AMD CPU speculation... and expert conjecture

[Rum]

Juan did you even watch the AMD Core Innovation Summit? Lisa Su and Jim Keller said that x86 is just as important as ARM or did you chose to tune that part out? Keller also said that the build process will be fast, he and his team are taking things from other builds/ arch that AMD perfected in the past and implement them in the new Arch. Your assumptions that the new x86 Arch won't outperform the K12 are asinine and unfounded.
One last thing I am sure that Keller has plenty of funds to build his new x86 toy, all your nonsense just sounds like sour grapes cuz it's not all about ARM!

Sounds like they are doing what Intel did when they scrapped crapburst and made Core 2. I really hope this new arch is a winner. Phenom II was a very solid chip, and still are pretty capable. More capable than faildozer, and not really all that far behind piledriver at similar clock speeds and core counts. Had they been able to take the memory controller improvements of faildozer/piledriver, and tweaked PhII in a way that it could handle the higher clock speeds, and core counts, I think we would have had a far better chip than what we were given.

It sure does look like that is what they are doing and why not, if you already have things that you are doing well then you might as well use them! Unfortunately they didn't go that way with Phenom ll, it could have been it's last hurrah but well know how that went... Blahdozer...

Great interview by Tech Radar with Huddy, looks like M$ approached them about Mantle and implementing that in DX12... AMD didn't steal it like some tried to say!
http://www.techradar.com/news/computing/pc/amd-s-richard-huddy-on-the-state-of-pc-graphics-mantle-2-and-apus-1255575/2#articleContent

A quote from the interview on page 2

Mantle 2 - if it takes us about a year to get through a Mantle iteration - then Mantle 2 will come around the same kind of timeframe as DirectX 12. DX 12 brings a lot of the goodness that Mantle brought. We had a lot of conversations with Microsoft about what we were doing with Mantle, and in those conversations, they said, 'OK, if you really can solve this problem of building a better throughput system that runs on Windows, then we'd like to take that to Windows as well and we'll make that the extra software functionality that comes in DX 12.' So that's how DX 12 has come about.

 

[blackkstar]

I would like to point one thing out...as at the time we did not know...

When Jim Keller said, "AMD are on track to catch Intel in high performance cores"

We expected that would be excavator likely...knowing what I know now...that would be the next uarch.

As for 50% better IPC with Intel, no...they have roughly 10-15% better IPC with better hardware prediction. I put this forth...

AMD is on the rise...we will see what comes...

But branch prediction is one of factors which influence IPC. Who cares how good the rest is if you can't utilize it?

So every software that relies on prediction (i.e. almost everything) is AMD-biased?

Not all software relies on branch prediction for performance. Rendering doesn't have a lot of branches and look how well Piledriver does there, specially in an environment where it's not using funky tools that aren't so good for AMD.

I believe what rocks is getting at is this: http://phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=4

You can see massive differences between the two. Sometimes, FX is extremely close. Sometimes, it's not even close. What he is trying to say is that the times that it is close are usually when the benchmark is not doing a lot of branches in code (John the Ripper, Rendering, etc) and when it falls flat on its face (some of the scientific ones), it's doing lots of branch prediction.

As far as I remember, things have always been like this with AMD. Even when they had K8 there were things they just really sucked at.

Which is why most reviews suck, it's way too easy to pick strong programs for a certain product you like and then end your conclusion with something like "wow, the product I like won 4/5 benchmarks, it's better! It averaged 40% better! WOW!" When if you used a different suite of programs, they'd be more even (or even switching places) and the conclusion would be completely different.

One thing about x86 vs ARM that I don't see mentioned much is what happens when you get added instructions in an environment you can use them. Haswell brought AVX2 and look what it did for Dolphin Emulator. It's ridiculously faster. ARM might be catching up to x86 when the x86 is running nothing higher than SSE2 code, but if you were to compare ARM IPC to Haswell IPC in something like Dolphin Emulator when Haswell is using AVX2, it'd be an absolute bloodbath.

Even Haswell vs IB is an absolute bloodbath. And ARM being RISC while x86 being CISC will always have that sort of situation. IPC is meaningless comparison if both CPUs can do 20 instructions in a given unit of time, but the x86 is doing FMA and all these operations in single instructions while the RISC is doing something like a simple FMA in a few instructions.

 

[szatkus]

I would like to point one thing out...as at the time we did not know...

