Intel updates microcode to block H87/B85 overclocking [BT]

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

cytg111

Lifer
Mar 17, 2008
26,206
15,619
136
It wasn't Intel's intention to OC on these boards, it's part of their product segmentation. The fact that mobo manufactures tried to go around Intel is pretty shameless, and as such they should bare the burden of appeasing any affected customers with a proper z87 motherboard..

- It is still Intel pulling the trigger no matter how we spin it. The end user dont give a crap about the blame game. In the end they're gonna end up with less of a product. And Intel has the gun residue all over. Intel *should* go "Oh that was not what we intended, oh well, lesson learned till next time" and not "James, my gun please"
 

BallaTheFeared

Diamond Member
Nov 15, 2010
8,115
0
71
- It is still Intel pulling the trigger no matter how we spin it. The end user dont give a crap about the blame game. In the end they're gonna end up with less of a product. And Intel has the gun residue all over. Intel *should* go "Oh that was not what we intended, oh well, lesson learned till next time" and not "James, my gun please"

It's still not Intel's fault no matter how you slice it.

The blame lies with the mobo manufactures who went outside of spec.

Same thing happened to EVGA when they tried to circumvent Keplers voltage limits, big daddy Nvidia brought the hammer down.

Of course Nvidia didn't go to any extremes like bios updates or windows updates to lock them out, but the principle is still the same. It's Intel's product, and they give you the right to sell it. Follow the rules or sell AMD boards only.

Mobo manufactures should step up and offer an upgrade through RMA for customers who wish to trade in their previously sold as overclockable non z87 board for a z87 board.
 

cytg111

Lifer
Mar 17, 2008
26,206
15,619
136
I still dont get it. I have a product, not a faulty one, it does not compute wrong, does not set things on fire, yet it gets an update from the factory that artificially shrinks its capabilites. Intel is pulling that trigger. If I was an enduser of said product, had endless $$ and nothing else to do with my time I think I'd like to try that one out in court.
 

SOFTengCOMPelec

Platinum Member
May 9, 2013
2,417
75
91
I can see both sides to the argument.

Intel has a business case to make, be it limiting their liability or maximizing their profitability, either way there are literally thousands of people whose livelihoods and jobs depend on Intel making the tough decisions needed to ensure the cash flow is healthy enough to keep paying those folks their salaries so they in turn can keep paying their mortgage and buy food and send their kids to school and pay taxes, etc.

On the other side, my absolute favorite rig build of all time was my dual-550MHz Celeron machine using the Abit BP6 mobo.



So not only were you buying the dirt-cheap CPU version from Intel (their Celeron line), you were (1) overclocking them from 333MHz to 550MHz (w00t!), and (2) putting them into a silly cheap dual-socket mobo and running them in SMP like a boss!

Now then, Intel obviously took measures to prevent this from happening again but for a brief period in time enthusiasts could build quite the powerful and cheap box.

So I have to ask myself, how would I have felt in 1999 if Intel reached out through Microsoft with a microcode "update for security" that totally borked my BP6 duallie and operationally made it a single-socket mobo running at near-stock clockspeeds?

I'd be pissed. Disgusted and pissed.

So I can understand the emotional response that Intel is getting from this total dick move on their part.

It is all about the upsell. Intel really is no different than your local car dealership when it comes to their marketing and sales strategies.

Get your enthusiasts to pay a premium for the "K" chip (upsell #1), then make them pay a premium to get the right chipset as needed to actually use the unlocked features of your CPU (upsell #2), and don't forget to lay it on heavy and thick that they better buy that "extended warranty" performance tuning plan (upsell #3).

Upsell, upsell, upsell.

If your livelihood depends on Intel having a healthy robust fiscal outlook then you are pleased to see Intel doing what it needs to ensure you have continued job security, that is a very natural and understandable human desire. I've been there (not at Intel, but at others) and it sucked to see my management not minding the shop well enough to safeguard the financial future of employees like myself.

But for the rest of the consumer market, this stuff just creates friction as it pits consumers against the seller in an adversarial way and the history of commerce shows this always, always, results in a detrimental outcome for the business that is waging war on its customer.

