AMD CPU speculation... and expert conjecture

[juanrga]

But it seems more interesting to use a R9 290/290X with Kaveri. Why? Because those two cards already support "system unified addressing" and may be the perfect complement for a hUMA HSA APU.
Finally, it is worth to notice that AMD has always made software. See the software section in their Wikipedia article. AMD now reinventing itself as a "gaming company" is focusing the software development to gaming.

yeah.. a more powerful gfx card will always deliver more gfx prowess. i combo'ed the most possible candidates for amd dual gfx.

But I didn't mention the 290/290X because are more powerful but "Because those two cards already support "system unified addressing" and may be the perfect complement for a hUMA HSA APU"

See how you don't read?

 

[blackkstar]

We won't be able to make any concrete deductions on how bad GF screwed AMD with bulk and SOI.

Yeah, the 7850k is not that far behind Trinity in clock speed, but look at what happened with Richland overclocking. I can nearly guarantee you that you won't be seeing a single 7850k breaking 5ghz on air. And Richlands that did that were common.

To be quiet honest I would be surprised if Kaveri broke mid 4ghz range for maximum overclocks. I have a feeling that AMD used the clock speed tweaks they used on Richland to salvage a bulk product. Meaning that whatever AMD did that let Richland break 5ghz while Trinity had no chance was applied to Kaveri to prevent Kaveri from having catastrophic clock speed. RCM comes to mind and I have a very good feeling if AMD didn't beta test what they did to get clock speeds up with Richland than we'd be looking at a 2ghz range Kaveri high end model.

I also don't expect Kaveri to overclock well at all because of what AMD did with FX. Clearly it was a better choice to release PD with a 20% clockspeed improvement than to release SR FX on bulk. That alone should speak volumes of why there is no SR FX.

I'm assuming things kind of went like this

1. AMD starts talking about Steamroller on AM3+ using 28nm SOI
2. Glofo completely screws it up with delays, tells AMD they can get their big 28nm SOI chip in mid to late 2014.
3. AMD decided to either go bulk with SR FX or just not to it at all
4. They find out clockspeed and overclocking hit that comes with bulk wouldn't make the chip 10% faster
5. Release PD parts with 20%+ clockspeed bump (FX 9000 series)

Take a look at S|A, Kaveri missed targets all over. It didn't miss targets because bulk is just as good as SOI.

I'm standing by the fact that AMD got completely screwed by GloFo and they can't make big parts, so they're trying to pretend like they don't care about making big parts to cover up the massive failure that's happening. Look at the video Cazalan posted, the guy is not excited about it at all. Low end product.

Kaveri is decent but it's not great, and it feels like AMD is aware of this by the fact that it's hyping up HSA instead of Kaveri.

 

[jdwii]

So i was finally able to use a piledriver base APU, my brother has it i told him to check it out for his new 400$ PC rig. He got an A10 5800K and 1866 dual channel ram. I installed multiple games what i find funny is windows really does treat the processor like a hyperthreaded CPU, I noticed the 1st and 3rd core was being used 40% and the other 2 where doing nothing. Also what i found amazing is a A10 with dual 1866 ram can actually play GTA4 at 720P at medium settings same as 360 and had decent frame rates With barely little trying i was able to overclock the APU to 4.2ghz with 1.325V and its 100% stable could shoot for higher clocks but again this is my brothers PC. This APU play's Call of duty MW3 in 720P high or 1080P medium with smooth gameplay, I kept getting more of my games hitman absolution was playing well at 1080P medium Deus ex played perfect and BF3 played decent. I have to admit this is why i like Amd over Intel for 400$ my brother can play fantastic games the I3 would barely break 15FPS unless i spend more money on an 80$ video card and get a lower quality board and by that time I could just get a 6 core FX and a 7770.

Still though i would not want this APU in my gaming rig and i would not recommend it on anything but midrange gaming rigs i would never put anything higher than a 150$(7850) Video card on the A10, i say this because at times i saw his processor load 2 cores to 100% and it happened on second life and a couple other games. If juanrga is correct and the 4 core kaveri can beat the 8350 in any 4 core load then yes a APU would work in a higher-end gaming setup and would probably kick some butt, when i overclocked the A10 to 4.2Ghz even the GPU score went up so yeah i think the A10 Trinity is bottlenecked by its own CPU.

 

[de5_Roy]

See how the PS4 regulates heat with these thermal images
http://vr-zone.com/articles/thermal-images-show-ps4-regulates-heat/64084.html

Spoiler

But I didn't mention the 290/290X because are more powerful but "Because those two cards already support "system unified addressing" and may be the perfect complement for a hUMA HSA APU"

See how you don't read?

huh? i know nothing about system unified addressing and i haven't seen it in action. that's why i ignored it, instead considered the gfx card(did i specify a model in my reply?) as traditional discreet gfx cards, in an Imaginary pc. now i see that your reply was a bait to throw insults at me. well, i regretfully admit that took the bait, you win.