When Jim Keller said, "AMD are on track to catch Intel in high performance cores"

We expected that would be excavator likely...knowing what I know now...that would be the next uarch.

As for 50% better IPC with Intel, no...they have roughly 10-15% better IPC with better hardware prediction. I put this forth...

AMD is on the rise...we will see what comes...

But branch prediction is one of factors which influence IPC. Who cares how good the rest is if you can't utilize it?

So every software that relies on prediction (i.e. almost everything) is AMD-biased?

Not all software relies on branch prediction for performance. Rendering doesn't have a lot of branches and look how well Piledriver does there, specially in an environment where it's not using funky tools that aren't so good for AMD.

I believe what rocks is getting at is this: http://phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=4

You can see massive differences between the two. Sometimes, FX is extremely close. Sometimes, it's not even close. What he is trying to say is that the times that it is close are usually when the benchmark is not doing a lot of branches in code (John the Ripper, Rendering, etc) and when it falls flat on its face (some of the scientific ones), it's doing lots of branch prediction.

As far as I remember, things have always been like this with AMD. Even when they had K8 there were things they just really sucked at.

Which is why most reviews suck, it's way too easy to pick strong programs for a certain product you like and then end your conclusion with something like "wow, the product I like won 4/5 benchmarks, it's better! It averaged 40% better! WOW!" When if you used a different suite of programs, they'd be more even (or even switching places) and the conclusion would be completely different.

One thing about x86 vs ARM that I don't see mentioned much is what happens when you get added instructions in an environment you can use them. Haswell brought AVX2 and look what it did for Dolphin Emulator. It's ridiculously faster. ARM might be catching up to x86 when the x86 is running nothing higher than SSE2 code, but if you were to compare ARM IPC to Haswell IPC in something like Dolphin Emulator when Haswell is using AVX2, it'd be an absolute bloodbath.

Even Haswell vs IB is an absolute bloodbath. And ARM being RISC while x86 being CISC will always have that sort of situation. IPC is meaningless comparison if both CPUs can do 20 instructions in a given unit of time, but the x86 is doing FMA and all these operations in single instructions while the RISC is doing something like a simple FMA in a few instructions.

These are multithreaded benchmarks at stock clocks. Single thread is more important for average usage. Hell, the software I'm developing right now won't run in multiple threads even if I really wanted.

And there we have +50% IPC:
http://media.bestofmicro.com/0/T/418637/original/per-clock.png

 

[jdwii]

Juanrga, you mean to say the Bulldozer team were "fired," not "hired." Seen the same mistake several times now!

Now expect 200+ reposts by Juan that "hire" actually means "fire".

Be nice he is from Spain they have a 25% unemployment rate its probably hard to tell the difference between hired and fired

 

[jdwii]

It is about equal in normal use, but the 6300 is cheaper allowing extra money to go towards gpu.

http://static.techspot.com/articles-info/827/bench/CPU_...

They cost the same plus Intel has better boards for less money?

I am seeing a 10 dollar minimum difference in price. to someone in this budget range it makes a difference.

In the U.S?
6300
http://www.amazon.com/AMD-FD6300WMHKBOX-FX-6300-Processor-Edition/dp/B009O7YORK
I3 haswell
http://www.amazon.com/Intel-i3-4130-FCLGA-Processor-BX80646I34130/dp/B00EUUKVXM

???????

Plus looking i have to spend 60$ on a AM3+ board instead of 45$ on a intel board just to get a 970 board. The Intel board i am seeing is a 51$ now instead of 45$
http://www.newegg.com/Product/Product.aspx?Item=N82E16813130731

This has 2 USB 3.0's, PCI-E and so on i used this board for the I5 build my friend got its a very good for so cheap. Amd needs a complete new refresh on their higher-end boards. I think the main reason why Intel boards are getting cheaper on the low-end is over power management since Intel uses about half the power. Not to mention intel has more integrated on their CPU
Personally i wouldn't get anything lower then a 6300 for gaming even the quad core PD without L3 for 80$ wouldn't cut it i would rather have a Pentium then since older games will be as fast as a I5 and newer ones would be about equal to a 750K and you can get a Pentium for just 60$ vs 80$(750K) and 20$ on 400$ builds is a huge difference

I used newegg, what TH uses for price reference.
i3 4130 = 124.99 http://tinyurl.com/pqfgfe3
fx 6300= 119.99 http://tinyurl.com/ohyrds4

And A 760G am3+ is less than 50 dollars further cutting the difference.