Intel needs to find ways to add value to customers, not find ways to reduce or minimize the value it is adding to its customers. This move, while justified, is an act of war on its own customers and those are the kind of war wounds that people remember for a long long time.

My answer might actually surprise you.

BOTH the things that you mentioned, actually happened to me.

I also loved getting the Abit BP6's (best to put a tiny fan on its too hot chipset). I think I remember going to the computer shop (or it might have been mail order), and buying like 10 Celerons at a time, because they were so relatively cheap (about $50 in US then I guess, and about £50 here). Sometimes worrying about the exact batch numbers, for best overclocking.
The advantage of buying 10, is that I could find the best "overclockable" ones to match up on the wonderful Abit BP6 motherboards.
The overclocks then were amazing, and most of the extra clock speed, really gave extra speed.
Often the £50 bottom chip, could be overclocked by about 50%, approaching the performance of the top £450 Pentium chip.

And then of course, you could have 2 of them on each board, when most people ONLY had one core.

Now the 2nd twist of the tail.

The windows update that kills the overclock, unbelievably, happened to me.
(Ok, I'm partly lying here, but let me explain what happened).

Of course I had to make up multiple Abit BP6 computers, as they were so cheap, and I wanted many, many cores. I won't embarrass myself here and admit to how many systems I joined together to sort of make one super computer.

Now, this is where I got hit with the "windows update, computer killer" effect.

I think I had to use windows NT to support the dual processors (there may have been other options, I'm not sure, can't remember).
So I networked them all together.

But Microsoft (if you think I hate Intel (I DON'T), then meet me in real life, and ask me what I think of Microsoft, and Bill Gates, and you will know that I am NOT 100% delighted with them, lol) was programmed to detect that I was networking these computers together, without using a server version of their software.

So to my HUGE annoyment, it seemed to automatically "CRASH"/stop, after a certain limit of network traffic/computer numbers. I don't know the exact criterion Microsoft used, but much, much later, I found articles on the internet, which seemed to confirm my suspicions, of these artificial limits, to prevent saving money buying Microsoft server editions.

Why multi computers ?
I like doing all sorts of computing stuff, such as (at that time, potentially), extensive prime95 mersenne prime searches, complicated Chess puzzle analysis, running my own programs for long test runs, etc etc.

Computers were relatively slow then, so multiple computers (many were not only slow then, but most were single core, single cpu), made more sense, than it does these days, as you can get a 6 core, relatively high speed computer, which will handle most things, on its own.

Back on topic:
Yes, there are indeed multiple sides to this issue. On the one side Intel protecting its earnings, and on the other, fairness to consumers.

I think a few years or so, ago, Intel trialled Internet upgradeable processors. You buy a relatively cheap Intel cpu, and have it in your computer. You then choose a better Intel cpu, and buy an upgrade code from Intel, via the internet.
The special code then changes options in the cpu, which improves its performance, and features, etc.

People absolutely HATED the idea, and I think it was a complete and total flop, due to massive unpopularity with customers.

Anyway, thanks for the informative reply!

EDIT: It was a very long time ago. I can't clearly remember what I did, and did not run on the computers. So take the "what I ran" list as VERY approximate. It may have applied to computers I had before, or after the BP6 era, I'm not 100% sure.
 
Last edited:

SPBHM

Diamond Member
Sep 12, 2012
5,066
418
126
Yes the fact does remain, the fact that Intel did not offer overclocking on non z87 motherboards... Which is the only fact that matters.

Cheapest z87 is already sub $100, how much lower were you hoping to go but stay above $50 and still feel the need to post about it?

Why is someone buying an overclocking $240 CPU and putting it in a $50 motherboard? What exactly is the use case for such a purchase?

oh, nice to see that newegg have a z87 board for $97 or so, but it doesn't change the fact that you could go even lower keeping the possiblity to OC, you had that choice, Intel is taking that away with an update.

