AMD CPU speculation... and expert conjecture

[gamerk316]

Therefore, I have to explain to you something said about BF4 and the i5-4670k using a... crysis 3 benchmark where the i5-4670k doesn't appear?

Revolutionary concept! I will try:

Spoiler

a chart @ 720p, revolutionary indeed. a haswell i5 will out perform the i5 3550 in the chart, so you can insert that there.

this sums up what is wrong with AMD fanboys. i'd suggest a permanent ban on such.

and you're not particularly talking about BF4 alone either. there is only 1 game where the fx are "competitive" (not even winners) and you hail them as "winning this round".

8 craps cores are crap.

I like the double standard where 720p benchmarks are accepted when Intel wins, but are rejected when AMD wins.

But no problem with giving one at your solicited 1080p

Spoiler

and here BF4 with the FX-8350 above the Haswell i5-4670k

Spoiler

Spoiler

Crysis 3:

TechSpot:

Toms:

PcLab:

All put the FX-8350 at about the level of the i5-3550/i5-3570. Lower tier BD's (6300, etc) do even worse, competing with the i7-920 and older Phenom II CPUs.


BF4:

GameGPU:

PcLab:

Both have the 8350 hovering around 2600k level performance. And like the Crysis benchmarks, lower tier PD's compete against aged CPUs, namely Nahalem and Phenom II.

The only reason the 8350 does as well as it does is its 4GHz clock, not its extra cores. Lower its speed to 3.5GHz, and you'd see performance fall in line with the other PD based chips.

 

[de5_Roy]

AMD turns 45
http://www.fudzilla.com/home/item/34654-amd-turns-45-gets-a-porsche

 

[Cazalan]

First product of the openpower initiative

http://www.zdnet.com/google-eyes-power-chips-amazon-arm-both-add-up-to-intel-headaches-7000028881/

It is shown once again that AMD head did the correct when abandoned competing in the traditional Opteron server products (Warsaw CPU is only for legacy customers) and migrated to new HSA APU (Berlin) plus ARM (Seattle) strategy.

Interesting link but your conclusions make little sense.

A) It's just a prototype not a product. Meaning they may not use it at all. Google creates and kills projects just as quickly.

B) It's using very big POWER cores not little cores. Making it closer to Warsaw than ARM parts.

C) AMD didn't have much choice. They don't have the money to compete with expensive big Xeon parts. They're in debt up to their eyeballs.

 

[palladin9479]

Both have the 8350 hovering around 2600k level performance. And like the Crysis benchmarks, lower tier PD's compete against aged CPUs, namely Nahalem and Phenom II.

The only reason the 8350 does as well as it does is its 4GHz clock, not its extra cores. Lower its speed to 3.5GHz, and you'd see performance fall in line with the other PD based chips.

The F8350 ($190 USD) competes with the i5 in pricing. Same as the FX6300 ($120 USD) competes with the i3. The only thing that matters here is price. Your attempting to compare the $190 USD chip to the$250 ~ $300+ USD chips which is foolish. The closest you can compare between then is the i5-4440/4570 vs the FX8350 and the i3-4130/4339 vs the FX6300/6350. And that's without getting to the FX8320 which is cheaper then the i5-4430 yet cheaper then the i3-4340.

Now when we discuss the number of "cores" needed, gaming benchmarks are thrown out the window because every last one of them is unrealistic. The rule of thumb I've followed, and that I recommend others follow, is to have 1~2 cores more then the most used by your most demanding game. Since most games use 2~3 cores you want 3~5 cores on the system. This lets you play the game without worrying about any background process interfering with you. Those background process's include things like skype, vent, firefox, antivirus, HW monitors, and all the services Windows likes to run in the background. Essentially your doing what console design's do, allocating 1~2 cores for "system / other". Otherwise we'd all be using silly things like i3's with dual 780's and wondering whats wrong with the system.

 

[truegenius]

Both have the 8350 hovering around 2600k level performance. And like the Crysis benchmarks, lower tier PD's compete against aged CPUs, namely Nahalem and Phenom II.

The only reason the 8350 does as well as it does is its 4GHz clock, not its extra cores. Lower its speed to 3.5GHz, and you'd see performance fall in line with the other PD based chips.

this type of comments bring peace to my mind and then a voice came from inside which shouts, bring on the die shrink k10

Both have the 8350 hovering around 2600k level performance. And like the Crysis benchmarks, lower tier PD's compete against aged CPUs, namely Nahalem and Phenom II.

The only reason the 8350 does as well as it does is its 4GHz clock, not its extra cores. Lower its speed to 3.5GHz, and you'd see performance fall in line with the other PD based chips.

