Intel CEO: AMDs time is soon over

Imagine calling someone a liar while using a source for a competing company instead of a source from the company the person was originally talking about.

July 23, 2021:


Intel =/= AMD

Even if the semi-conductor shortage is resolved late 2022 for AMD, that doesn't mean that Intel is going to be on that same time frame. Different companies have different contacts, different management, different resource usage, and different levels of efficiency. This comment was not the slam-dunk that you thought it was. Especially considering that this is just a giant guess for anyone, including AMD and Intel. AMD's CEO was likely painting a best-case-scenario, and Intel's is likely more realistic. We can't know for sure, but calling someone a liar based on the opinion of a CEO from a different company altogether was bad form.

How did I call them a liar? Their source was from April, and Intel got a lot of their fab issues ironed out.
 
Intel without competition became stagnant and lazy.
If it wasn't for AMD putting some pressure on them, we would still be stuck with 14nm++++, 4c8t for an i7, and no IPC gains for a whole decadde.
History repeats. It was the same with Athlon 64 and Pentium 4.
 
History repeats. It was the same with Athlon 64 and Pentium 4.

Not really. The Pentium 4 was an attempt at changing the CPU paradigm.
It traded IPC for higher clock speeds. But at the time no one really new how it would pan out.
Today, pretty much all CPU archs use much shorter pipelines, than the Pentium 4. In part, because everyone saw how big of a failure it was.
Even Intel, a few years after the end of the Pentium 4, admitted it was one of their worst archs.
In truth, the P4 was not a matter of laziness and stagnation, quite the opposite, it was a strong push in a very different direction.
But it ended up being the wrong decision. Today we have the power of hindsight, but that's something that not everyone knew.

Imagine that there was a time when people were arguing whether CISC or RISC was the better option.
Today that is settled, and all major CPU archs use RISC. Even X86, that started as CISC, when the P6 was released, it became a RISC with a beefy decoder unit.
 
I think Apple is pushing Intel just as much as AMD is.

Apple doesn't have an X86 license. And this is still the stronghold of Intel.
PCs with Apple ARM CPUs, are a only few % of the market.
The only company that competes directly with Intel on X86, is AMD.
 
Intels complete lack of innovation on the x86 has made that platform so bad Apple jumped ship. Jesus christ, nothing fucking happened in 10 years. I'll believe it when I see it but at this point I think they would have to do something extraordinary.
 
Intels complete lack of innovation on the x86 has made that platform so bad Apple jumped ship. Jesus christ, nothing fucking happened in 10 years. I'll believe it when I see it but at this point I think they would have to do something extraordinary.

It was worse than that. The bugs and security problems were the breaking point for Apple.
Reports from ex-Intel employees said that at a certain point, Apple was filling half of all bugs and security issues with the Core family of CPUs.
It's true that Intel has been almost stagnant on the IPC front, for several years.
But they were patching up a ton of holes.
Now Apple controls it's own silicon, and this gives them much better control over features, performance and security.
 
It was worse than that. The bugs and security problems were the breaking point for Apple.
Reports from ex-Intel employees said that at a certain point, Apple was filling half of all bugs and security issues with the Core family of CPUs.
It's true that Intel has been almost stagnant on the IPC front, for several years.
But they were patching up a ton of holes.
Now Apple controls it's own silicon, and this gives them much better control over features, performance and security.
Aye, I think this has made it hard for Intel to attract good people as well, Apples own silicon department seem to have som real talent - going from the M1 it'll be interesting to see if RISC could make a real comeback in the desktop performance space. The M1 has equal performance to Apples previous i7 CPU Macbooks with less than 30% of the power envelope. Anyway, anything like that would be far off in the future.

I'm just kinda mad at Intel how they pretty much halted the development of high performance CPUs for so long.
 
How did I call them a liar? Their source was from April, and Intel got a lot of their fab issues ironed out.

They said something. You replied with, "Untrue." You can quibble over semantics all you'd like, but that's the same as calling them a liar. Also, you're going to need a source for the claim that Intel has their fab issues ironed out, because the latest source I could find was only 2.5 months ago, and in that source (which I posted above) Intel's CEO was still saying that this supply issue can go into 2023.
 