 

[palladin9479]

So i was finally able to use a piledriver base APU, my brother has it i told him to check it out for his new 400$ PC rig. He got an A10 5800K and 1866 dual channel ram. I installed multiple games what i find funny is windows really does treat the processor like a hyperthreaded CPU, I noticed the 1st and 3rd core was being used 40% and the other 2 where doing nothing. Also what i found amazing is a A10 with dual 1866 ram can actually play GTA4 at 720P at medium settings same as 360 and had decent frame rates With barely little trying i was able to overclock the APU to 4.2ghz with 1.325V and its 100% stable could shoot for higher clocks but again this is my brothers PC. This APU play's Call of duty MW3 in 720P high or 1080P medium with smooth gameplay, I kept getting more of my games hitman absolution was playing well at 1080P medium Deus ex played perfect and BF3 played decent. I have to admit this is why i like Amd over Intel for 400$ my brother can play fantastic games the I3 would barely break 15FPS unless i spend more money on an 80$ video card and get a lower quality board and by that time I could just get a 6 core FX and a 7770.

Still though i would not want this APU in my gaming rig and i would not recommend it on anything but midrange gaming rigs i would never put anything higher than a 150$(7850) Video card on the A10, i say this because at times i saw his processor load 2 cores to 100% and it happened on second life and a couple other games. If juanrga is correct and the 4 core kaveri can beat the 8350 in any 4 core load then yes a APU would work in a higher-end gaming setup and would probably kick some butt, when i overclocked the A10 to 4.2Ghz even the GPU score went up so yeah i think the A10 Trinity is bottlenecked by its own CPU.

Yep APU's are amazing for SFF and budget systems. They are more economical then purchasing a separate CPU and dGPU but they come with the limitation that you need to design around them and get decent speed memory.

You should be able to bump the NM to 2.4Ghz and using 2166 memory. My 8350 can do 2.4Ghz CPU-NB which improves the iGPUs performance. Don't bother bumping up the iGPU's clock speed as it won't help much, the memory is where the bottleneck is happening not the core GPU clock.

 

[juanrga]

We won't be able to make any concrete deductions on how bad GF screwed AMD with bulk and SOI.

We know one thing for sure. If AMD had selected SOI, they couldn't present Kaveri at APU13 and start shipping it now, because GF has failed to execute their 28nm SOI plans. Their Fab 8 is not ready for mass production and Samsung is coming in their help, taking the control of the R&D dept.

Assuming GF gets its problems solved for late 2014 it would be... too late (haha), because then Kaveri would be competing against Haswell successor (Broadwell) and Intel promises a much more powerful iGPU for Broadwell. Then AMD would cancel Kaveri and release Carrizo.

Luckily AMD selected bulk and can now start to ship Kaveri for official launch at January. Also AMD is now tappingout 20nm chips on bulk and 16nm on bulk FINFET.

Yeah, the 7850k is not that far behind Trinity in clock speed, but look at what happened with Richland overclocking. I can nearly guarantee you that you won't be seeing a single 7850k breaking 5ghz on air. And Richlands that did that were common.

To be quiet honest I would be surprised if Kaveri broke mid 4ghz range for maximum overclocks. I have a feeling that AMD used the clock speed tweaks they used on Richland to salvage a bulk product. Meaning that whatever AMD did that let Richland break 5ghz while Trinity had no chance was applied to Kaveri to prevent Kaveri from having catastrophic clock speed. RCM comes to mind and I have a very good feeling if AMD didn't beta test what they did to get clock speeds up with Richland than we'd be looking at a 2ghz range Kaveri high end model.

As said before:

The CPU in Kaveri will be overclocked beyond 4.5 GHZ without problems. It will not break worldwide records either.

The former phrase is based in leaked docs. The late phrase means that I don't expect Kaveri to break the 8GHz mark. ;-)

I also don't expect Kaveri to overclock well at all because of what AMD did with FX. Clearly it was a better choice to release PD with a 20% clockspeed improvement than to release SR FX on bulk. That alone should speak volumes of why there is no SR FX.

I'm assuming things kind of went like this

1. AMD starts talking about Steamroller on AM3+ using 28nm SOI
2. Glofo completely screws it up with delays, tells AMD they can get their big 28nm SOI chip in mid to late 2014.
3. AMD decided to either go bulk with SR FX or just not to it at all
4. They find out clockspeed and overclocking hit that comes with bulk wouldn't make the chip 10% faster
5. Release PD parts with 20%+ clockspeed bump (FX 9000 series)

Before I speculated with AMD taking one of two possible ways. The second was "abandon FX-4000 and use Warsaw-like dies for FX-6000/8000/9000 refresh".