Intel offered B85 in way it could be used for overclocking, that's a fact ;)

use case is simple, trying to save as much money as possible for the most important parts (CPU and VGA if you are a gamer)


I also killed a P67 UD4, endless boot cycle just as the chipset recall went down :|

Or you could have gotten an i5-4670k and $180 quality motherboard with $80 off :hmm:

again, having as many options as possible is nice, if you settle down for a lower OC the risk is greatly reduced.




talking about Dual Celeron, it also reminds me of Tualatin vs Coppermine MBs... Intel was trying to sell new chipsets/MBs but it was nothing a pin mod couldn't solve...
 
Last edited:

cytg111

Lifer
Mar 17, 2008
26,206
15,619
136
So not only were you buying the dirt-cheap CPU version from Intel (their Celeron line), you were (1) overclocking them from 333MHz to 550MHz (w00t!), and (2) putting them into a silly cheap dual-socket mobo and running them in SMP like a boss!.

Off topic, i know, sorry, but as I recall it, for SMP to work you had to drill some holes in those celerons. Had a friend back in uni that did that with 300a's for dual 450. Did you do the drill hack? (at the same time I had a 300a running 504 .. also my favorite rig of all time).


... Sometimes worrying about the exact batch numbers, for best overclocking.

- Heh, did that too, I cherry picked a Malay part, did 504 with no hazzle too.. It did take me some time thou to find a vendor that would let me pick based on serial numbers ..
 
Last edited:

Idontcare

Elite Member
Oct 10, 1999
21,110
64
91
Now the 2nd twist of the tail.

The windows update that kills the overclock, unbelievably, happened to me.
(Ok, I'm partly lying here, but let me explain what happened).

Of course I had to make up multiple Abit BP6 computers, as they were so cheap, and I wanted many, many cores. I won't embarrass myself here and admit to how many systems I joined together to sort of make one super computer.

Now, this is where I got hit with the "windows update, computer killer" effect.

I think I had to use windows NT to support the dual processors (there may have been other options, I'm not sure, can't remember).
So I networked them all together.

But Microsoft (if you think I hate Intel (I DON'T), then meet me in real life, and ask me what I think of Microsoft, and Bill Gates, and you will know that I am NOT 100% delighted with them, lol) was programmed to detect that I was networking these computers together, without using a server version of their software.

So to my HUGE annoyment, it seemed to automatically "CRASH"/stop, after a certain limit of network traffic/computer numbers. I don't know the exact criterion Microsoft used, but much, much later, I found articles on the internet, which seemed to confirm my suspicions, of these artificial limits, to prevent saving money buying Microsoft server editions.

Why multi computers ?
I like doing all sorts of computing stuff, such as (at that time, potentially), extensive prime95 mersenne prime searches, complicated Chess puzzle analysis, running my own programs for long test runs, etc etc.

Computers were relatively slow then, so multiple computers (many were not only slow then, but most were single core, single cpu), made more sense, than it does these days, as you can get a 6 core, relatively high speed computer, which will handle most things, on its own.

Back on topic:
Yes, there are indeed multiple sides to this issue. On the one side Intel protecting its earnings, and on the other, fairness to consumers.

I think a few years or so, ago, Intel trialled Internet upgradeable processors. You buy a relatively cheap Intel cpu, and have it in your computer. You then choose a better Intel cpu, and buy an upgrade code from Intel, via the internet.
The special code then changes options in the cpu, which improves its performance, and features, etc.

People absolutely HATED the idea, and I think it was a complete and total flop, due to massive unpopularity with customers.

Anyway, thanks for the informative reply!

EDIT: It was a very long time ago. I can't clearly remember what I did, and did not run on the computers. So take the "what I ran" list as VERY approximate. It may have applied to computers I had before, or after the BP6 era, I'm not 100% sure.

I avoided the MS tax by clustering my rigs into a Beowulf using redhat Linux. MS was far more predatory than Intel back then, nowadays Intel has stepped up its game.

Off topic, i know, sorry, but as I recall it, for SMP to work you had to drill some holes in those celerons. Had a friend back in uni that did that with 300a's for dual 450. Did you do the drill hack? (at the same time I had a 300a running 504 .. also my favorite rig of all time).

