New AMD Roadmap for 2013 (donanimhaber.com)

[Ajay]

So am I the only one wondering about Richland? We expected a Steamroller APU with GCN and instead they put a Piledriver APU with Radeon Cores 2.0 on the roadmap. Is it Trinity 2.0 (the same way Brazos 2.0 appeared) or does Radeon Cores 2.0 mean GCN? If so, what is Radeon Cores 1.0 then? We have VLIW5 in 1st Gen Llano APU, VLIW4 in 2nd Gen Trinity APU and now Radeon Cores 2.0 in 3rd Gen. Is it GCN or VLIW4?

For sure Richland will be Tinity 2.0 which, AFAIK, which I thought was going to be a 28nm shrink, but may only be a new stepping with a very small performance bump. I'm hoping Radeon 2.0 means CGN. That would go along with the APU focus on GPU power over CPU power.

But what is coming after that?! If it's a new uArch, I can't see it coming any earlier than 2015 (and that's with the recycling of older design elements, say from Thurban).

 

[Ajay]

It seems like Keller won't be designing anything (he has too much under his belt for that, IMO), but, hopefully, he will help AMD not make the same mistake it did with Bulldozer.

From AMD's press release:

SUNNYVALE, CA--(Marketwire - Aug 1, 2012) - AMD (NYSE: AMD) announced today that Jim Keller, 53, has joined the company as corporate vice president and chief architect of AMD's microprocessor cores, reporting to chief technology officer and senior vice president of technology and engineering Mark Papermaster. In this role, Keller will lead AMD's microprocessor core design efforts aligned with AMD's ambidextrous strategy with a focus on developing both high-performance and low-power processor cores that will be the foundation of AMD's future products.
"Jim is one of the most widely respected and sought-after innovators in the industry and a very strong addition to our engineering team," said Papermaster. "He has contributed to processing innovations that have delivered tremendous compute advances for millions of people all over the world, and we expect that his innovative spirit, low-power design expertise, creativity and drive for success will help us shape our future and fuel our growth."

 

[ShintaiDK]

No you dont need to use specfp_rate for Multithreaded and Throughput. Both of my links above have Auto Parallel = YES.

http://www.spec.org/cpu2006/Docs/readme1st.html#Q15

I dont think you even understand what you post or link to. Let me try give you an example.

http://www.spec.org/cpu2006/results/res2011q1/cpu2006-20110117-14102.pdf (24 cores)
http://www.spec.org/cpu2006/results/res2011q1/cpu2006-20110117-14108.pdf (48 cores)

 

[CHADBOGA]

Yes they will offer desktop variants as well. It's just that SR based CPUs are 2014 now,probably due to GloFo's process node troubles. And the only change to the roadmap is Kaveri since Vishera was always on the 2013 perf. desktop roadmap and there was never a mention of a "new" FX chip based on SR core in 2013. It's likely to come in 2014 on a new socket with DDR4 memory and 4+Modules.

OBR's greatness is evident once more, as he was the first to publicly predict massive lateness for Steamloller. :awe:

 

[wlee15]

They cant cancel the license. Then Intel gets all the IP.

When the license is up for renewal, then the x64 is no longer under IP.

In regards to the copyright side both Intel and AMD have perpetual royalty-free non-transferable licenses to each other copyrights in regards to each other's instruction sets. These rights survive termination for any reason including material breach.

On the patent-side there are three ways to terminate with cause:
1)Termination due material breach
2)Termination due to change of control
3)Termination due to bankruptcy

In the first one the party that is not in breach keeps the rights from the agreement till the end of the term of the agreement or until they are also in breach of the contract. The party or parties in breach lose all their rights in the agreement except for the ones that survive termination such as the copyright ones above. If there is a change of control then the agreement ends and both companies lose their licenses to each other patents. Bankruptcy is the most complicated but it appears that AMD could enter bankruptcy and still retain the cross license agreement as long as there is no change of control(which would terminate the agreement) and retain the ability to grant Intel licenses to their patents. Of course both sides could agree to amend the agreement since it's in the interest of Intel to keep AMD away from patent trolls. (Remember patents don't give you the right to make something they give you the right to deny someone else the right to make something)

 

[mrmt]

In regards to the copyright side both Intel and AMD have perpetual royalty-free non-transferable licenses to each other copyrights in regards to each other's instruction sets. These rights survive termination for any reason including material breach.

Nope. All cases are treated equally as a material breach of the agreement. There is no possibility of AMD maintain access to Intel IP after being acquired.