Warsaw promises further reduction in power consumption compared to Opteron. I assume that are using fully enabled resonant clock mesh. Efficiency in the desktop is secondary, but reduction in power consumption could be used for increasing base/turbo clocks and the ability to overclock beyond 5GHz on air.

But the 2104 desktop roadmap shows that they are not doing that.

Then the big question is why is not AMD reusing the dies that are fabricating for Warsaw also for the FX line? Why the old FX-line is extended to 2014 without any refresh? Why the FX-4000 series continue being sold when the 4-core Opterons are abandoned and Warsaw is 12/16 core?

The more likely answer to all that is "lack of demand". We know that the FX line didn't sell well. Therefore, AMD must have an inventory of old products that they want to sell and, at the same time, there is no room for a refresh that 'nobody' would purchase. By nobody I mean less than 1% of total gamers or so.

We also have leaked docs. that said that the 9590 was the last FX chip.

My belief is that AMD is extending the FX line up to 2014 for replacing it with an ultra-high-end APU in 2015 (FM3 socket). Time will say.

Take a look at S|A, Kaveri missed targets all over. It didn't miss targets because bulk is just as good as SOI.

The S|A article is overly pessimistic. Several of us (including myself) discussed it in S|A forums.

 

[juanrga]

Slides from the Thief talk at APU13

http://i.imgur.com/nt5217p.png

Confirms that said here before:
- MANTLE is not specific for AMD hardware.
- High-end cards such as R9-290X are unused, ~50% idle is not unusual in current games.
- MANTLE eliminates several GPU and CPU bottlenecks
- "PC gamer will be able to achieve higher frame-rate with a lower spec CPU!"

 

[juanrga]

The reason why the GPU cores of Kaveri are not 900MHz but lower is related to bandwidth. The original Kaveri APU was planned to come with GDDR5 support (someone has said me that the final Kaveri die contains a disabled GDDR5 memory controller). If this is true the decision to abandon GDDR5 was made very very late.

Without GDDR5 the iGPU is memory bandwidth limited and increasing the frequency by a 25% (or more) will increase the power and the heat much more than the graphics performance.

There are data that support this hypothesis.

Kaveri has up to 512 shaders. This is the same number found in the HD 7750. This old card comes with two kind of memory DDR3 and GDDR5. The GDDR5 version has the cores @ 900MHz whereas the DDR3 uses lower frequency of 800GHz.

Therefore, this seems to conform that the drop in Kaveri iGPU frequencies is a consequence of the abandon of GDDR5 support and not of bulk vs SOI.

The difference in frequencies between the Kaveri iGPU and the DDR3 7750 is of 11%, but this cannot be attributed to SOI vs bulk because both are bulk. It could be attributed to Kaveri having a lower TPD.

 

[8350rocks]

We won't be able to make any concrete deductions on how bad GF screwed AMD with bulk and SOI.

We know one thing for sure. If AMD had selected SOI, they couldn't present Kaveri at APU13 and start shipping it now, because GF has failed to execute their 28nm SOI plans. Their Fab 8 is not ready for mass production and Samsung is coming in their help, taking the control of the R&D dept.

Assuming GF gets its problems solved for late 2014 it would be... too late (haha), because then Kaveri would be competing against Haswell successor (Broadwell) and Intel promises a much more powerful iGPU for Broadwell. Then AMD would cancel Kaveri and release Carrizo.

Luckily AMD selected bulk and can now start to ship Kaveri for official launch at January. Also AMD is now tappingout 20nm chips on bulk and 16nm on bulk FINFET.

Yeah, the 7850k is not that far behind Trinity in clock speed, but look at what happened with Richland overclocking. I can nearly guarantee you that you won't be seeing a single 7850k breaking 5ghz on air. And Richlands that did that were common.

To be quiet honest I would be surprised if Kaveri broke mid 4ghz range for maximum overclocks. I have a feeling that AMD used the clock speed tweaks they used on Richland to salvage a bulk product. Meaning that whatever AMD did that let Richland break 5ghz while Trinity had no chance was applied to Kaveri to prevent Kaveri from having catastrophic clock speed. RCM comes to mind and I have a very good feeling if AMD didn't beta test what they did to get clock speeds up with Richland than we'd be looking at a 2ghz range Kaveri high end model.

As said before:

The CPU in Kaveri will be overclocked beyond 4.5 GHZ without problems. It will not break worldwide records either.

