Intel stuck with $1.45 billion fine in Europe for unfair and damaging practices against AMD

Pretty sickening what Intel did to control the market and move it away from AMD

Intel stuck with $1.45 billion fine in Europe for unfair and damaging practices against AMD | ExtremeTech

 

Pretty sickening but you had to know something was going on. And all the sub 35w AMD CPUs going in massive chassis notebooks.

Beamed from my G2 Tricorder

 

Sickening? I don't know...

Just politics with $$$$$$$$$$$$$$$$$ instead of votes (to get to the public $$$$$$$$$$$$$$$$$).

As a consumer, I've always bought whatever improved from where I was previously. AMD was never in the running for my workflows.

 

Intel currently has a policy of subsidizing computer manufacturers' advertisements... IF they feature Intel's processor.

Brilliant marketing idea. Too bad AMD is too cash-starved to do the same thing.

 

If only AMD had a true competing CPU...

 

... so how much of this 1.45 billion goes back to AMD? None?

 

As this is 'damages' against AMD, I would assume (almost) all of it.

Money can't save AMD though; as the article states in the first post, Intel is years ahead and time is what AMD can't make up (easily).

 

If only AMD CPU's were as fast as Intel, battery efficiency as good, and they were offered in better grades of laptop than they usually are.

 

once upon a time when Netbust was sucking hard, AMD managed to capture 28% or so of the market share with the Athlon...

 

Yes, but not today or recently.

 

It would be OK if Intel told OEMs and hardware stores that if they buy big quantities of CPUs from them, they would get a discount on all the CPUs. Tons of companies does this. But Intel went way further, they said that they would get refund only if the OEMs and hardware shops stopped selling AMD CPUs.
It is directly anti competition and only profit the OEMs and shops because they get greater margins on the hardware they sell and put in the machines. As we have all seen, the Intel CPUs sold on the internet, the complete desktop systems you buy, they are all bloody expensive, which means the companies are the ones who net a good deal of money on this.

I also see people mention that AMD CPUs are sub par compared to Intel, maybe the money AMD should have gotten back in the days when Intel first sleazed their way to getting more sales, would have gone to R&D for AMDs CPU department? Which could have resulted in better products that could compete against Intel. Intel have a big foundry, one of the best in the industry if not the best, how much of the cost to operate the plant, comes from dirty money?

 

I wonder about this too. As I said above, at one point in the past AMD was very competitive and a threat to Intel. Really makes you wonder if there's any correlation between the start of Intel's dirty practices and AMD's decline.

 

Well, if not CPU front, but AMD had superior integrated graphics that Intel could not match up even until nowadays. These APUs with strong graphics could have been the bestbuy thin 12-14" laptops if manufacturers are making it, unfortunately we did not get any.

 

Well you can start to take a look on Anandtech`s review of AMD`s Athlon64 vs Intel`s Pentium 4.
Its the complete opposite of todays situation.
AnandTech | AMD Athlon 64 4000+ & FX-55: A Thorough Investigation

From the article in OP

So yeah I do wonder if the dirty play from Intel allowed them to crush AMD in R&D which eventually made Intel`s Core processors happen while AMD sadly didnt have the fund to keep on developing on full force. Of course, just speculation, and that 1.4 billion dollars is fractions of what AMD really should have gotten paid in fine today. Its sad that AMD doesnt even receive the money Intel is fined with

 

This is what I was going to say, and believe I've said it many times in the past, that there was some paying off OEM's and stores to prevent AMD products from getting on the market. The lack of that money from the get-go had a significant impact on their R&D. How can you compete when your competition pays off the retailers that will sell your products. The sad thing it isn't only Intel at fault but the OEM's and retailers that turned a blind eye to the practice.

It will still take a miracle to get Intel to pony up that money to AMD, let alone the fact that Intel had a net revenue in 2013 of over $52 Billion will hardly make a dent. They will just raise prices by 5% or more and we will pay for it in the end. I think Intel should be fined $10 Billion and be forced to sell chips at cost for a year, because I'm sure AMD has lost more than $1.5B if you take into account lost market share and R&D opportunity.

 

Sorry HT, but that is just not how it works. Especially on a decade old complaint. Especially on the second 'judgment' of over $1B...

And it has been shown (here?) that AMD won't receive the damages assessed? If not; just one thief stealing from another (EU vs. Intel...), again; politics with a capital $$$$$$$$$$$$$$$$$.


As for the article Cloudfire linked a couple of posts up; AMD was the more expensive option back then too. There is no 'good guy'... there is only what is 'good enough, right now' for consumers.


