AMD fans rejoice, Intel has their own Hector Ruiz

[frozentundra123456]

They took the wrong gamble on Bulldozer and it didn't pay off. It happens. But that doesn't mean AMD becomes less of a threat when it has more money if that's the argument you are somehow trying to make.

I was actually saying the opposite.

 

[mtcn77]

Bulldozer wasn't a gamble. It was an HPC gambit. CMT-ALU was compromised. All AMD roadmaps lead to APUs. Once reduced to a sole gpu fetch unit, the apu performs 20% better.
https://www.extremetech.com/computi...md-cpu-performance-by-20-without-overclocking

 

[theeedude]

Even dumber than buying Mobileye for ridiculous amounts of money, which is survivable for a behemoth like Intel, is taking their self driving car development out of Silicon Valley and moving it to a bunch of newly minted early retiree millionaires in Israel who probably couldn't care less financially if Intel strategy succeeds or not. Self driving cars are an AI problem. Most of top AI talent is in Silicon Valley. Good luck convincing top AI talent to:
- Move to Jerusalem for a job.
- Be the one hungry working guy in a crew of people who have already struck gold.
So Intel is limiting itself to homegrown Israeli talent, which is fine indeed, but not in comparison to the type of global top talent Silicon Valley attracts, while Intel's competitors are going to go poach away at what talent they still have in automotive space which doesn't want to relocate to Israel.

 

[William Gaatjes]

Even dumber than buying Mobileye for ridiculous amounts of money, which is survivable for a behemoth like Intel, is taking their self driving car development out of Silicon Valley and moving it to a bunch of newly minted early retiree millionaires in Israel who probably couldn't care less financially if Intel strategy succeeds or not. Self driving cars are an AI problem. Most of top AI talent is in Silicon Valley. Good luck convincing top AI talent to:
- Move to Jerusalem for a job.
- Be the one hungry working guy in a crew of people who have already struck gold.
So Intel is limiting itself to homegrown Israeli talent, which is fine indeed, but not in comparison to the type of global top talent Silicon Valley attracts, while Intel's competitors are going to go poach away at what talent they still have in automotive space which doesn't want to relocate to Israel.

Well, almost all the successful design iterations of the latest and greatest cpu of Intel since Nehalem already come from the design facility in Israel.
Intel US is hollowed out from the inside.

 

[William Gaatjes]

And do not forget Intel Capital. It has invested in many startups in Israel for some time now since the late half of the 90s.
Intel will find their way again.

 

[DrMrLordX]

Well, you have to admit, the Haifa campus has been pretty successful going back even further than Nehalem. I think they were at least partially responsible for Core2 . . . Dothan and Yonah came from them.

 

[formulav8]

Ironically, when they were in a much worse financial position they came up with a very good product in Zen.

I still think Ruinz was handed a strong company with good possibilities and he ran it in the ground while given a golden parachute to make sure he didn't get hurt. BD was supposed to lower die space/wafer costs with good threading performance-wise and back fired in regular computing. If at least Steamroller went to the 8 core model, AMD would have looked much better in multi-threading. The multi-core uptick with Steamroller was nice but AMD couldn't afford to make it an 8 core I'm guessing especially on the same socket.

 

[formulav8]

All AMD roadmaps lead to APUs

AMD was apparently looking to use the GPU's for floating point for many things instead of the CPU and it did not work out that way obviously.

 

[theeedude]

Well, almost all the successful design iterations of the latest and greatest cpu of Intel since Nehalem already come from the design facility in Israel.
Intel US is hollowed out from the inside.

The problem is existing Mobileye guys are going to be instant millionaires who care about Intel job about as much as a hobby, top talent in Israel would rather go to the next Mobileye than to Intel, plus US and international AI talent will have almost zero interest in relocating to Jerusalem.

 

[Jan Olšan]

I still think Ruinz was handed a strong company with good possibilities and he ran it in the ground while given a golden parachute to make sure he didn't get hurt. BD was supposed to lower die space/wafer costs with good threading performance-wise and back fired in regular computing. If at least Steamroller went to the 8 core model, AMD would have looked much better in multi-threading. The multi-core uptick with Steamroller was nice but AMD couldn't afford to make it an 8 core I'm guessing especially on the same socket.

I'm disappointed they wasted money on Seattle and didn't instead quickly throw together a BGA SoC like that based on 8 Excavator modules. That could have had some success actually, looking at Xeon-D. Not sure if it would have been profitable, but Seattle was completely useless by the look of it. The ideal thing to do was probably to not do anything there, but I guess they wanted to ride the ARM hype to please investors. It was definitely a mistake though, they needlessly spent money on promoting the server ARM platform and funding software development of it - despite the fact it is actually threat to them as x86 platform in servers is much better bet for them when they have Zen/Naples. They probably realised that already though, which is good. The money was wasted at the worst time though.

(I don't believe K12 would have been substantially better than Zen, and as such it was pointless to spend money on it when Zen was aiming to give the same goodies but on mature platform that doesn't need extensive investment just to be barely usable. To be frank, x86 barrier and technology/compatibility advantage is as much asset to AMD as it is to Intel and they should have protected it instead of undermining it).

 

[William Gaatjes]

The problem is existing Mobileye guys are going to be instant millionaires who care about Intel job about as much as a hobby, top talent in Israel would rather go to the next Mobileye than to Intel, plus US and international AI talent will have almost zero interest in relocating to Jerusalem.

Well, it does not have to be all that bad. If you are sort of a genius and you have no longer to worry about income like the mobileye guys do now, it is much easier to come up with great inventions. Although, it does take a certain amount of discipline not to be "living large" all the time.
I do not see why top talent would not work for Intel. Seems to be a big and serious name in Israel.
But i do agree about the location, the constant threat is not something i would enjoy in living. Add the orthodox religious Jews that are very likely also not fun to be with if you are not Jewish.

 

[jpiniero]

(I don't believe K12 would have been substantially better than Zen, and as such it was pointless to spend money on it when Zen was aiming to give the same goodies but on mature platform that doesn't need extensive investment just to be barely usable. To be frank, x86 barrier and technology/compatibility advantage is as much asset to AMD as it is to Intel and they should have protected it instead of undermining it).

Keller said that the K12 core was wider and more efficient than Zen although it would not have clocked anywhere near what Zen has. Although that was several years ago; how the projects compare now we'll never know.

 

[Murloc]

jerusalem sure doesn't sound as fun as Tel Aviv.

Anyway why wouldn't they be able to keep a team in California at the same time? It's not most efficient but they could work on different projects.

 

[Arachnotronic]

The problem is existing Mobileye guys are going to be instant millionaires who care about Intel job about as much as a hobby, top talent in Israel would rather go to the next Mobileye than to Intel, plus US and international AI talent will have almost zero interest in relocating to Jerusalem.

It's a 33% boost in the stock price, not a 10X one, some perspective is important