or simply having them run at slower clock speed).

Now, I've actually designed ICs, have first hand knowledge of SEM, FIBs and all you need to peek inside chips, and what not. The kind of stuff that chipworks (http://www.chipworks.com/) will charge you an arm and a leg to do.
I can't tell you WHY I know that stuff, but I can easily spot a newbie on the topic when I see one.


You are right that I don't know the latest practices for a certainty, but you can bank on the fact that they still use a similar practice today as what I described.

So you don't know, but we should 'bank' that you're right... :rolleyes
If you don't know, the easiest way is to get informed before opening your mouth. You might come across somebody that actually does know, and you will look really stupid.

[ElNono]

:lmao

WC, as usual, bringing his uneducated opinion and trotting it out for everyone to lulz at.

And :lmao at WC trying to prove he knows more than ElNono on the subject. :lmao

Meh... it's a pretty specific subject, so I won't outright laugh at him.
I think he just has the syndrome of 'I did a gig for Intel 15 years ago, so some of the stuff I overheard maybe still applies'.

[ElNono]

BTW, I have no problem disagreeing that you think what Intel is doing now is A-OK.
Just don't come telling me it's standard practice, because I know for a fact that's BS.

[mouse]

wild cobra you used to work on towers?

[MaNuMaNiAc]

Now I don't know shit about this subject, but it wouldn't surprise me that WC talks out of his ass. He does it in almost any thread he posts.

[TDMVPDPOY]

the 50bucks only makes sense if INTEL also covers warranty for those cpus when overclocked

[Cry Havoc]

I did work at two different angles in the semicondictor industry. Four years with an equipment manufacturer and working with Intel people, and another four years working for a Chip maker.

ElNono is wrong, and so are you.

You're an idiot.

You're the guy who's asking for advice on what to buy in another thread from "real geeks" (because apparently you don't know enough about, gee, CAD to buy a computer to run it), and you think you can hold a serious discussion against ElNono?

What the fuck is wrong with your brain? Seriously, how fucked up in the head can you be to post this:

"I am looking at replacing it with the Toshiba Satellite P505-S8025. It has the i5 processor at 2.4 ghz. it has an 18.4" screen at 1680 x 945.

Thoughts please. I will talk to her again tomorrow."

Because you don't know what you need processor/GPU-wise to run CAD, FUCKING CAD applications, and then to summarily decide you know more about processors than a guy who has shown more intricate knowledge of computers than anyone in The Geek Forum, myself included?

"Hi, I'm Wild Cobra, I'll ask for your advice on a completely basic computer-purchasing issue and simultaneously tell you I know more about them than the people whose advice I'm asking for."

WC would probably walk up to Lance Armstrong and ask him what kind of road bike is good for a beginner, then tell him he's wrong for riding on a Trek frame with Bontrager wheels.

You have serious issues, dude. Your ego must be the size of fucking Mars. I can't even fathom how warped your sense of self must be.

LOL @ using anything from 1994 to apply to today's systems over someone who's still actively in the field with applicable, demonstrable knowledge of modern computers.

[resistanze]

:lmao

[Cry Havoc]

You are right that I don't know

End of the fucking discussion.


:lmao

What can I say? Stupidity upsets me at times. Usually I just laugh at it, but in special cases of vast incompetence, I can get angry.


As it is, due to the complexities of 65nm and 45nm production and the fact that Intel refuses to use SOI on their chips, they get a sizeable amount of 'defective' batches. That is, processors that either can't max out on the speed because of parasitic capacitance or because one or more parts of the components didn't quite come out right (IE: a bad cache line). Some of those chips can be salvaged (IE: by shutting down said cache using a FIB, or simply having them run at slower clock speed).

For those that are wondering, this is very accurate and extremely well-put. You can listen to this guy, or the guy who is asking you what computer he should buy for his daughter.

:lmao

Epic thread status approaching. Can't wait to see WC's next post.

[ElNono]

BTW... on the crack-proof claim and Intel being an authority in the subject of tamper-resistant solutions:

Intel Threatens to Sue Anyone Who Uses HDCP Crack (http://www.wired.com/threatlevel/2010/09/intel-threatens-consumers/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A%20wired27b%20(Blog%20-%2027B%20Stroke%206%20(Threat%20Level)))

LOL HDCP... :lmao

[resistanze]

Hackers gonna hack.

[Cry Havoc]

Hackers gonna hack.

I've got a top of the line laptop to sell WC, if he's interested.

Yo check it out, it's insanely great, it's got a 28.8 bps modem, with an active matrix, man, a million psychedelic colors!