A confident one. Leaked benches, if real, have been suggesting a +40% advantage to Intel Vs. the 5950X which would be one of (if not THE) largest current-Vs-current gaps in history. A Zen 3+ refresh was rumored to head off these Intel chips, but there was talk of cancelation due to performance gains being uselessly disappointing (single digits), so I'm not sure where that stands now. A genuine Zen 4 could at best be rushed to land Q3 or Q4 2022, but achieving +40% over Zen 3 would require nothing short of a miracle, nevermind beating the 12XXX's. AMD rolled a pair of 6's for 8 years in a row with Intel being stuck on 14nm, and they still lost decidedly every single year save for the 5XXX's.

The "leaks" are contradictory.
As much as there is gains there may also be regressions. There's talk that Alder Lake will not be a winner in gaming.

I think AMD just need to pay a bit of attention to gaming, and they're doing this with Ryzen 3D, because for everything else they're very well.
Intel situation in servers and Exascale will not improve soon. Their next big thing, Sapphire Rapids can't compete with Zen 3 in performance and uses much more power (which means that it's hundreds of dollars more expensive to operate) and AMD is already putting the finishing touches on Zen 4.

Intel is on a hard position.
They will have better products to sell, but that use too much power making them unattractive on serves and too expensive for average consumers with their early adoption of DDR5.
 
I wrote off Intel and it's lackluster CPUs when I switched to my 3900x 2 years ago..

I have only 1 remaining laptop with an Intel CPU in it that's 3+ years old ...

Talk is cheap - Intel needs to put out a product that actually impresses and is remotely competitive before smack-talking about their competitors....
 
Not really. The Pentium 4 was an attempt at changing the CPU paradigm.
It traded IPC for higher clock speeds. But at the time no one really new how it would pan out.
Today, pretty much all CPU archs use much shorter pipelines, than the Pentium 4. In part, because everyone saw how big of a failure it was.
Even Intel, a few years after the end of the Pentium 4, admitted it was one of their worst archs.
In truth, the P4 was not a matter of laziness and stagnation, quite the opposite, it was a strong push in a very different direction.
But it ended up being the wrong decision. Today we have the power of hindsight, but that's something that not everyone knew.

Imagine that there was a time when people were arguing whether CISC or RISC was the better option.
Today that is settled, and all major CPU archs use RISC. Even X86, that started as CISC, when the P6 was released, it became a RISC with a beefy decoder unit.
Still it could be seen as a move of arrogance by Intel. A stupid idea they wouldn't have done if they hadn't been so full of their own success.
 
If I was an engineering manager at AMD I'd be using this in my next standup/team meeting.

Firstly, it shows recognition of the teams hard work that their main competitor is saying this! Which is always nice for a team to hear how good they have been performing.

Secondly, I'd be using it to drive them forward as a motivational tool.

As others have mentioned competition between these two is good for us consumers, the stagnation last decade was annoying.
 
Competition is good.

That being said, are price branded chips still basically similar in performance? The 11600k is basically on par with the 5600x, costs less, but uses more power.

I don't really get why people get so obsessive over CPUs.
The i5 11400 can actually outperform the 5600x and it's $125 cheaper. AMD is just way better at marketing its products. Can't wait to see what Intel's next batch can do
 
Not quite, the 3600x still outperforms it.

5600x is still the king of gaming, but we'll see what Intel has up its sleeve.

I hope my next PC in 5 years or so is all Intel/American.

5600x is the best card right now, for sure, but for the MRP, the 11400F was the best for the value. Didn't realize the price on it skyrocketed though! They should rename it the SkyRocket Lake
 
Still it could be seen as a move of arrogance by Intel. A stupid idea they wouldn't have done if they hadn't been so full of their own success.

I have to disagree. At the time no one knew how clock speeds would ramp up. We have the great power of hind sight, but they didn't, at that time and place.
Intel had some of the best engineers of the world, and still they thought the Pentium 4 could reach 10Ghz.
So they plowed in, with new fabrication processes and increasing the length of the pipeline.
But it didn't scale well, and the reason everyone now knows that, is because Intel tried it.
 
Top Bottom