The former phrase is based in leaked docs. The late phrase means that I don't expect Kaveri to break the 8GHz mark. ;-)

I also don't expect Kaveri to overclock well at all because of what AMD did with FX. Clearly it was a better choice to release PD with a 20% clockspeed improvement than to release SR FX on bulk. That alone should speak volumes of why there is no SR FX.

I'm assuming things kind of went like this

1. AMD starts talking about Steamroller on AM3+ using 28nm SOI
2. Glofo completely screws it up with delays, tells AMD they can get their big 28nm SOI chip in mid to late 2014.
3. AMD decided to either go bulk with SR FX or just not to it at all
4. They find out clockspeed and overclocking hit that comes with bulk wouldn't make the chip 10% faster
5. Release PD parts with 20%+ clockspeed bump (FX 9000 series)

Before I speculated with AMD taking one of two possible ways. The second was "abandon FX-4000 and use Warsaw-like dies for FX-6000/8000/9000 refresh".

Warsaw promises further reduction in power consumption compared to Opteron. I assume that are using fully enabled resonant clock mesh. Efficiency in the desktop is secondary, but reduction in power consumption could be used for increasing base/turbo clocks and the ability to overclock beyond 5GHz on air.

But the 2104 desktop roadmap shows that they are not doing that.

Then the big question is why is not AMD reusing the dies that are fabricating for Warsaw also for the FX line? Why the old FX-line is extended to 2014 without any refresh? Why the FX-4000 series continue being sold when the 4-core Opterons are abandoned and Warsaw is 12/16 core?

The more likely answer to all that is "lack of demand". We know that the FX line didn't sell well. Therefore, AMD must have an inventory of old products that they want to sell and, at the same time, there is no room for a refresh that 'nobody' would purchase. By nobody I mean less than 1% of total gamers or so.

We also have leaked docs. that said that the 9590 was the last FX chip.

My belief is that AMD is extending the FX line up to 2014 for replacing it with an ultra-high-end APU in 2015 (FM3 socket). Time will say.

Take a look at S|A, Kaveri missed targets all over. It didn't miss targets because bulk is just as good as SOI.

The S|A article is overly pessimistic. Several of us (including myself) discussed it in S|A forums.

I want proof of the following:
-AMD said they are replacing FX with Kaveri
-AMD said they are going to 20nm bulk
-AMD said Kaveri APUs are high end parts

Everything I have seen so far is all contradicting your claims.

 

[blackkstar]

We won't be able to make any concrete deductions on how bad GF screwed AMD with bulk and SOI.

We know one thing for sure. If AMD had selected SOI, they couldn't present Kaveri at APU13 and start shipping it now, because GF has failed to execute their 28nm SOI plans. Their Fab 8 is not ready for mass production and Samsung is coming in their help, taking the control of the R&D dept.

Assuming GF gets its problems solved for late 2014 it would be... too late (haha), because then Kaveri would be competing against Haswell successor (Broadwell) and Intel promises a much more powerful iGPU for Broadwell. Then AMD would cancel Kaveri and release Carrizo.

Luckily AMD selected bulk and can now start to ship Kaveri for official launch at January. Also AMD is now tappingout 20nm chips on bulk and 16nm on bulk FINFET.

Yeah, the 7850k is not that far behind Trinity in clock speed, but look at what happened with Richland overclocking. I can nearly guarantee you that you won't be seeing a single 7850k breaking 5ghz on air. And Richlands that did that were common.

To be quiet honest I would be surprised if Kaveri broke mid 4ghz range for maximum overclocks. I have a feeling that AMD used the clock speed tweaks they used on Richland to salvage a bulk product. Meaning that whatever AMD did that let Richland break 5ghz while Trinity had no chance was applied to Kaveri to prevent Kaveri from having catastrophic clock speed. RCM comes to mind and I have a very good feeling if AMD didn't beta test what they did to get clock speeds up with Richland than we'd be looking at a 2ghz range Kaveri high end model.

As said before:

The CPU in Kaveri will be overclocked beyond 4.5 GHZ without problems. It will not break worldwide records either.

The former phrase is based in leaked docs. The late phrase means that I don't expect Kaveri to break the 8GHz mark. ;-)

I also don't expect Kaveri to overclock well at all because of what AMD did with FX. Clearly it was a better choice to release PD with a 20% clockspeed improvement than to release SR FX on bulk. That alone should speak volumes of why there is no SR FX.