I never had to drill or otherwise modify my Celerons for BP6 operation. I think that came later after Intel made its move to "address" the undesired SMP OC'ing market.
 

SOFTengCOMPelec

Platinum Member
May 9, 2013
2,417
75
91
Off topic, i know, sorry, but as I recall it, for SMP to work you had to drill some holes in those celerons. Had a friend back in uni that did that with 300a's for dual 450. Did you do the drill hack? (at the same time I had a 300a running 504 .. also my favorite rig of all time).




- Heh, did that too, I cherry picked a Malay part, did 504 with no hazzle too.. It did take me some time thou to find a vendor that would let me pick based on serial numbers ..

I'm confused memory wise about the batch numbers situation, because other cpus I purchased, sometimes the batch number was a consideration.
A friend I had at the time, was also madly keen on overclocking the celerons, I think we split the cpus between ourselves, or something.
He may have been more keen on batch number checking.
I think my method was buy one from a supplier (mail order), and if it is a good batch number, buy more.
And potentially sell on the spare or not very overclockable cpus, or use them to make cheap low end computers for oneself, or others.

I did not need to drill them. But you mentioning the drilling seems to ring a bell, slightly.
Later it was the pencil trick, for reconnecting the tracks (which I managed to avoid).

The speed when you first play with a BP6 is amazing, because you run, and leave running an application, and the computer carries on working, the mouse moving fluidly, as if the computer was doing nothing.
Because it is running in the other (2nd) processor.
It felt sooooo powerful, a real step forward.

I'm still annoyed with Intel, because they won't let us have multiple cpus per motherboard, for home enthusiasts use (unless you want to spend $1,000's and $1,000's, just on the pair of multi socket enabled xeons from Intel, alone, and the motherboard can costs a fair bit as well).
 

galego

Golden Member
Apr 10, 2013
1,091
0
0
The company protecting their assets against 3rd party vendors who are essentially hacking their products and running them outside of spec.

Intel didn't advertise these chipsets as overclocking boards, the mobo companies did.

If 3rd party vendors were running out of spec, the warranty would be broken, the 3rd party vendors would be losing money with mobo replacements, and Intel would not need to release a microcode to correct the situation.

Here 3rd party vendors found the way to offer an extra to users, whereas running on spec. 3rd party vendors checked that the chipsets run safely before releasing them to the market.

To force users to pay for the extra, Intel is releasing a microcode update to close the hole that they leaved open in the spec.
 
Last edited:

SOFTengCOMPelec

Platinum Member
May 9, 2013
2,417
75
91
I avoided the MS tax by clustering my rigs into a Beowulf using redhat Linux. MS was far more predatory than Intel back then, nowadays Intel has stepped up its game.

Later on, and these days, that is similar to what I would do.
The crazy thing is if you write your own software, these days, there can be a huge amount of work, just to use all the cores and hyper threads available in just one computer, before you even start worrying about multiple computers.
(Size of "huge amount of work" depends on what the project does, for many things, it can be quite quick to use all the cpu power).

I love the relatively limit free Linux, because it does not prevent you connecting many computers together, and networking themselves all day, for many months, without a murmur of trouble.
The older versions of windows, were stretched to keep running for even just one week, without crashing (BSOD etc). E.g. Windows 98 (I think) was notoriously long term unstable.
One of the versions of windows hilariously crashed (BSOD) while Bill Gates was demoing the new windows OS version, to everyone.

In one of your older posts (in the last week or so), you said you crazily went and bought a $1,000 memory expansion (or something), because "you WANTED IT!", in the late 90's.
That is so, so typical of me, I find it very funny.
 
Last edited:
Aug 11, 2008
10,451
642
126
OK, I agree that if I bought a mobo and expected one set of features and found them later disabled, I would be upset.

What I dont see is how this benefits intel, since they dont sell mobos anymore. Seems to me it was more like mobo manufacturers trying to get an edge on each other. I could see how disabling turbo overclocking on non-k chips could benefit intel by forcing users to move up to k model chips, but this applies to "k" chips which were intended to be used on a different motherboard anyway, am I correct?
 