The F8350 ($190 USD) competes with the i5 in pricing. Same as the FX6300 ($120 USD) competes with the i3. The only thing that matters here is price. Your attempting to compare the $190 USD chip to the$250 ~ $300+ USD chips which is foolish. The closest you can compare between then is the i5-4440/4570 vs the FX8350 and the i3-4130/4339 vs the FX6300/6350. And that's without getting to the FX8320 which is cheaper then the i5-4430 yet cheaper then the i3-4340.

Now when we discuss the number of "cores" needed, gaming benchmarks are thrown out the window because every last one of them is unrealistic. The rule of thumb I've followed, and that I recommend others follow, is to have 1~2 cores more then the most used by your most demanding game. Since most games use 2~3 cores you want 3~5 cores on the system. This lets you play the game without worrying about any background process interfering with you. Those background process's include things like skype, vent, firefox, antivirus, HW monitors, and all the services Windows likes to run in the background. Essentially your doing what console design's do, allocating 1~2 cores for "system / other". Otherwise we'd all be using silly things like i3's with dual 780's and wondering whats wrong with the system.

we can put it in other way, if we compare pentium g series with fx top line fx9000
then in single/dual threaded tasks even pentium g stays ahead of fx
so here, 100$ chip is comparable to 700$ chip then it is wrong comparison
(btw you can buy other cheap locked i5- clock them upto their locked limit and then you can beat amd (even overclocked) at cheap price using less power, for example i5-4570 non k and clock it up to 3.8-4GHz base and 4-4.4 turbo and then it is better than fx8350 overclocked in sinlge and multi thread and costs same as fx8350 at newegg)

by this i mean both have advantage in their department, amd is selling cheap as their is no other choice

and not every thing in world is well threaded, supports HSA, Mantle
thus amd fails in anything which does not support these, if software supports these then you will roll else you will crawl
while intel and nvidia gives consistence performance throughout thus it won't make your pc looks like slow crap in some softwares, everyone loves performance , and everyone loves consistent performance even more

and thus other rule of thumb says that, cores needs to be fast octa are also present in meditek
(there is a humorous line i want to say, i don't know how to say this in English, so here it is in Hindi if someone can understand. cores fast honi chahiye octa core to meditek bhi hai)

AMD turns 45
http://www.fudzilla.com/home/item/34654-amd-turns-45-gets-a-porsche

well what is the purpose of such a pic in the article. Ground breaking !
they could have used something else more logical
or do they have already used it logically?

 

[jimmysmitty]

^ whats this ? mods are taking nap and you are using this opportunity

The main argument against BD has always been "Why does a newer processor with a 600MHz speed advantage and twice as many cores generally only do almost as good as a chip produced by Intel two years ago?". That's why BD has been so underwhelming. Part of that is because some people hyped the arch to death, promising an Intel killer. BD was never going to be that.

history is repeating itself, as now fanboys are promising apus as gaming dgpus killer

The issue is, when AMD can do that...you will probably be looking at GPUs with 8000+ GCN cores in the dGPU segment.

GPU performance is growing at roughly 30% gain per generation, while CPU performance is growing around 10% per generation. Eventually dGPU performance will be so staggeringly ahead of what you can fit on an APU (the gap will only increase) that no matter how you want to stack them the dGPU will be more performance per watt AND per dollar.

and here are some signs of more cores
AMD R9 390X Bermuda XTX 4224 cores
http://www.guru3d.com/news_story/amd_pirate_islands_radeon_r9_300_series.html

I remember the hype of Barcelona. Holy crap that was bad. It was a "native" quad core that would wipe the floor of the entire planet with Intels Conroe core CPUs. I am not sure how many of you guys were here for Barons little lectures on the superiority of native vs MCM that Intel did with its first quad cores.

As for the R9 300 series, we had insane numbers for Hawaii XT as well. Hell Fijis numbers resemble one of the rumors for Hawaii XT.

I am sure if it is a 20nm part, I doubt AMD will try to throw another space heater out on 28nm, it will have more cores. But is that the answer? Should they just keep throwing more cores or should they actually look into designing a more efficient arch that has less cores but performs better?

First product of the openpower initiative

http://www.zdnet.com/google-eyes-power-chips-amazon-arm-both-add-up-to-intel-headaches-7000028881/

It is shown once again that AMD head did the correct when abandoned competing in the traditional Opteron server products (Warsaw CPU is only for legacy customers) and migrated to new HSA APU (Berlin) plus ARM (Seattle) strategy.