BTW... on the crack-proof claim and Intel being an authority in the subject of tamper-resistant solutions:

Intel Threatens to Sue Anyone Who Uses HDCP Crack (http://www.wired.com/threatlevel/2010/09/intel-threatens-consumers/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A%20wired27b%20(Blog%20-%2027B%20Stroke%206%20(Threat%20Level)))

LOL HDCP... :lmao

Yeah, 4chan and Reddit lit up like Christmas trees when that crack was posted. Awesome. :lol

[DarkReign]

So weird that I find a thread mentioning HDCP.

Just this weekend I bought a bunch of BluRays and wanted to watch one of them on my PC. Fucker wouldnt play.

After scouring the net, I find out about HDCP. Turns out, you need an HDCP compliant monitor to watch BluRay movies. My monitor is 6 years old (still badass, though, imo) and I now have a BluRay RW drive that is near worthless until I buy a new one.

Such bullshit.

[ElNono]

So weird that I find a thread mentioning HDCP.

Just this weekend I bought a bunch of BluRays and wanted to watch one of them on my PC. Fucker wouldnt play.

After scouring the net, I find out about HDCP. Turns out, you need an HDCP compliant monitor to watch BluRay movies. My monitor is 6 years old (still badass, though, imo) and I now have a BluRay RW drive that is near worthless until I buy a new one.

Such bullshit.

Google 'HDFury2' and 'HDFury3'... monoprice.com has probably the best prices on them...

[resistanze]

I've got a top of the line laptop to sell WC, if he's interested.

Yo check it out, it's insanely great, it's got a 28.8 bps modem, with an active matrix, man, a million psychedelic colors!

I must say WC's new contributions to the Geek Zone in addition to the Political Forum has been nothing short of legendary. I wonder what kind of interesting i486 processors he could be selling in the Buy, Sell, Trade, Network.

[MannyIsGod]

:lmao

this site is endless entertainment.

[leemajors]

So weird that I find a thread mentioning HDCP.

Just this weekend I bought a bunch of BluRays and wanted to watch one of them on my PC. Fucker wouldnt play.

After scouring the net, I find out about HDCP. Turns out, you need an HDCP compliant monitor to watch BluRay movies. My monitor is 6 years old (still badass, though, imo) and I now have a BluRay RW drive that is near worthless until I buy a new one.

Such bullshit.

AnyDVD HD works great.

[DarkReign]

Google 'HDFury2' and 'HDFury3'... monoprice.com has probably the best prices on them...


AnyDVD HD works great.

Talk to me like I am an idiot...

Are you both saying these are workarounds for HDCP compliant monitors?

If so, you are frickin awesome.

[DarkReign]

Google 'HDFury2' and 'HDFury3'... monoprice.com has probably the best prices on them...

Oh! I get it.

HDCP must only work when an HDMI cable is plugged into the video card.

So youre saying (correct me if I am wrong) that using an HDMI converter is a workaround for not having a HDCP compliant monitor?

BTW, I use a DVI cord for my ancient monitor.

[leemajors]

Oh! I get it.

HDCP must only work when an HDMI cable is plugged into the video card.

So youre saying (correct me if I am wrong) that using an HDMI converter is a workaround for not having a HDCP compliant monitor?

BTW, I use a DVI cord for my ancient monitor.

DVI to HDMI cables won't work either. I ended up buying a lifetime license for AnyDVD HD, it bypasses/removes the HDCP crap. I bought a blu-ray drive for my comp last year and was really pissed when I couldn't output to my TV through my DVI to HDMI cable, but this program works great and is updated nearly weekly and I haven't run across any discs it can't decode. Also allows you to rip blu rays if you choose. Only drawback is it's $80, I think.

[Cry Havoc]

http://www.funnyforumpics.com/forums/This-Thread-Delivers/1/thread-delivers.jpg

[Wild Cobra]

Meh... it's a pretty specific subject, so I won't outright laugh at him.
I think he just has the syndrome of 'I did a gig for Intel 15 years ago, so some of the stuff I overheard maybe still applies'.
In many ways you are right. however, I am only speaking of binning to match supply with demand. I know there are differences from the past to today. I already pointed that out. If it makes you feel better to focus on that, and write a book about the newer areas then fine. Stroke your own ego.

Do yo disagree that that some chips are intentionally degraded? Or that some CPU packages are intentionally degraded?

In the past, some chips may have part of the stricture that fails, so they are electrically removed, and sold as a lesser product. At times, they even purposely degraded perfect working units for proper supply ratios to demand. This is an undisputed fact.

What is wrong with now doing this is firmware? Giving the user the option to buy a CPU that is a few hundred dollars and upgrading for less than a hundred more later. It saves the user from replacing the entire CPU at a few hundred dollars.

Why are you complaining about this?

[Wild Cobra]

You're an idiot.

You're the guy who's asking for advice on what to buy in another thread from "real geeks" (because apparently you don't know enough about, gee, CAD to buy a computer to run it), and you think you can hold a serious discussion against ElNono?