Intel may not have done the right thing (legally), but neither does any other company if they think they have a chance to get away with it.

Don't forget that the computer manufacturers were just as guilty in my view. It takes at least two to tango and in this case; it was an industry tangle.

 

HP started selling 4 devices with Beema A6 APUs

The catch?

They're in laptops. Specifically, 15.6" and 17" 768p laptops: Shop AMD


On a side note, Acer put the A10-7300 and a R7 M265 (which is most likely going to be bottlenecked by the CPU) in a 15.6" 768p laptop: Acer | Aspire E | Aspire E5-551G-T0JN | Overview


The only recent Kaveri laptops that are interesting are the Elitebook 700s. The only issue is that to upgrade an Elitebook 745 G2 from a 1600x900 resolution to 1080p, you have to go to the customizable models, which will automatically hike the costs from $929 to $1690.

Nearly $800 for a 900p to 1080p panel upgrade.

The Elitebook 725s and 755s are also quite expensive once you start to upgrade from the 768p base model. Such as charging nearly $100 for an 128GB SSD.

The Elitebook 700s are also limited to a 3-cell 50 WHr internal battery. You have to pay $219 for an external 60 WHr battery.

 

That was my point too. They were never integrated into thin and light form factors that Intel had with even higher TDP's.

 

To those who have said that AMD may have done better if manufacturers had put their chips in more attractive form factors - that may have been prevented precisely by Intel paying off manufacturers to not ship AMD. Some they paid to not ship AMD entirely, but it's entirely possible that they also paid (or even are still paying) manufacturers to not put AMD chips in certain form factors, while allowing them to ship some AMD chips so it doesn't look like an exclusivity agreement.

And if you think it's unplausible that Intel would pay to get exclusivity in certain segments, just look at what they did with the business market. Even today, you don't see many AMD chips in business environments, and it's not just because the old "you can't get fired for buying IBM" now applies to Intel.

Ultimately, this is part of the reason that I plan to buy AMD CPUs in the future. The other part is that I want there to still be at least two big x86 CPU companies, and AMD needs all the support it can get currently. In 2011, it still looked like Piledriver might be pretty good, so I was OK going with the i5-2500k and Radeon GPU at that point.

Also noteworthy is that Intel actually paid this fine in full in 2009 (according to their own rep, quoted by Ars Technica), so what this really means is they won't be getting that money back, unless they win on further appeal, not that they are paying that this year. I don't think they have much chance of winning on further appeal.

 

I just don't think the fine is stiff enough, and the money needs to go to AMD, at least a large portion of it. To pay $1.5B for 4-5 years of garnering the advantage to result in $50B+ in profits? That's a small price to pay to gain the exorbitantly competitive edge over AMD. The fines need to be substantial enough so that they won't do it again, not something that they can just write a check for and wipe their hands clean.

 

Well short of liquidating Intel's assets, I don't think the damage could ever be undone. Those 5 years have basically crippled AMD's R&D program, and ensured that AMD will always play second fiddle to Intel. Even if all that $50Bn of dirty money went to AMD, they're just far too behind at this point in the game to catch up with Intel.

...and typing all that just made my blood boil a little bit.

 

Awhile ago, a major banking company was fined close to $1 billion for being too risky with their trading.

They then wrote off a large portion of the fine as tax deductibles by using some loopholes in the tax codes.

Derp.

 

In the early days I always built desktops with AMD CPUs because they offered better performance than Intel while still being cheaper.
I would buy notebooks with AMD APUs too because they are at 35W vs Intel`s 45W, which will reduce heat. But the recent reviews of AMD APUs and 8970M results in heavy bottlenecks which lower the total performance down to a midrange GPU, makes me hesitate.
What AMD should do is reduce the IGP part of the APU down to minimal (enough to support 1080p decoding) and increase core clocks for the CPU and sell that APU for gaming notebooks with dedicated graphic cards. While selling the APUs with better IGP for ultrabooks and notebooks without dGPU. That should be enough to remove bottlenecks with high end mobile GPUs and being able to cater for all sorts of notebooks.

Then they can start putting pressure on the notebook OEMs to feature AMD APUs in the high end notebooks from Dell/Clevo etc. I can understand a dilemma will occur by that time for the OEMs, because the same model with Intel processors and AMD processors would mean two different motherboards. For OEMs its so much easier to just pick the strongest one of them and stick with it.

The whole thing is complicated but I would support AMD if I had the option.

 

They won't be in high end laptops until they get High End Processing Power, the integrated graphics are great. They need to be as cool operating as Intel and have comparable battery efficiency to be in anything but value laptops. In short, equal to Intel except for the superior integrated graphics.