Some time ago I predicted that new gen games would increase hardware requirements towards the top end. Wolfenstein recommends an i7 or FX 8-core

http://gamingbolt.com/wolfenstein-the-new-order-minimum-system-specs-revealed

AM1 vs Bay-Trail:

When looking at the benchmarks, it appears that the CPUs within the SoCs, each of which consists of four relatively slow cores (AMD Jaguar and Intel Silvermont), are closely matched with regards to performance. Sometimes Intel is slightly ahead, but most of the time AMD manages to be slightly faster. You can consider the CPU performance to be roughly half that of a Core i3 from the Haswell generation.

AMD clearly has the better graphical performance: the difference to Intel is larger in the 3DMark and game benchmarks. AMD appears to be better suited for multimedia work. In the short amount of time that we had, we were unable to run video tests, but considering the implemented technologies we are willing to predict that AMD will take the lead here as well. We will revisit this at a later point in time.

As for power consumption, both platforms are far more energy efficient than the Intel Socket 1150 or AMD Socket FM2(+) platforms. Intel is a little more energy efficient under load, but AMD wins when the processors are idle.

As far as we're concerned, AMD is also the victor with regards to connectivity: the AM1 platform has an extra USB 3.0 port, a Serial ATA 600 port instead of a Serial ATA 300 one, and support for DisplayPort 1.2.

The battle of AMD AM1 versus Intel Bay Trail-D ends, as far as we're concerned, in a resounding victory for AMD.

http://uk.hardware.info/reviews/5334/28/amd-am1-vs-intel-bay-trail-d-review-cheap-desktop-platforms-conclusion

Amazon is getting too big for its own good, much like Google. They are going to branch out too much and suffer in the end. Remember Ford? They did the same thing. They used to own many more car companies like Mercury (killed off), Aston Martin (still holds a small part but no longer full ownership), Volvo and many more. Turned out too much was holding them down and once they dropped back a bit to the more core companies, they started to profit again.

As for the news of Wolfenstein, it will not need a i7 or FX8000. Wanna know why? It is the id Tech 5 RAGE engine, the last engine Carmack coded before he left id to pursue whatever he wanted to do. I am sure they may have tweaked it a bit but I am 100% sure I will run it fairly well maxed out at 1080p with my i5 4670K. I ran RAGE maxed out on a Q6600 and a HD5870 1GB.

Either way at some point in time games will require a minimum quad core or 8 core but that is the nature of the beast in PC gaming. They move forward, slower than hardware but software eventually moves forward. When I did build my old Q6600 system with a HD 2900Pro 1GB, everyone else was using a Core 2 Duo E6000 or waited and got the E8400 and almost always a 8800GT 512MB or at best a 8800 Ultra with 768MB. Guess who didn't have to upgrade as soon?

Predicting that games will need more cores eventually is not that impressive.

AMD turns 45
http://www.fudzilla.com/home/item/34654-amd-turns-45-gets-a-porsche

I remember working at a community college on a work grant (same grant I used to pay for some college) and we had some AMD and Intel ROM chips back from when they did that and did not produce CPUs. You could easily flash the ROM by removing a sticker and exposing the window to daylight to wipe it. Was some cool stuff.

 

[de5_Roy]

^^ @truegenius: arm's big.Little configuration is good for mobile, but desktops and laptops don't really need them due to different usage model and due to the way power is supplied to pcs.
it could be good for tablets, like before i said 4 jag/puma cores doing heavy lifting and the 4 a.c.e.s in the igpu doing connected stand-by stuff and other lightweight tasks. intel was rumored to use the haswell+crystal well's igpu cores like that but for some reason that didn't pan out afaik. i think kaveri or mobile kaveri is capable in hardware, but might not get software support until carrizo's successor.

funny you should mention mediatek, they (and arm) could be looking to use hsa to bolster their own hmp (heterogenous multiprocessing) techs among other things.

i think you were trying to say "amd's octo cores should be fast since even mediatek has octo core cpus" or something similar.

 

[truegenius]

I remember the hype of Barcelona. Holy crap that was bad. It was a "native" quad core that would wipe the floor of the entire planet with Intels Conroe core CPUs. I am not sure how many of you guys were here for Barons little lectures on the superiority of native vs MCM that Intel did with its first quad cores.

is amd can use something like this in their apus ?
to produce separate cpu and gpu chips so that we can get pure cpu instead of harvested apu

i think you were trying to say "amd's octo cores should be fast since even mediatek has octo core cpus" or something similar.

exactly

due to my English and poor vocabulary (thats because of poor secondary memory) i wasn't able to create the sentence clearly
and at that time i was thinking, Y U no know english

 

[palladin9479]

we can put it in other way,

There is no other way. All that matters is price, period. It's the only way to fairly compare any consumer product, otherwise I could pull out 80~120K systems that would crush anything with a Core i7-4960X. They would be utterly useless for anything relevant to the discussion but they would sure make for great bar graphs.