What the fuck is wrong with your brain? Seriously, how fucked up in the head can you be to post this:

"I am looking at replacing it with the Toshiba Satellite P505-S8025. It has the i5 processor at 2.4 ghz. it has an 18.4" screen at 1680 x 945.

Thoughts please. I will talk to her again tomorrow."

Because you don't know what you need processor/GPU-wise to run CAD, FUCKING CAD applications, and then to summarily decide you know more about processors than a guy who has shown more intricate knowledge of computers than anyone in The Geek Forum, myself included?

"Hi, I'm Wild Cobra, I'll ask for your advice on a completely basic computer-purchasing issue and simultaneously tell you I know more about them than the people whose advice I'm asking for."

WC would probably walk up to Lance Armstrong and ask him what kind of road bike is good for a beginner, then tell him he's wrong for riding on a Trek frame with Bontrager wheels.

You have serious issues, dude. Your ego must be the size of fucking Mars. I can't even fathom how warped your sense of self must be.

LOL @ using anything from 1994 to apply to today's systems over someone who's still actively in the field with applicable, demonstrable knowledge of modern computers.
You're the idiot when you cannot acknowledge I already pointed out, and that was a practice then, and that some of those tactics will remain today.

[Cry Havoc]

http://freon.shackspace.com/misc/backpedaling.gif

[Cry Havoc]

ElNono is wrong, and so are you.


You are right that I don't know.


In many ways you are right.

[Cry Havoc]

Get a grip and think of the realities involved.

[leemajors]

What is wrong with now doing this is firmware? Giving the user the option to buy a CPU that is a few hundred dollars and upgrading for less than a hundred more later. It saves the user from replacing the entire CPU at a few hundred dollars.

Why are you complaining about this?

This isn't what Intel is doing here.

[Wild Cobra]

http://freon.shackspace.com/misc/backpedaling.gif
If you want to call it that, I don't care.

When I left the semiconductor industry, they were starting to deal with the realities requiring the low-k dielectrics that are now absolutely required. SOI (silicone on insulator) was an established process, but we weren't using it in 98 to 02, but in my prior job in the industry, we did experiment with sapphire wafers (SOS,) before 98.

I don't think anyone was selling smaller than a 130nm technology at the time. We were still doing 250nm in 2002, a little behind the leaders like Intel. We were just starting to use trench designs a couple years before I left, and planning to start commercial copper processes.

I understand ElNono's lengthy post. Did you? Layer deposition becomes more critical because CMP is only meant to compensate for step height ratios. Our problem as the company didn't buy good enough deposition equipment and our CMP equipment couldn't compensate for going smaller than 250nm. CMP is required for a flat surface for focusing in photo-lithography. All these things must come together, and more failures occur at this level of miniaturization. For that reason, the intentional binning of products into lower classed bins would be reduced, but I doubt it's entirely eliminated.

[MannyIsGod]

He doesn't get that and he doesn't understand how everyone jumping on him is doing so because he brings up irrelevant information and then acts like they're trying to prove him wrong on that.

Its not the first time he's done this, and I'm sure it won't be the last.

[Wild Cobra]

This isn't what Intel is doing here.
OK, what did I miss?

[MannyIsGod]

OK, what did I miss?

:lmao

[Cry Havoc]

ElNono is wrong, and so are you.


You are right that I don't know.


In many ways you are right.


OK, what did I miss?

:lmao

[ElNono]

Oh! I get it.

HDCP must only work when an HDMI cable is plugged into the video card.

So youre saying (correct me if I am wrong) that using an HDMI converter is a workaround for not having a HDCP compliant monitor?

BTW, I use a DVI cord for my ancient monitor.

The HDFury 2/3 will take the input of a HDMI signal (with and without HDCP) and output it to component/VGA, effectively allowing you to use your old monitor/tv with hardware that requires HDCP. Be warned that the HDCP consortium might at some point decide to revoke the key for the chip the Fury uses, thus rendering it useless. That said, it's been around for ages...

[ElNono]

In many ways you are right. however, I am only speaking of binning to match supply with demand. I know there are differences from the past to today. I already pointed that out. If it makes you feel better to focus on that, and write a book about the newer areas then fine. Stroke your own ego.

Do yo disagree that that some chips are intentionally degraded? Or that some CPU packages are intentionally degraded?

I dispute that notion when it comes to CPUs made by Intel at this point in time (which happens to be the topic at hand).


In the past, some chips may have part of the stricture that fails, so they are electrically removed, and sold as a lesser product. At times, they even purposely degraded perfect working units for proper supply ratios to demand. This is an undisputed fact.

In the past production yielded a good amount of perfectly working units. In the past, it was actually hard to rescue a bad chip, because the design was far from modular (I'm referring to the pre-68020 era).

