CNET: Is AMD still relevant?

[peonyu]

High end Cpus, imo AMD is not at all relevant anymore. Its been ages since they truely were relevant - Athlon 64 is what I mean here. And they were in the game with the Phenom by being competitive to an extent against the Intel Core series on price/performance basis.

Bulldozer is just awful however.

AMD's target market is the bottom feeder.

"I just want the cheapest computer that will run WoW and D3 @ 30-40fps"

And that's fine.. that's what they are there for. I just think it's hilarious when people try to act as though they are still somehow competitive with Intel.

The fact is Intel could put AMD out of business. They could easily price their chips even less, and just completely annihilate the price/performance benchmark, more than they do already. Intel *let's* AMD survive.

The problem there, is the I3 is priced pretty damn cheap for anything you look at it - laptop, desktop etc. And it smokes most of the AMD offerings close to it in price. AMD must have lost some of its better engineers after A64 came out for it to be so bad now.

 

[pelov]

Correct. Intel have only been competing with itself for several years now. And the current prices reflect the best volume/margin/profit ratio. So lowering or upping the price will damage revenue and profit.

Same with innovation. Intel will continue so, else you wouldnt buy the next CPU.

They're at the point where they absolutely need AMD, though. Intel wouldn't survive without another competitor in the x86 market, whether due to monopolistic business practices (again) or the market simply not wanting to buy into x86 because there is no other option. Think of it like what AMD did with GloFo. If AMD had kept the contract and didn't use that buyout clause they'd only have GloFo to make their chips and GloFo, though they did have a WSA to abide by, wouldn't care for anything past that because they know they'd get their money. As soon as AMD paid their way out of the contract it opened up their alternatives and GloFo has been forced to step their game up because they're now competing against TSMC/Samsung.

OEMs aren't going to be as willing to pay Intel with the knowledge that they are the only x86 show in town. It's bad business to buy into a monopoly like that. If Intel were to screw up or charge too much or slow down the OEMs would have nowhere else to go. If they were to be the only large x86 maker then you can throw legacy compliance advantage in the trash as it won't mean much. The OEMs would be far more willing to go ARM or MIPS or whatever if it means they aren't relying solely on Intel which has no competition.

Intel absolutely needs AMD to stay alive and barely breathing. That's the best way for Intel to conduct business.

 

[futurefields]

So what happens if AMD goes belly up? Does the government shut down Intel due to the "monopoly" ?

If Intel is the only company that can make good cpu's than so be it. Would be irrational to halt progress because they are a "monopoly" ...

Nobody is stopping AMD from hiring better engineers/making better products.

 

[Schmide]

Monopoly law can be vague at times. By x86 market share alone Intel is probably considered a monopoly now. The rise of arm and other non x86 computing has maybe slightly alleviated this, but they by far do not have free reign to do whatever they want as shown by the settlement of strong arm tactics against oems and AMD. The directive to remove brand check from their compilers also shows this.

The government doesn't shut down monopolies, it regulates them to prevent them from gouging consumers and stifling competition.

When you're as big as Intel you can have more weight on the economy than many small countries. With the cash reserves and reliance on their product, they could easily unfairly remove competitor(s) from the market.

Edit:

Nobody is stopping AMD from hiring better engineers/making better products.

One could say that Intel already did this. At one point in time AMD had a superior product. (slightly) Intel stifled growth by using a quota system to prevent OEMs from using AMD. This lead to a 1.8billion fine in Europe and a 1 billion settlement with AMD last year.

 

[georgec84]

AMD isn't going away.

 

[KompuKare]

One could say that Intel already did this. At one point in time AMD had a superior product. (slightly) Intel stifled growth by using a quota system to prevent OEMs from using AMD. This lead to a 1.8billion fine in Europe and a 1 billion settlement with AMD last year.

And another hypothetical what ifs (maybe not quite there with 'what if IBM had used a 68K in the original PC' but still an interesting question):

What if, back in the Athlon64 vs P4 days, both Intel and AMD had been at the same process node? The strongarm tactics were only one of the things stopping AMD from having a runaway success with Athlon64. The other part was that Intel was able to keep up due to having better fabs.

That's the real hurdle though: even if Intel messed up again and did another P4 (not likely), and if at that time by some miracle AMD had a better design, if Intel has a 2-3 year process advantage, would it matter?

 

[piesquared]

AMD isn't going away.