Pentiums will play games as long as you don't have anything else running, If as you say have built a few of them, you should already know this. So your either spreading fud or uneducated in what your talking about.
You Pick.

"if you have other stuff running"
When i game on lower-end hardware i put very little in the background i build many Pentium builds BTW. Very popular chips for the money i don't even think you can build a 750K build and have it come close to the Pentium when only spending 400$ it would mean you have to drop your GPU budget or lower your ram to 4GB vs 8. Also i said 970 board with the same features as the Intel board including 3.0 sata 6 and so on.

 

[jdwii]

Jim Keller has had tons of time in the industry working on projects since K7...if he says it will be better, I have no reason to doubt. Though, like those I know at AMD, I consider it a win if they are within 5%.

Current Intel arch is about 50% faster than current AMD offers in regards to IPC, if next x86 AMD arch gains 50% it will put them right where Intel is sitting NOW, and i bet Intel will also bring a good bunch of extra performance on Skylake, therefor i do not expect AMD to even match Intel with their new Arch, but 50% gains in IPC is quite good none the less.

Per core per clock Steamroller is around 35% slower compared to the Haswell design under several benchmarks of mine. PD yes its more closer to the 50% number, and the phenom is around 10% faster per clock per core. The only time this is not true is when programs use certain instruction sets that the phenom doesn't have.

 

[anxiousinfusion]

FF is coming earlier than I thought: http://www.digitimes.com/news/a20140701PD201.html But of course, Apple is going to continue node hogging.

 

[-Fran-]

I'll side with palladin on the IPC thing. You need a lot of context to understand what "IPC" is really telling you (which comes from RISC based OPs).

You need to simplify the comparison and check overall performance at stock clocks and the price the CPU sells at, plus the round of OC capabilities. All other metrics you can produce/see are not that important in the face of a performance race.

Notebook space and SFF cares about power, so you can add power into the mix for those markets.

Point is: as much as we would like for "IPC" to be an oranges to oranges comparison between PD and SB/IB/H, it is not. We all know (or we should, by now) that process matters a lot in how a CPU performs, so IPC isn't able to tell you that side of the picture, making it an incomplete metric to measure "performance". More into that, Intel and AMD won't have the same fabrication nodes in the short term (can't see into the future, but I'm sure it won't be in the long either, haha), so fabrication process DOES MATTER.

Cheers!

 

[etayorius]

Me, as a Gamer... i would NEVER touch the Pentium or the i3 CPUs, JUST NO... i can`t really speak for the non gamer AVG user.

Unless you`re playing games from 2008 and older, then yes grab a Pentium or an i3. otherwise... just grab the i7, i5, FX or PhenomII with at least 4 cores.

 

[szatkus]

Me, as a Gamer... i would NEVER touch the Pentium or the i3 CPUs, JUST NO... i can`t really speak for the non gamer AVG user.

Unless you`re playing games from 2008 and older, then yes grab a Pentium or an i3. otherwise... just grab the i7, i5, FX or PhenomII with at least 4 cores.

Tomb Raider from 2013 runs well on Pentium (even on Phenom II X2). Just like many other games. Not everyone plays Crysis 3.

 

[8350rocks]

Luckily, we now have 2 roughly comparable chips in terms of core count and frequency: The 8350 and 4790k.

8350rocks: are you saying that the 8350 is within 10 to 15% of the 4790k on average in cpu tests (I.e. excluding gpu limited tests)? If so, then I'll be asking myself why I've waited so long to upgrade the amd way instead of being tempted to go intel for the first time in 18 years.

You are comparing Apples to watermelons. the 4790k is, in AMD terms, an 8 module CPU, and you are comparing that to a 4 module CPU.

I can misconstrue things too...though...do you want to compare the 4790k to an 8 module Opteron at the same clockspeed with quad channel memory and all the other insane server features that come with it...?

 

[8350rocks]

I would like to point one thing out...as at the time we did not know...

When Jim Keller said, "AMD are on track to catch Intel in high performance cores"

We expected that would be excavator likely...knowing what I know now...that would be the next uarch.

As for 50% better IPC with Intel, no...they have roughly 10-15% better IPC with better hardware prediction. I put this forth...