I'm assuming things kind of went like this

1. AMD starts talking about Steamroller on AM3+ using 28nm SOI
2. Glofo completely screws it up with delays, tells AMD they can get their big 28nm SOI chip in mid to late 2014.
3. AMD decided to either go bulk with SR FX or just not to it at all
4. They find out clockspeed and overclocking hit that comes with bulk wouldn't make the chip 10% faster
5. Release PD parts with 20%+ clockspeed bump (FX 9000 series)

Before I speculated with AMD taking one of two possible ways. The second was "abandon FX-4000 and use Warsaw-like dies for FX-6000/8000/9000 refresh".

Warsaw promises further reduction in power consumption compared to Opteron. I assume that are using fully enabled resonant clock mesh. Efficiency in the desktop is secondary, but reduction in power consumption could be used for increasing base/turbo clocks and the ability to overclock beyond 5GHz on air.

But the 2104 desktop roadmap shows that they are not doing that.

Then the big question is why is not AMD reusing the dies that are fabricating for Warsaw also for the FX line? Why the old FX-line is extended to 2014 without any refresh? Why the FX-4000 series continue being sold when the 4-core Opterons are abandoned and Warsaw is 12/16 core?

The more likely answer to all that is "lack of demand". We know that the FX line didn't sell well. Therefore, AMD must have an inventory of old products that they want to sell and, at the same time, there is no room for a refresh that 'nobody' would purchase. By nobody I mean less than 1% of total gamers or so.

We also have leaked docs. that said that the 9590 was the last FX chip.

My belief is that AMD is extending the FX line up to 2014 for replacing it with an ultra-high-end APU in 2015 (FM3 socket). Time will say.

Take a look at S|A, Kaveri missed targets all over. It didn't miss targets because bulk is just as good as SOI.

The S|A article is overly pessimistic. Several of us (including myself) discussed it in S|A forums.

The thing is that you are only stipulating that there is a difference between Warsaw and PD. It could just be a rebrand with nothing done to the chip at all, and PD DT might just be Warsaws.

I am asking around in forums from recent FX purchasers to check batch numbers. You can tell what year and week the chip came from from that number and if the chips are old, AMD probably has a lot of inventory. If the batch number isn't indicating the chip is old, AMD might not have old parts laying around.

Just for reference, I bought my FX 8350 in mid October and it was a 1236, which means mid summer for me. So a few months of delay before chip was born to customer's computer.

If anyone else bought an FX recently please chip in with your batch number.

 

[noob2222]

^^@8350

ill add to that.

what fab is doing 16nm

Broadwell isn't even a DT chip. how is it competing in the DT market?

30% of AMD chips sold were FX ... how is that "didn't sell well"

Steam survey doesn't account for 100% of the market.

 

[blackkstar]

Getting a few answers from people who bought FX lately.

They are getting 1330s range batch numbers.

People who bought FX when it first came out where getting 1220 to 1240 range.

It sounds to me like AMD is still moving chips. If batch numbers are still that high there is absolutely no way they could meet demand with existing inventory for at least a year and a month, which is what you're suggesting.

If that were the case we'd still be seeing 1200s and early 1300s shipping from major retailers.

 

[8350rocks]

My batch number is 1311, though it was from quite a bit earlier in the year...(early spring/April-ish)

 

[jdwii]

Trying to decide what to get a New 8350fx or a new 280x? I have a Phenom II X6 OC 3.9ghz and a 6950 2GB card.

 

[con635]

http://www.youtube.com/watch?v=axyHkKn_e80
mmm dont know about this, fake??? Says it kaveri 7850, bf, med, high, ultra

 

[8350rocks]

http://www.youtube.com/watch?v=axyHkKn_e80
mmm dont know about this, fake??? Says it kaveri 7850, bf, med, high, ultra

I usually have a salt shaker handy for youtube stuff like that...typically you can't get any realistically good idea of what they are actually using.

 

[8350rocks]

Trying to decide what to get a New 8350fx or a new 280x? I have a Phenom II X6 OC 3.9ghz and a 6950 2GB card.

Your 6950 isn't bad...honestly...it's probably about even with a 7870 (give or take) would be my guess, so you're not in bad shape there.

You could get a new 8350 for less money...

Tough call, I would think the 8350 might be more benefit just for the newer instruction sets supported; however, a 280x would support MANTLE for BF4 and all the other games coming with it.

 

[tracker45]

Any word on how mantle performs yet ?

And does a game have to be made with Mantle for it to be "activated".
Does this mean older games cannot get a performance boost ?

 

[tracker45]

http://www.youtube.com/watch?v=axyHkKn_e80
mmm dont know about this, fake??? Says it kaveri 7850, bf, med, high, ultra

Would Mantle increase the fps on an APU ?

 

[8350rocks]

No word on performance yet.

Yes, MANTLE has to be implemented, that means older games would not get a boost from it.

Theoretically, MANTLE should boost performance on Kaveri APUs.