 

I really wish there were some AMD-based laptops that were of decent quality. I see AMD products I like, just no finished laptops that appeal to me that include AMD processors. If I found one, I would snatch it up in an instant.

AMD was cheaper than Intel, always. During the whole fiasco, Intel barely budged from their prices, and actually reacted to the Athlon 64 FX by releasing the far more expensive yet far weaker Pentium 4 Extreme Edition. The only reason people bought those things were to have clockspeed bragging rights.

When AMD's license to produce x86 processors was going to expire, Intel fought tooth and nail not to renew it in order to completely kill off AMD. They were court-ordered to renew the license. A few years later they came back to AMD begging for the license to AMD64...

I agree. Very guilty.

 

Qing Dao,

See:
AnandTech | AMD Athlon 64 4000+ & FX-55: A Thorough Investigation

 

That is incorrect. The mistakenly say that the Pentium 4 560 was the flagship desktop processor at the time, but the $1000 Pentium 4 Extreme Edition actually was.

Then you can even look at the benchmarks and see that at every price point, the AMD processors performed much better. There is simply no way one can say that the AMD processors cost more.

 

Back then, AMD was on top. Not currently. Their best common processor is equivalent to roughly an Intel Core I3.

 

Ahhh the P4, precious memories. I had a 3.2ghz unit with Hyperthreading. What a turd it was. Eventually fried it with too much vcore.
I've never had the pleasure of using the Athlon chips but the Pentium 4 was dreadful regardless of how many GHz it could push.
AMD had a winner and sadly wasn't allowed to capitalise.

 

AMD had something with the Llano mobile APU's. Then I don't know what happened. Trinity was supposed to build on that. They had unlocked TDP, clock speed, and voltage. You could tune it to your heart's content and the higher end 45W APU's could compete with an i5 and even a quad i7 in some cases: http://forum.notebookreview.com/har.../652836-intel-beat-out-amd-4.html#post8401304

 

According to Anandtech, Trinity had a 10% increase in multi threaded and 20% increase in single threaded. Intel had a 20% increase in multi and 10% increase in single with Ivy bridge vs Sandy.
If we look at Kaveri top dog FX-7600P vs Richland A10-5750M, we have a 3.6% increase in single threaded and 10% in multi. Kaveri = Steamroller architecture. Richland = Piledriver architecture. So its two different architectures.
Compare Intel`s Haswell architecture against Sandy Bridge architecture, we get a 6% increase in single threaded and 4% increase in multi.

So both companies have had a really horrible development on the CPU side. All thanks to the stupid APU/IGP race. I`m not very interested in the upcoming Broadwell CPU from Intel either because I know its still using the Haswell architecture. Even when Skylake architecture comes to take over for Haswell, I have a big feeling the CPU improvements will be extremely dissappointing.

I truly wish Kaveri doesnt bottleneck the R9 M290X and the R9 M295X, which we will see when MSI makes a notebook with those two. Because if it doesnt bottleneck, that notebook can finally be a valid alternative against the expensive Intel + R9 M290X/M295X. And who knows, maybe more OEMs will finally open their eyes for AMD alternatives in their notebooks.
One thing is that OEMs traditionally pick Intel CPUs for their notebooks out of a safe habit since Intel`s dirty tactics began, but I also think there are some internal testing done on past AMD APUs with the dGPUs and OEMs just couldnt approve the usage of the AMD APUs because of the poor CPU performance. They can`t risk their reputation by selling a notebook that doesnt operate optimal like it should.

 

Kaveri will definitely bottleneck any high end mobile GPUs, even the FX-7600p, unless if you have a game that is very GPU limited.

I would prefer if manufacturers release some decent 1080p laptops that meet Ultrabook specs but have AMD APUs. HP's Elitebook 700s are the only ones that meet the ultrabook spec, but they charge nearly $800 for an upgrade from 900p to 1080p.

 

Sounds like a lot of assumption there. We wont really know until anyone does a review
Yeah, would be excellent to see the APUs with midrange GPUs as well. Not just AMD`s GPUs but also Nvidia. And ultrabooks of course. 35W is not too high heat and we know the APU wont bottleneck anything there

 

Well AMD was shaping up to something exciting for consumers. Basically a thin and light notebook with a max 45W TDP APU that was gaming capable and not cost a fortune. Intel made some fairly decent improvements with their IGP, but their IGP also throttles on their lower TDP CPU's. Vizio is the only one that did anything decent with them, and they are not notebook manufacturers, but the result was a reasonably affordable 14" notebook with a 45W TDP AMD APU.