AMD is on the rise...we will see what comes...

But branch prediction is one of factors which influence IPC. Who cares how good the rest is if you can't utilize it?

So every software that relies on prediction (i.e. almost everything) is AMD-biased?

I would like to point one thing out...as at the time we did not know...

When Jim Keller said, "AMD are on track to catch Intel in high performance cores"

We expected that would be excavator likely...knowing what I know now...that would be the next uarch.

As for 50% better IPC with Intel, no...they have roughly 10-15% better IPC with better hardware prediction. I put this forth...

AMD is on the rise...we will see what comes...

But branch prediction is one of factors which influence IPC. Who cares how good the rest is if you can't utilize it?

So every software that relies on prediction (i.e. almost everything) is AMD-biased?

Not all software relies on branch prediction for performance. Rendering doesn't have a lot of branches and look how well Piledriver does there, specially in an environment where it's not using funky tools that aren't so good for AMD.

I believe what rocks is getting at is this: http://phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=4

You can see massive differences between the two. Sometimes, FX is extremely close. Sometimes, it's not even close. What he is trying to say is that the times that it is close are usually when the benchmark is not doing a lot of branches in code (John the Ripper, Rendering, etc) and when it falls flat on its face (some of the scientific ones), it's doing lots of branch prediction.

As far as I remember, things have always been like this with AMD. Even when they had K8 there were things they just really sucked at.

Which is why most reviews suck, it's way too easy to pick strong programs for a certain product you like and then end your conclusion with something like "wow, the product I like won 4/5 benchmarks, it's better! It averaged 40% better! WOW!" When if you used a different suite of programs, they'd be more even (or even switching places) and the conclusion would be completely different.

One thing about x86 vs ARM that I don't see mentioned much is what happens when you get added instructions in an environment you can use them. Haswell brought AVX2 and look what it did for Dolphin Emulator. It's ridiculously faster. ARM might be catching up to x86 when the x86 is running nothing higher than SSE2 code, but if you were to compare ARM IPC to Haswell IPC in something like Dolphin Emulator when Haswell is using AVX2, it'd be an absolute bloodbath.

Even Haswell vs IB is an absolute bloodbath. And ARM being RISC while x86 being CISC will always have that sort of situation. IPC is meaningless comparison if both CPUs can do 20 instructions in a given unit of time, but the x86 is doing FMA and all these operations in single instructions while the RISC is doing something like a simple FMA in a few instructions.

blackkstar basically hit the nail on the head...

 

[szatkus]

Luckily, we now have 2 roughly comparable chips in terms of core count and frequency: The 8350 and 4790k.

8350rocks: are you saying that the 8350 is within 10 to 15% of the 4790k on average in cpu tests (I.e. excluding gpu limited tests)? If so, then I'll be asking myself why I've waited so long to upgrade the amd way instead of being tempted to go intel for the first time in 18 years.

You are comparing Apples to watermelons. the 4790k is, in AMD terms, an 8 module CPU, and you are comparing that to a 4 module CPU.

I can misconstrue things too...though...do you want to compare the 4790k to an 8 module Opteron at the same clockspeed with quad channel memory and all the other insane server features that come with it...?

So FX-8350 competes with i3 (4 modules vs 4 threads)? Neat

 

[jdwii]

It is about equal in normal use, but the 6300 is cheaper allowing extra money to go towards gpu.

http://static.techspot.com/articles-info/827/bench/CPU_...

They cost the same plus Intel has better boards for less money?

I am seeing a 10 dollar minimum difference in price. to someone in this budget range it makes a difference.

In the U.S?
6300
http://www.amazon.com/AMD-FD6300WMHKBOX-FX-6300-Processor-Edition/dp/B009O7YORK
I3 haswell
http://www.amazon.com/Intel-i3-4130-FCLGA-Processor-BX80646I34130/dp/B00EUUKVXM

???????