 

[Cazalan]

Just for reference, I bought my FX 8350 in mid October and it was a 1236, which means mid summer for me. So a few months of delay before chip was born to customer's computer.

The 6300 Opterons (and thus FX series) are server die/parts and have long life production. They'll be available for at least 4 more years according to AMD website with an EOL plan to add an additional 2 years. That puts you through 2017 at a minimum.

http://www.amd.com/us/Documents/6000_Series_product_brief.pdf

AMD has some high inventory but they're certainly not sitting on 4 years worth of FX/Opteron parts.

 

[juanrga]

We won't be able to make any concrete deductions on how bad GF screwed AMD with bulk and SOI.

We know one thing for sure. If AMD had selected SOI, they couldn't present Kaveri at APU13 and start shipping it now, because GF has failed to execute their 28nm SOI plans. Their Fab 8 is not ready for mass production and Samsung is coming in their help, taking the control of the R&D dept.

Assuming GF gets its problems solved for late 2014 it would be... too late (haha), because then Kaveri would be competing against Haswell successor (Broadwell) and Intel promises a much more powerful iGPU for Broadwell. Then AMD would cancel Kaveri and release Carrizo.

Luckily AMD selected bulk and can now start to ship Kaveri for official launch at January. Also AMD is now tappingout 20nm chips on bulk and 16nm on bulk FINFET.

Yeah, the 7850k is not that far behind Trinity in clock speed, but look at what happened with Richland overclocking. I can nearly guarantee you that you won't be seeing a single 7850k breaking 5ghz on air. And Richlands that did that were common.

To be quiet honest I would be surprised if Kaveri broke mid 4ghz range for maximum overclocks. I have a feeling that AMD used the clock speed tweaks they used on Richland to salvage a bulk product. Meaning that whatever AMD did that let Richland break 5ghz while Trinity had no chance was applied to Kaveri to prevent Kaveri from having catastrophic clock speed. RCM comes to mind and I have a very good feeling if AMD didn't beta test what they did to get clock speeds up with Richland than we'd be looking at a 2ghz range Kaveri high end model.

As said before:

The CPU in Kaveri will be overclocked beyond 4.5 GHZ without problems. It will not break worldwide records either.

The former phrase is based in leaked docs. The late phrase means that I don't expect Kaveri to break the 8GHz mark. ;-)

I also don't expect Kaveri to overclock well at all because of what AMD did with FX. Clearly it was a better choice to release PD with a 20% clockspeed improvement than to release SR FX on bulk. That alone should speak volumes of why there is no SR FX.

I'm assuming things kind of went like this

1. AMD starts talking about Steamroller on AM3+ using 28nm SOI
2. Glofo completely screws it up with delays, tells AMD they can get their big 28nm SOI chip in mid to late 2014.
3. AMD decided to either go bulk with SR FX or just not to it at all
4. They find out clockspeed and overclocking hit that comes with bulk wouldn't make the chip 10% faster
5. Release PD parts with 20%+ clockspeed bump (FX 9000 series)

Before I speculated with AMD taking one of two possible ways. The second was "abandon FX-4000 and use Warsaw-like dies for FX-6000/8000/9000 refresh".

Warsaw promises further reduction in power consumption compared to Opteron. I assume that are using fully enabled resonant clock mesh. Efficiency in the desktop is secondary, but reduction in power consumption could be used for increasing base/turbo clocks and the ability to overclock beyond 5GHz on air.

But the 2104 desktop roadmap shows that they are not doing that.

Then the big question is why is not AMD reusing the dies that are fabricating for Warsaw also for the FX line? Why the old FX-line is extended to 2014 without any refresh? Why the FX-4000 series continue being sold when the 4-core Opterons are abandoned and Warsaw is 12/16 core?

The more likely answer to all that is "lack of demand". We know that the FX line didn't sell well. Therefore, AMD must have an inventory of old products that they want to sell and, at the same time, there is no room for a refresh that 'nobody' would purchase. By nobody I mean less than 1% of total gamers or so.

We also have leaked docs. that said that the 9590 was the last FX chip.

My belief is that AMD is extending the FX line up to 2014 for replacing it with an ultra-high-end APU in 2015 (FM3 socket). Time will say.

Take a look at S|A, Kaveri missed targets all over. It didn't miss targets because bulk is just as good as SOI.

The S|A article is overly pessimistic. Several of us (including myself) discussed it in S|A forums.

I want proof of the following:
-AMD said they are replacing FX with Kaveri
-AMD said they are going to 20nm bulk
-AMD said Kaveri APUs are high end parts

Everything I have seen so far is all contradicting your claims.

