AMD: Partner meeting on April 23 in preparation of Navi and Ryzen 3000 CPUs launch

insider info of upcoming partner meeting:

AMD: We about to rekt Intel!
Partners: Sounds good!

 

Meanwhile, Intel be like

Spoiler

 

[

 

I saw that last month AMD put many Guides and marketing and Training material for it's partners so they can learn how AMD's product compare to Intel and Nvidia [CPU/GPU]. AMD investing now in it's partners to push sales higher.
https://www.reddit.com/r/Amd/comments/bahbzq/mindfactory_sales_data_for_march_2019/

 

Very nice. The overall sales are way over the 2/3-1/3 line.

 

And Germans love good engineered products

 

BRING IT ON!! I need info.......GIMME GIMME.....GIMME!!.!.!.!.

 

I kind of wonder what AMDs long term plan against Intel is. The core count increase was a nice "surprise" but it's really only going to last so long. With Intel's recent restructuring of it's design team and massive capex increase - how is AMD going to sustain it's momentum against a company that has 10x the revenue and significantly more clout in the industry?

 

The key is to keep spiting out meaningful updates at affordable prices.
Intel will have to spend big on development to cut off AMD progress, meanwhile AMD as a company is growing and gaining market and mindshare.

Moore's law has died, it has gotten difficult and expensive to keep pushing in the same direction.
AMD went the chiplet/glue route, and Intel will follow it.

What will be next? 3D stacking? New ways of arranging cores? Put DDR inside the package? Who knows?

 

I've wondered the same thing. Right now, AMD is in a very good position since they've basically proved to everyone that Intel has been ripping people off and was deliberately holding back, but as you alluded to, once Intel releases something to replace the aging Core i# architecture, I'm not sure how AMD is supposed to respond to that (assuming it is actually a substantially better replacement).

On the other hand, there's only so much money can buy. We're reaching the limits of physics for silicon transistors. Case in point: Intel has been very slow to get 10nm working adequately. Although I think there are a number of ways x86 could be improved upon, at this point, that would require significant changes to software and operating systems, which Intel traditionally avoids wherever possible.

 

That can't be true, they make Mercedes....

 

this is the more-than-$64,000 question.

but AMD has agility going for it and the most nimble move was actually the one i criticized the most - selling the fabs.
now that Intel has discovered the joys of contract work (an increasing part of it's product offerings) it has suffered from supply chain issues as well as node issues. Apple is even questioning the use and availability of Intel 5g modems. and it was the (4g) modems which caused a delay in the current Intel lineup.
Intel is and will continue to be the market leader, but by steadily decreasing margins for the next 3-4 years. after that, if AMD doesn't have its ducks in a row, there'll be trouble.

 

Haha Apple is questioning their continued use with Intel, period. There's rumors that they're going to transition entirely to ARM.
Though, if Apple ceased to exist tomorrow, I don't think that would really impact Intel all that much. Apple and Intel, to my understanding, never really got along that well because Apple is very demanding and Intel doesn't have enough of an incentive to cater to their demands.

 

By simply keeping the R&D pedal to the metal at all times. Having 10x the revenue (Intel is highly diversified outside of the PC arena, remember) has never been a defense against being bested by AMD technically. But this time at AMD, unlike the two-year period years back when AMD was besting Intel with it's A64 line, while Intel was trying to push Itanium and keep everyone else on a 32-bit desktop (I will never forget Intel's infamous, "You don't need 64-bits on the desktop!" ad campaign targeting the A-64), AMD is being run by the right people! Su has one foot in engineering, the other in the retail markets--she's a huge boon for AMD. After the A64, after Intel licensed x86-64 from AMD and proceeded to beat AMD at its own game, AMD went through a succession of CEO's without a clue, literally. Reminded me of C= a lot during those years. All of that is history, now, thankfully--or I don't think there would be an AMD around to talk about! One last comment: Intel reminds me a lot of Apple corporate--Intel really is not accustomed to having to compete with anyone except AMD, and the last time that happened was many years ago. Intel is huge and extremely cumbersome--will let go 15,000 employees in a year and not feel it, etc. AMD is primed, otoh, leaner and more nimble competitively and far less weighted down by other-than-PC concerns. As far as "clout" goes, Intel doesn't have as much as you might imagine. When AMD introduced the original Athlon, Intel pulled out all of the stops and dirty tricks it could manage to keep OEM motherboard makers away from using AMD cpus, and to keep companies like Dell from selling any AMD products by paying them not to do so. All total, it was never enough to knock AMD out of the box. Intel's going to have its work cut out for it from now on, imo.