Plus looking i have to spend 60$ on a AM3+ board instead of 45$ on a intel board just to get a 970 board. The Intel board i am seeing is a 51$ now instead of 45$
http://www.newegg.com/Product/Product.aspx?Item=N82E16813130731

This has 2 USB 3.0's, PCI-E and so on i used this board for the I5 build my friend got its a very good for so cheap. Amd needs a complete new refresh on their higher-end boards. I think the main reason why Intel boards are getting cheaper on the low-end is over power management since Intel uses about half the power. Not to mention intel has more integrated on their CPU
Personally i wouldn't get anything lower then a 6300 for gaming even the quad core PD without L3 for 80$ wouldn't cut it i would rather have a Pentium then since older games will be as fast as a I5 and newer ones would be about equal to a 750K and you can get a Pentium for just 60$ vs 80$(750K) and 20$ on 400$ builds is a huge difference

I used newegg, what TH uses for price reference.
i3 4130 = 124.99 http://tinyurl.com/pqfgfe3
fx 6300= 119.99 http://tinyurl.com/ohyrds4

And A 760G am3+ is less than 50 dollars further cutting the difference.

Pentiums will play games as long as you don't have anything else running, If as you say have built a few of them, you should already know this. So your either spreading fud or uneducated in what your talking about.
You Pick.

"if you have other stuff running"
When i game on lower-end hardware i put very little in the background i build many Pentium builds BTW. Very popular chips for the money i don't even think you can build a 750K build and have it come close to the Pentium when only spending 400$ it would mean you have to drop your GPU budget or lower your ram to 4GB vs 8. Also i said 970 board with the same features as the Intel board including 3.0 sata 6 and so on.

well then i understand why you like pentiums so much. specifically reducing background programs is not how normal people use there pc just to play a game.. Most people have AV, and browser tabs open while playing games. Try it yourself, I have and watch the fps jump around .
$400 i will play the game

A8 5600k = $ 99.99 @ newegg
ASRock FM2A55M = $ 42.99

That is $143.00 for cpu/board/ graphics show me a pentium combo at that price

Well in that case i would not recommend that i don't even have all that stuff running on my 8 core. Just not me to have stuff open that doesn't need to be open the most i do is run Pandora and a game or maybe video capturing and a game(which a 4 core would be better on).
However i recommend people with such builds to not do that and go for max GPU performance with doable CPU performance and the dual core Pentium+7770(250X) does the work great and under my testing does better then the A10 7850K but in general performance it even feels slower to me but for gaming its better. Some newer games like watch dogs show the Pentium next to a quad core steamroller.

Also for the combo part that is a nonsense comparison since your part is highly weaker in GPU performance you are more talking about 350$ gaming rigs not 400$, anything under 400$(to weak) i don't do except for the new Amd Athlon AM1 for HTPC
"256 Radeon (shader) cores"
Lol not even GCN doesn't support directx 12 or mantle.

 

[8350rocks]

Luckily, we now have 2 roughly comparable chips in terms of core count and frequency: The 8350 and 4790k.

8350rocks: are you saying that the 8350 is within 10 to 15% of the 4790k on average in cpu tests (I.e. excluding gpu limited tests)? If so, then I'll be asking myself why I've waited so long to upgrade the amd way instead of being tempted to go intel for the first time in 18 years.

You are comparing Apples to watermelons. the 4790k is, in AMD terms, an 8 module CPU, and you are comparing that to a 4 module CPU.

I can misconstrue things too...though...do you want to compare the 4790k to an 8 module Opteron at the same clockspeed with quad channel memory and all the other insane server features that come with it...?

So FX-8350 competes with i3 (4 modules vs 4 threads)? Neat

i5/i7 stop trolling.

 

[colinp]

You are comparing Apples to watermelons. the 4790k is, in AMD terms, an 8 module CPU, and you are comparing that to a 4 module CPU.

I can misconstrue things too...though...do you want to compare the 4790k to an 8 module Opteron at the same clockspeed with quad channel memory and all the other insane server features that come with it...?

What? The 4790k is a quad core with hyperthreading.

 

[jdwii]

Luckily, we now have 2 roughly comparable chips in terms of core count and frequency: The 8350 and 4790k.

8350rocks: are you saying that the 8350 is within 10 to 15% of the 4790k on average in cpu tests (I.e. excluding gpu limited tests)? If so, then I'll be asking myself why I've waited so long to upgrade the amd way instead of being tempted to go intel for the first time in 18 years.

You are comparing Apples to watermelons. the 4790k is, in AMD terms, an 8 module CPU, and you are comparing that to a 4 module CPU.

I can misconstrue things too...though...do you want to compare the 4790k to an 8 module Opteron at the same clockspeed with quad channel memory and all the other insane server features that come with it...?