 

Wasn't the deal they made something like they would cross-license all of their patents? I wonder what would've happened if AMD had decided not to license AMD64? Obviously some lawsuits and Intel trying to revoke the x86 license, but Itanium was clearly not taking over the mainstream by then, so it would've been interesting to see whether AMD64 still took off if AMD kept it exclusive. Intel probably would've made their own, different backward-compatible x86 extension, like they did with Vanderpool for visualization technology. But then what would have Microsoft done? Release Windows for both AMD64 and Intel64 and require programs to be recompiled for both? Or just stuck with Itanium and AMD64, since AMD64 came first? I'm sure Microsoft would not have been thrilled with the prospect of mainstream 64-bit programs compiled for AMD not working on Intel and vice versa. Acceptable perhaps with Itanium in 1999, when it was supposed to be the x86 successor and they had x86 emulation for it, but not good if there are two incompatible x86 successors. And by the time AMD64 did come in reality, it was clear Itanium would not in fact be the x86 successor, so there was still just one successor.

An interesting hypothetical, and probably even more interesting if Longhorn had shipped on time and it hadn't taken until XP x64 for their to be a somewhat-mainstream x64 Windows option.

Ah yes, I also had a Pentium 4, though 2.66 GHz without Hyperthreading. Still fast compared to the Pentium II it succeeded, but had I known as much about technology in 2003 as I did in 2007, I probably would've wound up with an Athlon and a video card that wasn't a GeForce4 MX.

Where exactly are you getting your numbers from? You're saying that Ivy Bridge is 20% faster in multithreaded vs. Sandy Bridge, which is being generous from what I usually hear. But then you say Haswell is only 4% faster than Sandy Bridge in multi-threaded. If both are true, that means that Haswell is about 13% slower than Ivy Bridge in multithreaded. I know Haswell is not exactly a roaring success from the performance front, but I'm pretty sure that it's not 13% slower than Ivy Bridge on average.

The FX-7600P vs. A10-5750M difference is also lower than what I've typically been hearing for the CPU, which is closer to 20% faster versus Richland. Granted, there aren't that many reviews of Kaveri yet, and it varies a fair amount by benchmark. But these numbers seem to be unduly harsh to Kaveri and Haswell, and pretty generous to Ivy Bridge. I tend to hear 5-10% difference for Ivy Bridge, and often inferior overclocking compared to Sandy Bridge - and that is why I decided it would be silly to spend money on upgrading my desktop beyond the i5-2500k. (Haswell would require a new motherboard, so it's even farther out of the picture given its meager benefits)

You are likely right that the FX-7600P will still bottleneck the high-end Radeons, which is unfortunate. Although I think it's clear that AMD isn't really trying to make the fastest CPU out there. While I would've loved to see Kaveri have a 50% faster CPU than Richland, if it really is about 20%, that's still a better increase than we've seen from any x86 manufacturer in years. So it's hard to complain, particularly with the IGP benefits. And while I'd also like to see AMD and Intel release more pure-CPU parts for those of use who are more than okay with dedicated cards, realistically, the single-threaded benefits of that would likely be pretty small. They couldn't just jack up the clocks to 6 GHz by leaving off the IGP.

What would be more realistic would be a quad-module (8-integer-core) CPU, with likely a somewhat higher price and die size than the FX-7600P, and similar clocks. Which wouldn't solve single-threaded bottlenecks. Although it'd still be pretty sweet if multi-threaded CPU tasks were your bottleneck. Kind of like how Vishera is a good option of the desktop if heavily-multithreaded workloads are your problem and you don't want to pony up for Intel's more-than-four-core CPUs.

Edit: They could also, say, make a three-module (6-integer-core) part without the GPU, and with more transistors dedicated to branch prediction and other tasks that increase the IPC (instructions per cycle). But aside from improvements like increasing cache sizes, increasing the IPC would require significant R&D resources, and any low-hanging-fruit has likely already been picked. AMD doesn't really have the R&D budget for such low-volume products that wouldn't necessarily be a success anyway (unlike the custom silicon for consoles, where they have a pretty-much-guaranteed return, even if with low margins).

 

Lol yeah, who didn't get burnt by the infamous P4 + Geforce 4 MX combo. Good lord, the Geforce 4 MX was shocking, it was so bad beyond belief. For me, the fail-train continued to the Geforce 6200 LE.

 

P4 Northwood + GF4 MX440 here at one point. Yep, count me in as victim of that scam too.