NTMBK

Lifer
Nov 14, 2011
10,454
5,839
136
OK, I agree that if I bought a mobo and expected one set of features and found them later disabled, I would be upset.

What I dont see is how this benefits intel, since they dont sell mobos anymore. Seems to me it was more like mobo manufacturers trying to get an edge on each other. I could see how disabling turbo overclocking on non-k chips could benefit intel by forcing users to move up to k model chips, but this applies to "k" chips which were intended to be used on a different motherboard anyway, am I correct?

Intel still sells the chipsets to the mobo manufacturers, and I expect that they charge more for Z series chipsets over B and H series. Enabling overclocking on B and H cuts into their Z series sales.
 

Atreidin

Senior member
Mar 31, 2011
464
27
86
What a dick move. And people are defending Intel treating us all like crap. Wow. I'm not sure which is worse.
 

KompuKare

Golden Member
Jul 28, 2009
1,228
1,597
136
I think a few years or so, ago, Intel trialled Internet upgradeable processors. You buy a relatively cheap Intel cpu, and have it in your computer. You then choose a better Intel cpu, and buy an upgrade code from Intel, via the internet.
The special code then changes options in the cpu, which improves its performance, and features, etc.

People absolutely HATED the idea, and I think it was a complete and total flop, due to massive unpopularity with customers.

Why do I have this feeling that this idea will be coming back for BGA soldered motherboards soon?

The alternatives are either 1) Intel vastly reducing their SKUs or 2) motherboard manufacturers, distributors and retailers having an inventory management nightmare.
 

piasabird

Lifer
Feb 6, 2002
17,168
60
91
I think it is a good idea because all of the people who overclock processors have a tendency to send in their burned up hardware in and expecting a refund on the hardware they destroyed by overclocking. They just drive up prices for people who do not overclock.
 

TerryMathews

Lifer
Oct 9, 1999
11,464
2
0
To the people defending Intel in this:

I don't believe in any market segment would the end user blame the chip set manufacturer. Not that these boards would've ended up in retail box computers but even if they did the consumer would just blame the manufacturer. Enthusiasts? Blame the mobo maker. I've never seen anyone say "Damn Intel" when their ECS motherboard dies.

This situation seems very equivalent to Sony removing Other OS from the PS3. Which I think most informed people chalked up as an example of an action that was legal but clearly consumer unfriendly.

I can think of a number of situations in the past where Intel has tolerated "off-label" use of their products. PC-DL and NCCH-DL. Dual CPU support in the socketed P3 Celerons.

I wonder what the difference is between then and now? :sly:
 

VirtualLarry

No Lifer
Aug 25, 2001
56,587
10,225
126
A company that makes any profit can be said to be exhibiting anti-consumer behavior. Kinda a pointless tangent you are going on here.
You know that's an untrue hyperbole.

Corporate charters are supposed to be "for the public good".

If they're not - time to dissolve them.

Intel has to deal with warranty issues for its chipsets, what for-profit business wouldn't seek to limit its liabilities?

By "damaging" the capabilities of a product AFTER it has been purchased by the citizen?

Surely you cannot be supporting Intel's actions on this? Intel is clearly playing dirty pool here.

If they go through with this, I foresee big class actions, and possible further FTC sanctions.
 
Last edited:

Charles Kozierok

Elite Member
May 14, 2012
6,762
1
0
Again, unless I'm missing something, this is Intel taking steps to ensure that its products are used as designed. If I'm supposed to get up in arms over that, well, it ain't happening.

People who don't like Intel's decisions in how to position its products have the option of buying someone else's products.
 

SOFTengCOMPelec

Platinum Member
May 9, 2013
2,417
75
91
You know that's an untrue hyperbole.

Corporate charters are supposed to be "for the public good".

If they're not - time to dissolve them.



By "damaging" the capabilities of a product AFTER it has been purchased by the citizen?

Surely you cannot be supporting Intel's actions on this? Intel is clearly playing dirty pool here.