Do you read? Because in the same message that you reply I wrote:

the old FX-line is extended to 2014 without any refresh [...]

My belief is that AMD is extending the FX line up to 2014 for replacing it with an ultra-high-end APU in 2015 (FM3 socket). Time will say.

About the second question, it was also answered before. but in another post. I will repeat this quote from AMD:

We are typically at the leading edge across the technology nodes. We are fully top-top-bottom in 28nm now across all of our products, and we are transitioning to both 20nm and to FinFETs over the next couple of quarters in terms of designs. So we will continue to do that across our foundry partners. […] We will do 20nm first and then we will go to FinFETs

The third of your questions was also answered before. I will repeat again part of what I already said...

Unlike you, I don't need AMD saying me that it is high-end / performance APU, I can obtain that conclusion by myself.

FX-8350: 256 GFLOP
Kaveri: 856 GFLOP = 118+737 GFLOP

856 >> 256 which means that the new APU will outperform the old 8350 CPU with HSA enabled software.

As shown in my BSN* article, the CPU in Kaveri will perform like an SB/IB i5 with ordinary software.

As shown at APU13, the Kaveri CPU can perform like a 8350 at stock or a i7-4770k at stock with MANTLE enabled software.

As leaked benchmarks show, the Kaveri CPU will outperform a 8350 in integer workloads with software up to four threads

But in any case since you ask about what AMD said, here is what AMD says directly or indirectly:

Su said the Kaveri chips will bring improvements over competing processors from larger rival Intel in such areas as graphics capabilities, overall performance and power efficiency. Kaveri also signals a significant step forward in AMD's heterogeneous computing vision of ramping up the parallel computing capabilities in system through the tight integration of multi-core CPUs and GPUs on the same piece of silicon."We believe this really unlocks a different level of … computing," Su said during her opening keynote address at the developer conference.

The demand for greater parallel computing capabilities is building through all levels of computing, from mobile devices and PCs to cloud servers and high-performance computing systems, she said. Driving this demand are such factors as cloud computing, big data and greater mobility. AMD's APU architecture is aimed at addressing that demand.

"The days of single-threaded performance are over," Su said. She noted that between 2011 and 2012, AMD shipped more than 80 million APUs, a number she expects to grow to more than 150 million in 2014 and more than 300 million within a few years after that, illustrating the growing demand for such heterogeneous computing capabilities in the industry.

AMD has recently updated its product roadmap and is set to release its Hawaii-based GPUs at the end of September, Kaveri-based APUs for the high-end segment and Kabini-based APUs for the entry-level segment in the first quarter of 2014, according to sources from the upstream supply chain.

Aimed at high-end desktops and laptops, Kaveri is the follow-up to last year's top-of-the-line Richland series and marks AMD's first heterogeneous APU - where a CPU and GPU are housed on the same piece of silicon to boost performance and speed.

[...]

Further details weren't available, but AMD said the new APU is due in the first six months of next year and that it will demo devices at CES in January.

 

[juanrga]

We won't be able to make any concrete deductions on how bad GF screwed AMD with bulk and SOI.

We know one thing for sure. If AMD had selected SOI, they couldn't present Kaveri at APU13 and start shipping it now, because GF has failed to execute their 28nm SOI plans. Their Fab 8 is not ready for mass production and Samsung is coming in their help, taking the control of the R&D dept.

Assuming GF gets its problems solved for late 2014 it would be... too late (haha), because then Kaveri would be competing against Haswell successor (Broadwell) and Intel promises a much more powerful iGPU for Broadwell. Then AMD would cancel Kaveri and release Carrizo.

Luckily AMD selected bulk and can now start to ship Kaveri for official launch at January. Also AMD is now tappingout 20nm chips on bulk and 16nm on bulk FINFET.

Yeah, the 7850k is not that far behind Trinity in clock speed, but look at what happened with Richland overclocking. I can nearly guarantee you that you won't be seeing a single 7850k breaking 5ghz on air. And Richlands that did that were common.

To be quiet honest I would be surprised if Kaveri broke mid 4ghz range for maximum overclocks. I have a feeling that AMD used the clock speed tweaks they used on Richland to salvage a bulk product. Meaning that whatever AMD did that let Richland break 5ghz while Trinity had no chance was applied to Kaveri to prevent Kaveri from having catastrophic clock speed. RCM comes to mind and I have a very good feeling if AMD didn't beta test what they did to get clock speeds up with Richland than we'd be looking at a 2ghz range Kaveri high end model.

As said before:

The CPU in Kaveri will be overclocked beyond 4.5 GHZ without problems. It will not break worldwide records either.

