Chris Hook leaves AMD, starting new role by the end of April

erek

[H]F Junkie
Joined
Dec 19, 2005
Messages
10,894
"Hi Everyone, As some of you know, I made the decision recently to leave AMD to pursue a new role outside the company (which I start at the end of April). AMD has been a great company to work for, and I owe a tremendous amount to them both professionally and personally. Changing jobs was an incredibly hard decision to make since I’ve been with AMD/ATI since I was still in my late 20s, which is so long ago that ATI still had a smoking room, there was only one TV in the office (a small 14” black and white version we crowded into a room to watch 9/11 on), transistor size was still measured in microns, and 320×240 was considered ‘high res’.

The most rewarding experience I had at AMD was reigniting Radeon over the past couple of years with Radeon Technologies Group, growing its market share from high teens to low thirties, and achieving a record revenue quarter late last year. It was a ton of work, but I’ve never had so much fun, visited so many great places, or worked with such amazing people. I’m also incredibly grateful to the tech press, who have been my second family for the past couple of decades, and I’ll always remember our adventures in places like Tunis, Iceland, Macau, Ibiza and the USS Hornet, to name a few. I hope there will be more of that one day.

I’ll fill you all in on my new opportunity in the coming days and weeks. In the mean time, I’m going to enjoy a few days of vacation where I don’t have to be constantly checking a work smart phone to see what’s going on back in the office (that hasn’t happened since I got my first Blackberry in 2002).

Cheers, Chris"
 
The problem wasn't in the people. The problem was the lack of resources.

As in most projects, you get a budget.

Then based on the budget, you set realistic goals of what you want achieve.

The problem with Vega?

Those goals weren't achieved. None of the major new features even work properly.

Now, the people responsible are getting the ax.
 
As in most projects, you get a budget.

Then based on the budget, you set realistic goals of what you want achieve.

The problem with Vega?

Those goals weren't achieved. None of the major new features even work properly.

Now, the people responsible are getting the ax.

The problem is when you set resources and goals for the GPU division, then reduce the resources because the CPU division is eating them, and you leave the GPU people with pants down so they have to release what they can.

If you want to imagine that people has worked designing good graphics during more than one decade suddenly doesn't know anymore how to do graphics or what is a "budget", do it, but you aren't fooling me.
 
The problem is when you set resources and goals for the GPU division, then reduce the resources because the CPU division is eating them, and you leave the GPU people with pants down so they have to release what they can.

If you want to imagine that people has worked designing good graphics during more than one decade suddenly doesn't know anymore how to do graphics or what is a "budget", do it, but you aren't fooling me.

These are just excuses.

Other companies layoffs too.

You have to make due with what you have.
 
Didn't he recently get a promotion as well ?

Well he could check his smart phone but there is nothing much going on at the moment in RTG anyway :)
 


I saw that there doing something else but you can't say that RTG was on any kind of roll but a downwards spiral.


When things are going well, senior management pretty much leaves you alone. You can bend the rules and do what you want.

It’s when things go badly that senior management steps in and changes things.
 
Would it be like the case where the head of the Austin team leading the cat cores (Bobcat, Jaguar) was poached by Samsung, and the rest of the team went to join their comrade at Samsung? By the few accounts out there (nobody seems willing to talk in absolutes), it seemed the team wasn't happy at AMD. At Samsung, they recently released the M3 core (which is a very fast CPU, when not hampered by Samsung's software and marketing teams in a pale attempt to force performance parity between the 845 and the Exynos).
 
Would it be like the case where the head of the Austin team leading the cat cores (Bobcat, Jaguar) was poached by Samsung, and the rest of the team went to join their comrade at Samsung? By the few accounts out there (nobody seems willing to talk in absolutes), it seemed the team wasn't happy at AMD. At Samsung, they recently released the M3 core (which is a very fast CPU, when not hampered by Samsung's software and marketing teams in a pale attempt to force performance parity between the 845 and the Exynos).

Well, poaching is less of a concern now.

Back then, AMD (as a whole) looked like it was on a bottomless spiral, so not a lot of people wanted to be here.
 
not a whole lot of f ks were given. RTG does need new blood with workable ideas that won't overspend budget to achieve outdated performance.
 
I hope Intel buys AMD. God I've been wishing for something like this to Happen.

Lot of people left AMD ... what .... 3 years ago or so? 2 1/2?