Check here

http://forums.anandtech.com/showpost.php?p=34190679&postcount=45

 

[AtenRa]

I dont think you even understand what you post or link to. Let me try give you an example.

http://www.spec.org/cpu2006/results/res2011q1/cpu2006-20110117-14102.pdf (24 cores)
http://www.spec.org/cpu2006/results/res2011q1/cpu2006-20110117-14108.pdf (48 cores)

Yeap, i have misunderstood, you're right

Those are the best i could find for each CPU

AMD Opteron 6174
SPECfp2006_rate = 316

AMD Opteron 6274
SPECfp2006_rate = 372

Opteron 6274 is almost 18% faster than 6174.

 

[Tsavo]

Not to be blunt, but who says Keller is anything like his old days today?

He is 53 now, not 42 or 43.

Except 53 isn't old.

 

[cytg111]

Except 53 isn't old.

Now, you got posts under your belt, so you're not new ...
Fact check; 95% probability: Of all the possible futures for AMD, ShintaiDK will paint the bleakest possible future for said company, no matter what circumstances you propose.
Very predictable and thus not very informative, either way you put it.

 

[wlee15]

Nope. All cases are treated equally as a material breach of the agreement. There is no possibility of AMD maintain access to Intel IP after being acquired.

Check here

http://forums.anandtech.com/showpost.php?p=34190679&postcount=45

That is the old agreement.

Here's the new one.

http://www.sec.gov/Archives/edgar/data/2488/000119312509236705/dex102.htm



3.4 Intel Copyright License to AMD. Subject to the terms of this Agreement, including without limitation Section 5.2(e), Intel grants to AMD, for use in or with an AMD Licensed Product, licenses under Intel’s copyrights in any Processor instruction mnemonic for an instruction developed by Intel, and the related opcodes, instruction operand mnemonics, byte format depictions and short form description (not to exceed 100 words) for those instructions, to copy, have copied, import, prepare derivative works of, perform, display and sell or otherwise distribute such mnemonics, opcodes and descriptions in user manuals and other technical documentation. No other copyright license to AMD is provided by this Agreement other than as set forth in this paragraph, either directly or by implication or estoppel.

3.5 AMD Copyright License to Intel. Subject to the terms of this Agreement, including without limitation Section 5.2(e), AMD grants to Intel, for use in or with an Intel Licensed Product, licenses under AMD’s copyrights in any Processor instruction mnemonic for an instruction developed by AMD, and the related opcodes, instruction operand mnemonics, byte format depictions and short form description (not to exceed 100 words) for those instructions, to copy, have copied, import, prepare derivative works of, perform, display and sell or otherwise distribute such mnemonics, opcodes and descriptions in user manuals and other technical documentation. No other copyright license to Intel is provided by this Agreement other than as set forth in this paragraph, either directly or by implication or estoppel.

5.3 Survival. The provisions of Sections 1, 2, 3.4, 3.5, 4.1, 5.2(d), 5.3, 6 and 7 will survive any termination or expiration of this Agreement as a whole.

 

[Phynaz]

That is the old agreement.

Here's the new one.

http://www.sec.gov/Archives/edgar/data/2488/000119312509236705/dex102.htm

From your link:

In the event of any termination of this Agreement pursuant to Section 5.2(a), and subject to the provisions of Section 5.2(e), the rights and licenses granted to any terminated Licensed Party(ies), including without limitation the rights granted under Section 3.8(d), shall terminate as of the effective date of such termination, but the rights and licenses granted to the non-terminated Licensed Party(ies) (including without limitation the Terminating Party and all of its non-terminated Subsidiaries) shall survive such termination of this Agreement subject to the non-terminated Licensed Party’s(ies’) continued compliance with the terms and conditions of this Agreement.

The 2009 agreement says the same thing as the older agreement.

 

[cytg111]

..the rights and licenses granted to the non-terminated Licensed Party(ies) (including without limitation the Terminating Party and all of its non-terminated Subsidiaries) shall survive such termination of this Agreement

I've read that five times now, and it still makes zero sense to me.. in effect i feel tossed into a triple negative with five maybe's ... wtf.

 

[wlee15]

From your link:

In the event of any termination of this Agreement pursuant to Section 5.2(a), and subject to the provisions of Section 5.2(e), the rights and licenses granted to any terminated Licensed Party(ies), including without limitation the rights granted under Section 3.8(d), shall terminate as of the effective date of such termination, but the rights and licenses granted to the non-terminated Licensed Party(ies) (including without limitation the Terminating Party and all of its non-terminated Subsidiaries) shall survive such termination of this Agreement subject to the non-terminated Licensed Party’s(ies&#8217 continued compliance with the terms and conditions of this Agreement.

The 2009 agreement says the same thing as the older agreement.

(ii) In the event of any termination of this Agreement pursuant to Section 5.2(c), and subject to the provisions of Section 5.2(e), the rights and licenses granted to both Parties under this Agreement, including without limitation the rights granted under Section 3.8(d), shall terminate as of the effective date of such termination.