The former phrase is based in leaked docs. The late phrase means that I don't expect Kaveri to break the 8GHz mark. ;-)

I also don't expect Kaveri to overclock well at all because of what AMD did with FX. Clearly it was a better choice to release PD with a 20% clockspeed improvement than to release SR FX on bulk. That alone should speak volumes of why there is no SR FX.

I'm assuming things kind of went like this

1. AMD starts talking about Steamroller on AM3+ using 28nm SOI
2. Glofo completely screws it up with delays, tells AMD they can get their big 28nm SOI chip in mid to late 2014.
3. AMD decided to either go bulk with SR FX or just not to it at all
4. They find out clockspeed and overclocking hit that comes with bulk wouldn't make the chip 10% faster
5. Release PD parts with 20%+ clockspeed bump (FX 9000 series)

Before I speculated with AMD taking one of two possible ways. The second was "abandon FX-4000 and use Warsaw-like dies for FX-6000/8000/9000 refresh".

Warsaw promises further reduction in power consumption compared to Opteron. I assume that are using fully enabled resonant clock mesh. Efficiency in the desktop is secondary, but reduction in power consumption could be used for increasing base/turbo clocks and the ability to overclock beyond 5GHz on air.

But the 2104 desktop roadmap shows that they are not doing that.

Then the big question is why is not AMD reusing the dies that are fabricating for Warsaw also for the FX line? Why the old FX-line is extended to 2014 without any refresh? Why the FX-4000 series continue being sold when the 4-core Opterons are abandoned and Warsaw is 12/16 core?

The more likely answer to all that is "lack of demand". We know that the FX line didn't sell well. Therefore, AMD must have an inventory of old products that they want to sell and, at the same time, there is no room for a refresh that 'nobody' would purchase. By nobody I mean less than 1% of total gamers or so.

We also have leaked docs. that said that the 9590 was the last FX chip.

My belief is that AMD is extending the FX line up to 2014 for replacing it with an ultra-high-end APU in 2015 (FM3 socket). Time will say.

Take a look at S|A, Kaveri missed targets all over. It didn't miss targets because bulk is just as good as SOI.

The S|A article is overly pessimistic. Several of us (including myself) discussed it in S|A forums.

The thing is that you are only stipulating that there is a difference between Warsaw and PD. It could just be a rebrand with nothing done to the chip at all, and PD DT might just be Warsaws.

From AMD news release:

“Warsaw” will provide significantly improved performance-per-watt over today’s AMD Opteron™ 6300 family.

[...]

Designed for enterprise workloads, it will offer improved performance-per-watt, which drives down the cost of owning a “Warsaw”-based server while enabling seamless migration from the AMD Opteron 6300 Series family.

From S|A article on AMD presentation of server roadmap:

In my chat with Mr. Feldman he seemed confident in AMD’s ability to make “Warsaw” a worthwhile refresh from AMD current “Abu Dubai” chips, by citing a focus on increasing efficiency. “Warsaw” will be similar to “Richland” in the sense that it takes advantage of process improvements to increase based clock speeds, and more aggressive power management to reduce consumption and offer higher TurboBoost clocks.

As I said above I am confident that Warsaw includes a fully enabled resonant clock mesh technology.

 

[juanrga]

^^@8350

ill add to that.

what fab is doing 16nm

Broadwell isn't even a DT chip. how is it competing in the DT market?

30% of AMD chips sold were FX ... how is that "didn't sell well"

Steam survey doesn't account for 100% of the market.

My mistake it is not 16nm, AMD is tapping out 14nm FinFET.

Kaveri is for DT and laptop, Moreover, Broadwell-D is for desktop.

Source for your "30% of AMD chips sold were FX"?

 

[szatkus]

FX-8350: 256 GFLOP
Kaveri: 856 GFLOP = 118+737 GFLOP

It's just some kind of marketing approximation.

"GFLOPs calculations developed by AMD Performance Labs measuring compute capacity for the AMD A10-5800K desktop APU which is 736 GFLOPS. AMD GFLOPs calculated using GFLOPs = CPU GFLOPs + GPU GFLOPs = CPU Core Freq. (3.8GHz) X Core Count (4) X 8 FLOPS + GPU Core Freq.(800MHz) X DirectX® 11 capable Shader Count (384) X 2 FLOPS."

"theoretical GFLOPS calculated by AMD as 856 for AMD A10-7850K with AMD RadeonTM R7 Series Graphics. GFLOPS = CPU GFLOPS + GPU GFLOPS = CPU Core Freq (3.7 Ghz) x Core count (4) x 8 + GPU Core Freq (720MHz) x Radeon Core (512) x 2"

You see? Different CPU and GPU cores, same formula. I don't think that Kaveri is only 16% better than Trinity. Looks like GFLOPS on slides are worthless.