How much longer until we have competent competition from AMD?

nVidia is pulling away at an unbelievable speed that the problem is going to become impossible to overcome without a vast sum of money and expertise.

And we all know that when AMD does announce their new graphic cards that they will still be way behind nVidia in performance.
 
I hope Intel buys AMD. God I've been wishing for something like this to Happen.

Lot of people left AMD ... what .... 3 years ago or so? 2 1/2?

How much longer until we have competent competition from AMD?

nVidia is pulling away at an unbelievable speed that the problem is going to become impossible to overcome without a vast sum of money and expertise.

And we all know that when AMD does announce their new graphic cards that they will still be way behind nVidia in performance.

I am sure the FTC would approve /s
 
I hope Intel buys AMD. God I've been wishing for something like this to Happen.

Lot of people left AMD ... what .... 3 years ago or so? 2 1/2?

How much longer until we have competent competition from AMD?

nVidia is pulling away at an unbelievable speed that the problem is going to become impossible to overcome without a vast sum of money and expertise.

And we all know that when AMD does announce their new graphic cards that they will still be way behind nVidia in performance.

Go bring a couple of billion dollars to AMD and preferably a couple of billion more while you are at it. AMD has been strapped for cash and unless you are a miracle worker and can spend 1 billion 10 times. How come most of you guys still not realize that when you don't have any money you need to make ends meet .

Check the video I posted look at the new title for Suzanne Plummer you don't do that if your GPU division is working as intended.

You know why there are nearly no players in the GPU playing field because the amount of money needed for it is silly and only deep pockets would allow development of a GPU.
 
Go bring a couple of billion dollars to AMD and preferably a couple of billion more while you are at it. AMD has been strapped for cash and unless you are a miracle worker and can spend 1 billion 10 times. How come most of you guys still not realize that when you don't have any money you need to make ends meet .

Check the video I posted look at the new title for Suzanne Plummer you don't do that if your GPU division is working as intended.

You know why there are nearly no players in the GPU playing field because the amount of money needed for it is silly and only deep pockets would allow development of a GPU.

It is why Nvidia are able to throw everything into it, they have no CPU of substance to develop and AMD's principal business is CPU design which itself costs a fortune. You get the feeling that AMD would rather break HPC markets than build a gaming GPU and rightly so. The question is what can AMD do with its graphics licence to make it profitable without being a drain? The easiest thing is to licence out, do custom solutions and focus on scientific GPU's only.

Gaming is not big enough money wise to waste money on.
 
When things are going well, senior management pretty much leaves you alone. You can bend the rules and do what you want.

It’s when things go badly that senior management steps in and changes things.

Things are not going well 'thanks' to the senior management.

It is why Nvidia are able to throw everything into it, they have no CPU of substance to develop and AMD's principal business is CPU design which itself costs a fortune. You get the feeling that AMD would rather break HPC markets than build a gaming GPU and rightly so. The question is what can AMD do with its graphics licence to make it profitable without being a drain? The easiest thing is to licence out, do custom solutions and focus on scientific GPU's only.

Gaming is not big enough money wise to waste money on.

Nvidia has ARM CPUs (with both private and licensed muarch) for mobile/automotive, is in partnership with IBM/Cavium and others for HPC, and is placing a foot on the RISC-V ISA.

AMD GPUs have been focusing on compute since the first day and have failed badly. AMD marketshare in HPC is minimal.

Have you noticed that AMD no longer mentions HSA?
 
The technology behind HSA is used heavily, most of it is wrapped into their ROCm branding now.

The idea's behind HSA transformed into vulkan's SPIRV. And no copy pointers in opencl.

LLVM, and GCC both are still actively working on HSAIL support.

And it is generally used to the benefit of opencl as well with no copy pointers.

https://rocm.github.io/
 
Monopolies are not allowed to create themselves, Intel will never buy AMD.it won't be allowed to happen as mentioned by the ftc
 
From the outside looking in, AMD's problem looks like they just bite off more than they can chew. They can *either* build a competitive GPU (see the original GCN releases) OR build a competitive CPU (see Ryzen/TR, or K8), but they can't seem to do both at the same time. When they try to do both, one or the other fails miserably. They don't have the resources.
 
They need to fix one, generate some cash and then fix the other. It took 6 years after the failure of FX to get a decent cpu from AMD, it may be another couple of years before we see a decent gpu that screams buy me
 
You mean that Raja is going to poach a good deal of people from RTG which are rather tired from fighting windmills ?