Now which product you select depends on your particular application and situation, there is no "best" overall. This is what I keep pointing out yet the cheerleaders of both sides refuse to consider reason. If someone wants to be biased they can weight whatever metric they want to use in such a way that it always favors their team, which is why I ignore anyone who attempts to use such biased weighting. I only take situation specific results into account. First define the specific situation, then select the best realistic configuration for that situation without bias or prejudice. Anything else is just cheer leading and gets instantly dismissed.

And I wasn't joking about the needed core count. You will need 1~2 additional cores to run OS, applications and any background tasks on. That ensures the windows scheduler doesn't start preempting your game's heavy threads. This is the reason that Pentium G's are useless for realistic gaming, you can twink the system to make a nice bargraph, but actual day to day use would suck. i3's have a similiar problem, they don't have anymore power then a Pentium G but HT enables them to squeak out a bit more efficiency. The haswell i3 deserves special mention because the additional ALU enables it to get more out of HT then IB/SB, you can treat it as a 3 core CPU for all intents and purposes. For the AMD line, multiple the core count by .8 to get an approximate. This multiplier is only in effect when there is more load then available modules. All the AMD desktop CPU's are multiplier unlocked while only the upper mid range Intel CPU's do that. This actually creates two separate markets since the 8350 ($190 USD) is in a different bracket then the i5-4670K ($235 USD), they don't compete with each other.

All of the fighting is really about people wanting to brand the fx8350 as an upper end CPU when it reality it's priced solidly in the lower mid range. The fx6300 is low end territory and the 760K is the budget CPU. The fx9 series is a bad joke and I have no idea what idiot at AMD gave the green light for them to be marketed. When each CPU is judged in their respective price categories without bias or weighting, they actually do quite well. The FX8 is more of a niche as rarely do people in the lower mid range need eight cores, but those who do can get a really good bargain. The fx6 though is a solid chip for low end builds.

 

[juanrga]

The only reason the 8350 does as well as it does is its 4GHz clock, not its extra cores. Lower its speed to 3.5GHz, and you'd see performance fall in line with the other PD based chips.

Agree. The difference with the FX-6300 is due to the higher clock.

However, (i) the game uses more than four threads, (ii) and the FX-8350 is above the i5-4670k.

First product of the openpower initiative

http://www.zdnet.com/google-eyes-power-chips-amazon-arm-both-add-up-to-intel-headaches-7000028881/

It is shown once again that AMD head did the correct when abandoned competing in the traditional Opteron server products (Warsaw CPU is only for legacy customers) and migrated to new HSA APU (Berlin) plus ARM (Seattle) strategy.

Interesting link but your conclusions make little sense.

A) It's just a prototype not a product. Meaning they may not use it at all. Google creates and kills projects just as quickly.

B) It's using very big POWER cores not little cores. Making it closer to Warsaw than ARM parts.

C) AMD didn't have much choice. They don't have the money to compete with expensive big Xeon parts. They're in debt up to their eyeballs.

A) Google has not founded the OpenPower foundation for developing only prototypes. Google has real interest on using this technology.

B) I didn't mention anything about big or little.

C) If AMD were still managed by the old headless head, they would release some Steamroller FX/Opteron with 12 or more cores, and AMD would spiral into the red up to going out of business.

Luckily they are not in AMD anymore. Current AMD server head did make the correct steps: (i) To understand that Bulldozer was a failure and abandon the entire family for future servers, (ii) to understand the market is moving away of x86, (iii) understand that AMD actually cannot compete against IBM or Intel on big CPU cores and has to pick alternative markets such as HSA APUs and ARM Seattle.

 

[Cazalan]

I'm fine with AMD cancelling their MCM server line, I just wish they had continued with a HEDT.

 

[jimmysmitty]

I remember the hype of Barcelona. Holy crap that was bad. It was a "native" quad core that would wipe the floor of the entire planet with Intels Conroe core CPUs. I am not sure how many of you guys were here for Barons little lectures on the superiority of native vs MCM that Intel did with its first quad cores.

is amd can use something like this in their apus ?
to produce separate cpu and gpu chips so that we can get pure cpu instead of harvested apu

i think you were trying to say "amd's octo cores should be fast since even mediatek has octo core cpus" or something similar.

exactly

due to my English and poor vocabulary (thats because of poor secondary memory) i wasn't able to create the sentence clearly
and at that time i was thinking, Y U no know english

Yes they can. Intels 1156 CPUs with IGPs were a MCM style, CPU and GPU on the same chip but different/separate dies connected via the DMI interconnect. If Intel did it, AMD can too.