Edit: Also note that in the 2001 like the 2009 version bankruptcy(and the various liquidation and assignment to creditors actions) and change of control are listed under a different subsection than the material breach so the one party gets to keep all their rights doesn't apply in those cases. Also section 8.2 (which is heavily redacted) explains the exception that allow the transfer of rights can be assigned to another party.

 

[Phynaz]

It does not.

(ii) In the event of any termination of this Agreement pursuant to Section 5.2(c), and subject to the provisions of Section 5.2(e), the rights and licenses granted to both Parties under this Agreement, including without limitation the rights granted under Section 3.8(d), shall terminate as of the effective date of such termination.[/b]

So now go read read sections 5.2(c)(d), Which is what I quoted.

 

[wlee15]

So now go read read sections 5.2(c)(d), Which is what I quoted.

5.2(c)(i) does not apply here because Change of Control is not a material breach of contract. That's why they you know they list two mutually exclusive outcomes in 5.2(c)(i) and 5.2(c)(ii).

 

[Tuna-Fish]

Why would Keller move to AMD to go from being a lead architect to creating ancillary IP for ARM SOCs? I just don't get this. If AMD was building a custom ARM SOC, then OK, maybe they threw a ton of money at him, but since AMD is using a bog standard ARM SOC, Keller must be bored out of his mind.

It takes ~5 years to design a new custom CPU core. AMD is going for a licensed core *now*, but it doesn't mean that is their plan forever -- it could be that Keller is really busy designing the next gen of high-powered ARM chips, but since those won't be out in years, AMD wants to have an ARM product to sell before that.

 

[Haserath]

I've read that five times now, and it still makes zero sense to me.. in effect i feel tossed into a triple negative with five maybe's ... wtf.

In the event AMD terminates, Intel's employees implode trying to understand the patent litigation. In effect, both companies are lost in the cross-fire(pun not unintended maybe).

Or it could mean Intel keeps the rights to AMD's patents while AMD(or the company that owns them) loses rights to Intel's patents.

 

[podspi]

There was no promise from AMD. Just a gossip website saying that AMD was going to do this, without any sources identified.

I could have sworn this was not true, but as it turns out AMD didn't have steamroller on any roadmaps except APU all the way back in February (!).


Oh well. It is a shame they are slipping from that 1yr cadence, I was impressed at their 2011/2012 execution. We'll see whether Trinity 2.0 will be enough to go against Haswell...

 

[Vesku]

OBR's greatness is evident once more, as he was the first to publicly predict massive lateness for Steamloller. :awe:

Predicting Global Foundries will be late delivering production volume on a new node is not a very risky prognostication.

 

[CHADBOGA]

Predicting Global Foundries will be late delivering production volume on a new node is not a very risky prognostication.

You do run a big risk from upset fanboys. D:

 

[krumme]

Its an illusion that price will change. Simply because higher price=lower profit for Intel due to lower volume.

Well you are excused because:

1. You are not into economics, where supply, demand and price elasticity defines what you pay for your cpu

2. You are a young boy and have not experienced the pre 1995 pricing structure for cpu

3. You do not work for a huge global company with the balls and competences to dictate the technology development, controlling the oem market, buying competitors, having a huge bad ass sales force and so on.

What do you think the production cost differences is between Intel 3750 and 3770? And how do you think the 3770 is priced? There simply is not competitor for the 3770. Now imagine there was no competitor for the rest of the line. And what would happen for investment in new process nodes for the mid, high-end segments?

If you dont think competition does anything for pricing, and thereby investments and innovation, try go to North Korea.

 

[MisterMac]

Well you are excused because:

1. You are not into economics, where supply, demand and price elasticity defines what you pay for your cpu

2. You are a young boy and have not experienced the pre 1995 pricing structure for cpu

3. You do not work for a huge global company with the balls and competences to dictate the technology development, controlling the oem market, buying competitors, having a huge bad ass sales force and so on.

What do you think the production cost differences is between Intel 3750 and 3770? And how do you think the 3770 is priced? There simply is not competitor for the 3770. Now imagine there was no competitor for the rest of the line. And what would happen for investment in new process nodes for the mid, high-end segments?

If you dont think competition does anything for pricing, and thereby investments and innovation, try go to North Korea.

Of course.

The thing is - we're just not in 1995.
PC sales aren't forecasted to explode(and by that extent servers) - so there's no massive battle in the arena AMD competes in.
Also there's no cyrix or any other players.


If AMD disappeared tomorrow - Intel wouldn't do a damn thing to upset a sure profit market.
Because they want to attack new markets - intel's worst competitor in the "PC" market is Intel now.