If they go through with this, I foresee big class actions, and possible further FTC sanctions.

Exactly.

Changing a product "AFTER" it has been released to customers, is a MAJOR thing.

It no longer belongs to Intel, it is completely none of their business.

Their interference of it, would surely be a criminal offence in most countries (I think).

The clicking of these crazy legal nonsense things, which can be 20 .. 200 pages long, in tiny writing, which most sane people, click "Yes, I READ THIS (Lies), yes, I agree to it (you would NOT if you had read it). May NOT stand up in court, as far as I am aware.
Even if legally it does stand up, angry customers are something, most sensible companies do not want.

Going for absolute PROFIT, and ignoring all other things, can, sooner or later, go disastrously wrong for the company.

A somewhat secret windows update, which in the background, potentially removes PERMANENTLY a significant amount of the computers performance, could cause Intel major headaches in the longer term.
If things get too scary, I would have thought Microsoft would pull the plug, and not do it for intel, as they could get hit by the cross-fire as well.
 

VirtualLarry

No Lifer
Aug 25, 2001
56,587
10,225
126
I see most of the people in this thread think profit is evil. Lord help you if you ever try to start your own businesses.

Profit is not evil. Aiding and abetting consumer fraud to maintain your profit margin is.

For both you and IDC, I offer this analogy:

It's not wrong to open a lemonade stand.

It's not even wrong to buy up all the lemons in the grocery store, and then offer them to other lemonade stands for a profit.

But sending in armed thugs to rough up the new lemonade stand owner that found an alternative source of lemons out of town is.
 

SOFTengCOMPelec

Platinum Member
May 9, 2013
2,417
75
91
Again, unless I'm missing something, this is Intel taking steps to ensure that its products are used as designed. If I'm supposed to get up in arms over that, well, it ain't happening.

People who don't like Intel's decisions in how to position its products have the option of buying someone else's products.

But the customer paid a small fortune to buy the most expensive cpu in the consumer range (non-extreme 2011) 'K', which is specially for overclockers.

When you overclock, it is at your own risk, anyway.

Intel don't say, you can overclock up to 5GHz, on ALL our chips, guaranteed.

Any overclocking is NOT guaranteed by Intel anyway.

But they do offer a paid for guaranteed replacement service on some/all of their cpus.
 

TerryMathews

Lifer
Oct 9, 1999
11,464
2
0
Again, unless I'm missing something, this is Intel taking steps to ensure that its products are used as designed. If I'm supposed to get up in arms over that, well, it ain't happening.

People who don't like Intel's decisions in how to position its products have the option of buying someone else's products.

Lol.

So should Intel be able to disable anything it wants to that isn't "as designed"?

How about using a laptop CPU in a desktop?

More PCIe video cards than the chipset supports? (Think BTC mining)
 

VirtualLarry

No Lifer
Aug 25, 2001
56,587
10,225
126
If they are disabling overclocking on non-k chips, that is unfortunate, but technically they never advertised them as being overclockable did they?

Did you ever see the Intel video (from the Sandy Bridge era) about how easy it was to overclock, that even Grandma could do it?

Intel certainly advertised overclocking with "K" CPUs with that video.
 

SenK9

Junior Member
Jul 14, 2013
8
0
66
That is a pretty low move. Probably not as low as not having full virtualization capability on K models. If they did I would have gladly spent more on a 4770K instead of a non K.
 

piasabird

Lifer
Feb 6, 2002
17,168
60
91
I have my opinion and you have yours. I have no compassion for overclockers. I think overclockers are the problem, not Intel. However, you can get rediculous about this subject. Sometimes too many limitations on how a builder adjusts things in the BIOS, could result in difficulty in purchasing the correct RAM. Since INTEL started developing chipsets they also made it possible to use a wide variety of RAM for things like low profile and low voltage RAM. I also in the past have purchased a K series CPU for the better integrated video. So I would not want to be told I can not use a K series chip on a specific motherboard but I could live with it if there were retail chips with better video options. You can get kind of rediculous on this.