You can only be "Don Quixote" so long, until you realize in a world full of Ferraris, Lamborghinis, Maseratis, etc., you are still riding a donkey. ;)
 
From the outside looking in, AMD's problem looks like they just bite off more than they can chew. They can *either* build a competitive GPU (see the original GCN releases) OR build a competitive CPU (see Ryzen/TR, or K8), but they can't seem to do both at the same time. When they try to do both, one or the other fails miserably. They don't have the resources.

AMD can do both. Main problem is bad management and internal wars are wasting resources. Just count all the projects have been canceled in last five years. Lots of millions of dollars have gone.
 
AMD can do both. Main problem is bad management and internal wars are wasting resources. Just count all the projects have been canceled in last five years. Lots of millions of dollars have gone.

Maybe. But all big companies have tremendous assloads of waste. If you're Intel, you can afford it. If you're AMD, you really can't.
 
AMD being a small company is fighting two giants (Intel and NVidia) which is very hard because both companies have really deep pockets contrary to AMD. AMD fails to conquer either of companies. I am still impressed what they have achieved with their budget. I wish another, bigger, american company will buy AMD.
 
AMD being a small company is fighting two giants (Intel and NVidia) which is very hard because both companies have really deep pockets contrary to AMD. AMD fails to conquer either of companies. I am still impressed what they have achieved with their budget. I wish another, bigger, american company will buy AMD.

When Intel or AMD gets bought the license for x86 or X86_64 will need renegotiation for either company. So no one can swoop in and take the cake.
 
I don't get this RTG is a disaster sentiment when the only place they're not competitive is in the very high end. Everything below Vega64 is competitive in price/performance (at msrp at least), and Raven Ridge is without peer. But yeah, AMD is a total disaster because they can't compete with the 1080ti.

That must make Nvidia a complete disaster because they can't compete in igp. Fire Huang, his igps are so bad he's completely locked out of PC igp and is only in the weakest of the 3 consoles! What a complete disaster!
 
I don't get this RTG is a disaster sentiment when the only place they're not competitive is in the very high end. Everything below Vega64 is competitive in price/performance (at msrp at least), and Raven Ridge is without peer. But yeah, AMD is a total disaster because they can't compete with the 1080ti.

That must make Nvidia a complete disaster because they can't compete in igp. Fire Huang, his igps are so bad he's completely locked out of PC igp and is only in the weakest of the 3 consoles! What a complete disaster!

It's a sort of trickle down mentality. Since Nvidia rules the roost at the ultra high end, they have a branding advantage in the mid-range and entry-level, where AMD still competes effectively on performance and price. Halo products often do that. However, Raven Ridge is indeed without peer - except from Intel's own Hades Canyon with Vega (or is it a Vega/Polaris hybrid?) unit. But how big is that market really? Most folks who omit buying a discrete GPU don't really care about games or GPU performance, so Intel's integrated units suffice. For ultra-budget gamers and small form factor folks, though... definitely, Raven Ridge is excellent.
 
I don't get this RTG is a disaster sentiment when the only place they're not competitive is in the very high end. Everything below Vega64 is competitive in price/performance (at msrp at least), and Raven Ridge is without peer. But yeah, AMD is a total disaster because they can't compete with the 1080ti.

That must make Nvidia a complete disaster because they can't compete in igp. Fire Huang, his igps are so bad he's completely locked out of PC igp and is only in the weakest of the 3 consoles! What a complete disaster!

The problem with Radeon RX Vega 56/64 is that they are too expensive to make.

They use huge dies and very expensive HBM2 memory.

The only times they ever sold at MSRPs were when AMD was handing out retailers rebates.
 
The problem with Radeon RX Vega 56/64 is that they are too expensive to make.

They use huge dies and very expensive HBM2 memory.

The only times they ever sold at MSRPs were when AMD was handing out retailers rebates.

Agree with first two points. Third point... no. Because of crypto, AMD still made assloads of money, as did the retailers doing heavy markup on them, and the things sold well above MSRP (even if not right away). Crypto saved AMD from a marginal Vega product.
 
Agree with first two points. Third point... no. Because of crypto, AMD still made assloads of money, as did the retailers doing heavy markup on them, and the things sold well above MSRP (even if not right away). Crypto saved AMD from a marginal Vega product.

AMD didn't make a lot of money; retailers did.

Wholesale prices of video cards (that retailers pay) were make long in advance of the mining sham.
 
Back
Top