However, we're not in the past anymore.

The die for a single core vs a dual core vs a quad core is pretty different. It makes sense, because the cost of creating the die design has gone down considerably from times past, and they can pack much more dies of a single processor type on a wafer that way.

There's also the fact that with AMD supplying compatible chips, the price points you see at this time actually paint a good picture of yield of different dies. An Extreme Edition type of i7 CPU costs an arm and a leg in general because it's really hard for Intel to obtain a chip of such quality with the current manufacturing process. They don't need to cripple those to, say, supply i7 960 CPUs for half the price, simply because most of the yield on the wafer are going to be i7-960 quality dies.


What is wrong with now doing this is firmware? Giving the user the option to buy a CPU that is a few hundred dollars and upgrading for less than a hundred more later. It saves the user from replacing the entire CPU at a few hundred dollars.

You need to first understand that you're comparing orange to apples.

To give an analogy that I think it's apt, this is like Ford selling you a car that maxes out at 40mph, and you having to pay extra to 'unlock' the extra speed.
You might think that it's a good business model for Ford, but it's undeniably lame as hell.


Why are you complaining about this?

I'm not complaining. I think it's super lame.

[ElNono]

If you want to call it that, I don't care.

When I left the semiconductor industry, they were starting to deal with the realities requiring the low-k dielectrics that are now absolutely required. SOI (silicone on insulator) was an established process, but we weren't using it in 98 to 02, but in my prior job in the industry, we did experiment with sapphire wafers (SOS,) before 98.

I don't think anyone was selling smaller than a 130nm technology at the time. We were still doing 250nm in 2002, a little behind the leaders like Intel. We were just starting to use trench designs a couple years before I left, and planning to start commercial copper processes.

I understand ElNono's lengthy post. Did you? Layer deposition becomes more critical because CMP is only meant to compensate for step height ratios. Our problem as the company didn't buy good enough deposition equipment and our CMP equipment couldn't compensate for going smaller than 250nm. CMP is required for a flat surface for focusing in photo-lithography. All these things must come together, and more failures occur at this level of miniaturization. For that reason, the intentional binning of products into lower classed bins would be reduced, but I doubt it's entirely eliminated.

BTW, to this day, Intel CPUs still do not use SOI. I don't know if it's a patent issue or what exactly, but baffles the mind.

[DarkReign]

DVI to HDMI cables won't work either. I ended up buying a lifetime license for AnyDVD HD, it bypasses/removes the HDCP crap. I bought a blu-ray drive for my comp last year and was really pissed when I couldn't output to my TV through my DVI to HDMI cable, but this program works great and is updated nearly weekly and I haven't run across any discs it can't decode. Also allows you to rip blu rays if you choose. Only drawback is it's $80, I think.

I googled your advice, seems a competent option.

Graci

[DarkReign]

The HDFury 2/3 will take the input of a HDMI signal (with and without HDCP) and output it to component/VGA, effectively allowing you to use your old monitor/tv with hardware that requires HDCP. Be warned that the HDCP consortium might at some point decide to revoke the key for the chip the Fury uses, thus rendering it useless. That said, it's been around for ages...

...and I am not that bright to have known about it before today. Yesterday was the first time I had ever heard of HDCP.

Way it goes.

[Wild Cobra]

This isn't what Intel is doing here.
What did I miss?

Intel websites confirm -- that lets you download software to unlock extra threads and cache on the new Pentium G6951 processor. Hardware.info got their hands on an early sample of the chip and discovered it's actually a full 1MB of L3 cache that's enabled plus HyperThreading support, which translates to a modest but noticeable upgrade. This isn't exactly an unprecedented move, as chip companies routinely sell hardware-locked chips all the time in a process known as binning, but there they have a simpler excuse -- binned chips are typically sold with cores or cache locked because that part of their silicon turned out defective after printing.
What didn't I understand?

I was explaining this was a common practice of the past, and it appears they continued I think it was SX models that sometimes came from perfectly good DX models, but that's a long time to remember specifics. Just couldn't re-enabled the feature once it was taken out.

What is the price difference between the different G6951 processors anyway? I don't know, and there are too many search results from this controversy.

Would you prefer to buy a lesser one for less money, then later pay more money than you did at first for a whole new one, or pay less to upgrade?

You and the computer makers have an option. Without the G6951, you would be buying the G6950 and upgrading to the G6952 or a different CPU at full price ind installation if you pay someone else to do it.

[Wild Cobra]

BTW, to this day, Intel CPUs still do not use SOI. I don't know if it's a patent issue or what exactly, but baffles the mind.
I've been out of the loop with Intel folks for I think 9 years, but they use to have advanced proprietary methods others don't use. The probably still do.

It was interesting during the short time work I did contract work at Intel's Portland Development facility.