They have to deliver good enough gains - if they want people to go haswell or broadwell. They have to also hope the ecosystem delivers enough usage of a CPU to justify it for business.

Something which Apple has done indirectly - you tell a dude with a macbook to work on a Core2 Pentium doing anything with 1 GB ram.

If you can't see the differences in the market now and in 95 - you should really get your degree(s) back asap.

 

[pelov]

 

[ShintaiDK]

Well you are excused because:

1. You are not into economics, where supply, demand and price elasticity defines what you pay for your cpu

2. You are a young boy and have not experienced the pre 1995 pricing structure for cpu

3. You do not work for a huge global company with the balls and competences to dictate the technology development, controlling the oem market, buying competitors, having a huge bad ass sales force and so on.

What do you think the production cost differences is between Intel 3750 and 3770? And how do you think the 3770 is priced? There simply is not competitor for the 3770. Now imagine there was no competitor for the rest of the line. And what would happen for investment in new process nodes for the mid, high-end segments?

If you dont think competition does anything for pricing, and thereby investments and innovation, try go to North Korea.

Since you are obviously into economics *cough*. Maybe you should tell me the ROI on factories and the volume needed to be profitable. Intel had zero competition the last 6 years and CPU price increases are simply a price index correction today. Reason is the CPus currently are priced for the maximum profit in terms of margins and volume.

I know exactly the history of CPU prices. And I know exactly why CPUs are cheaper than every today. The last price hike we had was before Core 2, since AMD was capacity constrained and could charge a high premium. But else prices have gone down with increased volume. And lower prices again increases the volume that again increases the profit. Untill the market is saturated and lowering price wont increase volume in a way that positively affect profit.

I dont know why you wish to talk about 2 SKUs only, since the entire ROI is based on all SKUs combined and their volume.

Profit = margins*volume.

Now whats best, sell 50mio CPUs at a 500$ average margin. Or to sell 400mio CPUs with an average 100$ margin. or lastly, sell 500mio CPUs with a 50$ margin?

That shouldnt be rocket science. Intel already priced it for optimal profit.

Oh...and competition and low volume was why CPU (and other components) prices was so high in the old days.

HDs havent been cheaper, yet only what..2 left? Memory havent been cheaper even tho one goes bankrupt after the other and competition dissapears.

Its a false assumption that competition is always good. Its maybe good in 7 or 8 times out of 10. But a horrible flop in the last 20-30%. In the last 20-30% you either get low quality and unreliable products in an effort to save cost. Or you get stagnation in a segment due to the risk vs reward factor.

Of course.

The thing is - we're just not in 1995.
PC sales aren't forecasted to explode(and by that extent servers) - so there's no massive battle in the arena AMD competes in.
Also there's no cyrix or any other players.


If AMD disappeared tomorrow - Intel wouldn't do a damn thing to upset a sure profit market.
Because they want to attack new markets - intel's worst competitor in the "PC" market is Intel now.

They have to deliver good enough gains - if they want people to go haswell or broadwell. They have to also hope the ecosystem delivers enough usage of a CPU to justify it for business.

Something which Apple has done indirectly - you tell a dude with a macbook to work on a Core2 Pentium doing anything with 1 GB ram.

If you can't see the differences in the market now and in 95 - you should really get your degree(s) back asap.

Exactly. Intel both needs give to the consumer a reason to upgrade (innovation) and to sell it at a price the consumer is willing to pay and think its worth to upgrade for. Else people will just keep using their old CPUs and Intel will go belly up.

 

[Ferzerp]

While ShintaiDK is typically extremely biased, the core of what he says is correct.

Intel needs to maintain volume, and since the desktop/laptop market is nearly saturated, the only way to do that is via product replacement. As CPUs have such long lifespans, the only way to keep the money flowing is to offer enough of an upgrade to entice users to replace their current systems at an acceptible price. As strong as the company is, it wouldn't take many years of a $1,000 low end CPU (this is the typical boogeyman those arguing for welfare purchases of AMD product use, right?) to ruin the company. They'll also be under intense federal scrutiny if/when they're the only player left in the market, though the effect of that could be little or lots based on how useless our government is.

We've been hearing "CPUs are fast enough". Well, if prices got out of hand, yes, they would be nearly universally decided to be fast enough, and Intel's cash cow would be gone. They need that volume.

What you will likely not see, though, are any core2 level performance increases. In the absence of competition, they *are* incentivized to only advance at the rate that maximizes profit, and this does, sadly, mean steady incremental improvements at a flat price.

The doom and gloom stories assume purely inelastic demand. CPU longenvity and the current baseline being plenty for most applications actually makes CPU demand extremely price elastic.