 

I won't even look more than a year ahead, really. Things change so fast! But I didn't think AMD should have sold its FABs, either--I also complained about that. But in retrospect I was wrong. Was likely the best thing at the time. AMD's problem up to now--post Jerry Sanders--was top management, imo. Until Su, none of the people they hired to run the company had a clue. With the right management there's no stopping AMD, whereas Intel is sitting on enough $ not to have to worry about bad management all that much. Intel failed with Rdram, with Itanium, and many other things that literally cost them billions in investments. But they had enough $ to stay afloat nonetheless. The one thing I think AMD cannot afford is mistakes on the order of those that Intel has made through the years.

 

If Apple drops Intel, they'll likely be dropping the Mac and general PC lines altogether, imo. ARM is fine for cell phones and low-powered devices like iPads and the like, but it doesn't hold a flame in terms of raw horsepower to either AMD or Intel's current x86. ARM is great for low-powered devices, which was its designed purpose (originally for embedded uses), but it's not and likely never will be a brute processor, imo. If, otoh, Cook were to decide to transition all of Apple's current x86 PC products to AMD, then the transition would be virtually seamless--with none of the compatibility problems that always hobbled the Mac every time Jobs became incensed because a cpu manufacturer would not do his bidding--and dropped them to move the Mac somewhere else. Now that Apple is no longer a personality cult, it will have to survive on its products in the marketplace. My thought is that Cook and company would love to stick strictly with cell phones and can everything else, but we shall see.

 

Not necessarily. Remember, Apple used to use PPC before they switched to Intel. They went through I think 2 or 3 versions of OSX that were PPC compatible (I think only one of them supported both PPC and Intel). Nowadays, most people who use a Mac are also mostly only using Apple software, and of course Apple is going to recompile their software for ARM if necessary. Other really big software provides (like Adobe or MS) will quickly conform with whatever Apple does. Apple's Xcode is also pretty competent software, where I assume most other Apple-exclusive devs could easily recompile their programs for ARM without much effort. Don't forget that a great deal of Mac-compatible programs are open-source, so they too can be recompiled.
And then there's web-based applications and iOS compatibility, which will give ARM-based Macs plenty of "backup" options.
That just leaves stuff like games and old programs behind. However, I wouldn't be surprised if Apple makes a new "Rosetta", which was the tool they made back around 2006 which allowed you to run PPC Mac binaries on Intel. Besides, Intel-based Macs already perform the worst of any OS for gaming, and now that Apple is in a feud with Nvidia, that makes gaming an even less attractive feature for Macs.

It's also worth taking a look at Apple's existing hardware. Except for stuff like the $5000+ Mac Pros, most of their current Intel computers are mediocre at best. Apple has their own in-house design for ARM CPUs, so if they add some bonus SSE instructions and crank up the clocks to 3GHz+, performance is going to be a non-issue for them. For the average Mac user, the GPU could do all of the real heavy-lifting.

EDIT:
One last thing to point out is Apple's exclusive graphics API called "Metal" (their competitor to Vulkan and DX12) while deliberately trying to ditch Vulkan and OpenGL. That's not a whole lot different than switching to ARM (in terms of complicated software compatibility), and yet, they're getting away with it. I think even Nvidia is supporting Metal, though don't quote me on that.

It wouldn't be as seamless as you'd expect. If you've ever attempted to build an AMD "hackintosh", you'll quickly realize that there are enough differences in AMD's hardware where Apple would still need to do quite a bit of work (assuming they don't just steal the work done by the hackintosh community). Honestly, I don't think going to ARM would be a whole lot more difficult, since Apple already has a great deal of investment in ARM and their own ARM-based architecture.
I would argue Apple is very much still a personality cult, it's just not as prominent as it once once.

 

I believe that if AMD sustains good earnings, it can compete strongly with Intel, winning is another thing (it's not necessary).

Ryzen 3000 (Zen2) is going to be good, I have no doubt, but I'm not sure about prices, I guess it will be as competitive as the old ones. Either way I'm going to upgrade to Ryzen 3000, probably in august.

 

I agree - AMD just needs to focus on improving revenue. Even if Intel's revenue doesn't go down, as long as AMD's is steadily increasing, that's what matters most.