Correct. In fact, AMD is gaining traction and becoming even more relevant. Trinity and the rest of their APU's are fantastic products that give the consumer the most gratifying experience. And thanks to OpenCl, Direct Compute and C++ AMP, users get the best performance in the majority of their applications, best battery life, along with best graphics performance on the market.

 

[piesquared]

And another hypothetical what ifs (maybe not quite there with 'what if IBM had used a 68K in the original PC' but still an interesting question):

What if, back in the Athlon64 vs P4 days, both Intel and AMD had been at the same process node? The strongarm tactics were only one of the things stopping AMD from having a runaway success with Athlon64. The other part was that Intel was able to keep up due to having better fabs.

That's the real hurdle though: even if Intel messed up again and did another P4 (not likely), and if at that time by some miracle AMD had a better design, if Intel has a 2-3 year process advantage, would it matter?

Process advantage doesn't matter all that much even now. AMD will sell Millions of APU's on 40nm and Millions of APU's on 32nm and Millions of traditional CPU's on 32nm. The process advantage is vastly blown out of proportion.

 

[KompuKare]

Not a matter of deserving, just that the unlocked nature of the core 2 low end was fantastic for OCing consumers and helped Intel wipe out memories of P4. OCing has always been popular in the enthusiast community and by locking down chips Intel has set the minimum entry price to OCing at around $200.

Don't have to have a feeling of entitlement to lament how much headroom is locked away in the SB Celerons and Pentiums.

Ah, good that's pretty much exactly what I meant. Well worded, which is important on this forum. Anyone here who questions the value of Intel's K-series CPUs gets a pounding here. Actually I never really question the value 2500K/2600K vs the $600-$1000 CPUs of yore but I have never been in the market for $200 CPU.

And as I said earlier, if Intel is dependent on a constant upgrade cycle and nothing they have atm appeals to me, I might have to go back to AMD. Actually, my current CPU is 'fast enought'(TM) so idle power is possible equally important and Trinity does very well there.

 

[Makaveli]

Consumers could care less about benchmarks, they are meaningless. They're only relevant to the shrinking online benchmarking addicts. Consumers care about user experience and how they interact with their devices on a daily bases. If their system gives them off the chart SuperPi numbers but stumbles and stutters in a game they like to play, do you honestly believe they would give a rats ass about the benchmark? If benchmarking is your holy grail, run some OpenCl or C++ AMP compiled benchmarks. Benchmarking can give you any number of different results.

I do have to agree with this.

Most people don't care, don't even know how to read a benchmark properly and just need stuff that works. This is the general public and most of the sheep apple calls customers. And sadly they out number us 1000:1 if people in general cared more about how stuff works and why it does what it does we would probably see better products out there and less crap and knock offs.

As they say ignorance is bliss!

 

[sm625]

Intel may be selling chips for less than they were 10 years ago, but dont fool yourself, they arent any cheaper. Consider this image:

Would that be a good deal if it was sold as a 99 cent value meal? You cant use the "oh but the tech is better" argument because the tech is better in a lot of things and they've all dropped substantially in price.

 

[Ferzerp]

Wow, totally grasping at straws now, aren't you?

The value of a processor is based on its physical size? That's a good one.

 

[Nemesis 1]

Intel may be selling chips for less than they were 10 years ago, but dont fool yourself, they arent any cheaper. Consider this image:

Would that be a good deal if it was sold as a 99 cent value meal? You cant use the "oh but the tech is better" argument because the tech is better in a lot of things and they've all dropped substantially in price.

Really . The fastest Price Drop I ever seen was in 2006. When intel introduced core. AMD went from overpriced CPUs to bargain basement as intel introduced core at bargain pricies compared to AMD overpriced hotrod

 

[VirtualLarry]

Process advantage doesn't matter all that much even now. AMD will sell Millions of APU's on 40nm and Millions of APU's on 32nm and Millions of traditional CPU's on 32nm. The process advantage is vastly blown out of proportion.

Oh really? What about BD's lackluster performance, because AMD couldn't get them to clock high enough at launch? Not to mention, their high power consumption.

Also, AMD could make a higher profit per chip, if they were using a more advanced fab process, such that they could fit more chips per wafer.

 

[piesquared]

Oh really? What about BD's lackluster performance, because AMD couldn't get them to clock high enough at launch? Not to mention, their high power consumption.

Also, AMD could make a higher profit per chip, if they were using a more advanced fab process, such that they could fit more chips per wafer.

You can't ignore the cost of Fabs and bleeding edge wafers. It's not as simple as you're suggesting but i'm sure you know that.