Good thing I learned from that debacle, and every hardware upgrade since then has been sensible and a decent step up from before. 9800 Pro, 7800 GT, 8800 GT 512MB, 5870 (fried from insane OC/OV 24/7) -> 6950 (RMA), 7950. P4 Prescott, Athlon 64 X2 4400+, C2Q Q9300, i5-2500K.

Well, everything except the Prescott, which wasn't much faster than the Northwood it replaced and ran like a furnace. But in my defense, that was given to me free by a friend who upgraded.

And come to think of it, the Northwood core wasn't half-bad as Pentium 4, being competitive with the Athlon XP of the time and coming before NetBurst achieved the infamy we remember it for today. And it was a massive step up from the AMD K6-2 I had before that.

But you, on the other hand...that 6200 LE...what were you thinking?!?

 

Is it almost as bad as buying a laptop with a GT 610m paired with an Ivy Bridge i7 (35W TDP)?

My dad almost did that because he noticed it had 2GB of VRAM. I explained to him that it was like installing 16GB of RAM on an 2007 Intel Atom netbook. Aka, the CPU is going to choke long before the RAM is even close to be used up.

 

Uber low-end cards are uber low-end cards, so I guess so.

 

I was really late to the Intel game. My dad bought his first Intel CPU at the end of 2007. I bought my first Intel CPU in August 2008.

The Geforce2 MX was my introduction to Nvidia after being solely with 3dfx previously on Macs and PC's. It did just fine playing games for several years. In my first self-funded PC build, I went with an nForce2 motherboard with GeForce4MX IGP. It was awesome actually. I don't think there has ever been an IGP as relatively powerful (compared to discreet cards) as that one. I think the problem was that they kept selling the GeForce4 MX for several more years as their entry-level card, but when it was new, it had 50% of the performance of the top-level card.

 

The problem with the GeForce 4 MX was that they were GeForce 4 in name only and lacked the D3D8 programmable pixel and vertex shaders of the GF3 and GF4 Ti, which was huge during a time when games using these new features were just beginning to take off. Even in fixed function games, the GF4 MX performed well behind the the GF3 and GF4 Ti, although it was a nice speed bump from the GF2 MX it was based off of. Hell, IIRC in D3D7 games it even beat out the FX 5200, another awful card (just like the rest of the FX series) that I had the misfortune of using for a short period of time before returning it for the monster that was the Radeon 9800 Pro.

 

yeah but theres a slim difference between Uber low end and just bad. The AMD 5570 is uber low end but a great design. The 6200LE was just a terrible design, way too unbalanced and anemic to be useful compared to plain IGP. It was probably cobbled together from the absolute scraps left over from making the actual 6200. Worst $70 I've ever spent.

Oh I forgot to mention, I still didn't learn my lesson until my first laptop, the venerable HP DV6-1045ee. God, that monstrosity weighed in at 3.5kg and had the GPU muscle to match...a Geforce 9200M GS........with DDR2, DDR2???!!!

After that fiasco, I vowed to completely master the realm of technology as so I'm never burnt again.

 

Well there are always some decent entry-level gaming GPU's <$100, but I'm talking about the bottom-of-the-barrel cards that are outperformed by IGP's and completely worthless from a price/performance standpoint no matter how cheap. You know, the behemoths of Nvidia and similar that are essentially rebadges of the same trash generation after generation that OEM's stick in office machines just so they can advertise about having a dGPU and stick a cute Nvidia/AMD sticker on the front panel. I've literally seen order-of-magnitude performance increases from just $10 price difference in that <$100 price bracket, I kid you not.

 

LOL, epic...

 

Here's something to read: ASUS A73 Series A73SD-TS72 Notebook Intel Core i7 2670QM (2.20GHz) 8GB Memory 750GB HDD NVIDIA GeForce GT 610M 17.3" Windows 7 Home Premium 64-Bit - Newegg.com

ASUS A73 Series A73SD-TS72

i7 2670QM (2.20GHz), 17.3", GT 610M (also has 2GB of VRAM, but Newegg doesn't mention that)

In fact, the customer reviews for the 15.6" cousin are relatively positive: ASUS A53SD-ES71 Notebook Intel Core i7 2670QM (2.20GHz) 6GB Memory 750GB HDD NVIDIA GeForce GT 610M 15.6" Windows 7 Home Premium 64-Bit - Newegg.com

Some of the reviews: "My boyfriend bought this for me as a gaming laptop, and I have to say, though it won't run them on all high settings, it still runs them extremely well!"

"Very quick start up, less than half a minute. Runs many games well including Team Fortress Two online. Does not overheat even when running for 5+ hours."

And this is why manufacturers occasionally get away with crap designs. Why stop doing it when most of your customers are taking it hook, sinker and line?