Does someone from Amd actually say this or their fanbase? Never heard anyone from Amd state otherwise.

 

[vmN]

How the hell is the 4790k a 8 module CPU? It is literally the weirdest statement I have heard to far.

 

[jdwii]

How the hell is the 4790k a 8 module CPU? It is literally the weirdest statement I have heard to far.

I think he means something i never heard anyone from Amd state.
I thnk he meant a module is equal an Intel core with HT that is what the usual fanbase claims, for example when the cores are all pushed with software a Quad core FX competes with a dual core with HT.

 

[vmN]

Well, you cannot just "translate" a certain architecture over to another like that. It is way more complex.

Just don't call it an 8 CMT core CPU, then I'll be happy. I don't care if there is an AMD employee running around calling it, it is simply wrong.

 

[szatkus]

Luckily, we now have 2 roughly comparable chips in terms of core count and frequency: The 8350 and 4790k.

8350rocks: are you saying that the 8350 is within 10 to 15% of the 4790k on average in cpu tests (I.e. excluding gpu limited tests)? If so, then I'll be asking myself why I've waited so long to upgrade the amd way instead of being tempted to go intel for the first time in 18 years.

You are comparing Apples to watermelons. the 4790k is, in AMD terms, an 8 module CPU, and you are comparing that to a 4 module CPU.

I can misconstrue things too...though...do you want to compare the 4790k to an 8 module Opteron at the same clockspeed with quad channel memory and all the other insane server features that come with it...?

So FX-8350 competes with i3 (4 modules vs 4 threads)? Neat

i5/i7 stop trolling.

You said: i7 ~ 8 modules.
So i3 ~ 4 modules. Right?

 

[jdwii]

TR: It doesn't seem like the 'mobile gaming is killing PC gaming' doom-and-gloom talk is as heavy as it was a year or so ago.
RH: We've learned that these things co-exist. We tend to think that when a new technology comes along it must be displacing something else and ruining the market for that. The PC has been disrupted. It's probably more that the PC, the replacement cycle has gone a little bit longer rather than people don't own PCs or don't want PCs. I have a smartphone, I have a console, I have a PC and I upgrade each of them when the time comes."

My favorite part of this so far. Juan likes to think of replacements smarter people at Amd like to think of newer markets that co-exist with the older stuff
http://www.techradar.com/news/computing/pc/amd-s-richard-huddy-on-the-state-of-pc-graphics-mantle-2-and-apus-1255575
http://www.techradar.com/news/gaming/richard-huddy-scathes-nvidia-gameworks-talks-amd-s-new-attitude-1255538?src=rss&attr=all

 

[de5_Roy]

kinda looks like mantle 2 is aiming at linux ecosystem, if you read huddy's words on mantle's supposed death a bit differently.
good news for steam os.

 

[jdwii]

kinda looks like mantle 2 is aiming at linux ecosystem, if you read huddy's words on mantle's supposed death a bit differently.
good news for steam os.

it needs to linux needs something better then just openGL.
Also i never knew Amd would hire some guy who sounds like he is getting angry over gameworks(which i own a nvidia card and never knew about it yet) i do agree he should get mad i always hated how Nvidia did that BS. Its like a Edison kind of thinking instead of tesla kind of thinking.

 

[de5_Roy]

Chipset solution vendors seek 28nm capacity at UMC, Globalfoundries and SMIC
http://www.digitimes.com/news/a20140702PD203.html
TSMC currently holds over 80% of the global 28nm process market, that's some intel-class marketshare, lol.

it needs to linux needs something better then just openGL.
Also i never knew Amd would hire some guy who sounds like he is getting angry over gameworks(which i own a nvidia card and never knew about it yet) i do agree he should get mad i always hated how Nvidia did that BS. Its like a Edison kind of thinking instead of tesla kind of thinking.

weren't edison and tesla nvidia's brands/codenames?
huddy is doing pro-amd p.r. he seems to be fixing/reinforcing amd's weak points in isv relationship area.
the techradar interview(text) i didn't notice any new technical/technological details on...anything, other than the publicly known stuff. notice how he says that amd has a healthy dgfx business and segues to gfx cards instead of discussing marketshare or revenue - stuff like that.
my biggest takeaway was mantle 2 and "elsewhere, there will be amd graphics". the latter one may be aimed at semicustom arm/x86 as well as arm ecosystem. it sounded interesting.