 

[juanrga]

Now when we discuss the number of "cores" needed, gaming benchmarks are thrown out the window because every last one of them is unrealistic. The rule of thumb I've followed, and that I recommend others follow, is to have 1~2 cores more then the most used by your most demanding game. Since most games use 2~3 cores you want 3~5 cores on the system. This lets you play the game without worrying about any background process interfering with you. Those background process's include things like skype, vent, firefox, antivirus, HW monitors, and all the services Windows likes to run in the background. Essentially your doing what console design's do, allocating 1~2 cores for "system / other". Otherwise we'd all be using silly things like i3's with dual 780's and wondering whats wrong with the system.

A very good recommendation, and the connection with what new consoles are doing is excellent.

we can put it in other way, if we compare pentium g series with fx top line fx9000
then in single/dual threaded tasks even pentium g stays ahead of fx
so here, 100$ chip is comparable to 700$ chip then it is wrong comparison

Anyone purchasing a 8-core CPU for using only single/dual threaded tasks has selected the wrong chip.

But what you are missing here is that people who purchase a 8-core CPU uses it for well-threaded tasks and/or for multitasking. This people detect a lag problem when use Intel CPUs with two or four cores.

Either way at some point in time games will require a minimum quad core or 8 core but that is the nature of the beast in PC gaming. They move forward, slower than hardware but software eventually moves forward. When I did build my old Q6600 system with a HD 2900Pro 1GB, everyone else was using a Core 2 Duo E6000 or waited and got the E8400 and almost always a 8800GT 512MB or at best a 8800 Ultra with 768MB. Guess who didn't have to upgrade as soon?

Predicting that games will need more cores eventually is not that impressive.

I agree on that it has no merit. However, people in this thread was predicting that new games couldn't scale above two threads and that i3 would be all what one needs to play them. I am giving benchmarks and news that show that we are migrating from a situation where FX-8350 couldn't math an i3 to a new situation where the FX can be on top of an i5-4670k.

 

[juanrga]

^^ @truegenius: arm's big.Little configuration is good for mobile, but desktops and laptops don't really need them due to different usage model and due to the way power is supplied to pcs.
it could be good for tablets, like before i said 4 jag/puma cores doing heavy lifting and the 4 a.c.e.s in the igpu doing connected stand-by stuff and other lightweight tasks. intel was rumored to use the haswell+crystal well's igpu cores like that but for some reason that didn't pan out afaik. i think kaveri or mobile kaveri is capable in hardware, but might not get software support until carrizo's successor.

It seems evident that you are referring only to the old IKS/cluster version of big.little, because modern big.little HMP is good for anything from mobile to supercomputers.

AMD current head already mentioned he expect to see big.little HMP in servers and Intel TCO mentioned that they want use big.little HMP for HPC. The rumor is that Intel will introduce big.little HMP in desktops/laptops for 2015/2016.

Big.little HMP requires single-ISA approach, which is absent in Haswell graphics, but will appear in Skylake.

 

[jimmysmitty]

^^ @truegenius: arm's big.Little configuration is good for mobile, but desktops and laptops don't really need them due to different usage model and due to the way power is supplied to pcs.
it could be good for tablets, like before i said 4 jag/puma cores doing heavy lifting and the 4 a.c.e.s in the igpu doing connected stand-by stuff and other lightweight tasks. intel was rumored to use the haswell+crystal well's igpu cores like that but for some reason that didn't pan out afaik. i think kaveri or mobile kaveri is capable in hardware, but might not get software support until carrizo's successor.

It seems evident that you are referring only to the old IKS/cluster version of big.little, because modern big.little HMP is good for anything from mobile to supercomputers.

AMD current head already mentioned he expect to see big.little HMP in servers and Intel TCO mentioned that they want use big.little HMP for HPC. The rumor is that Intel will introduce big.little HMP in desktops/laptops for 2015/2016.

Big.little HMP requires single-ISA approach, which is absent in Haswell graphics, but will appear in Skylake.

I haven't seen Intel looking into big.little but I have seen an article on them adding some extra offline cores that will kick in if a core goes bad, as they are fearing that as we get to the smaller process nodes the current will burn the CPU cores out and having a backup wouldn't be a bad idea.

 

[juanrga]

I haven't seen Intel looking into big.little but I have seen an article on them adding some extra offline cores that will kick in if a core goes bad, as they are fearing that as we get to the smaller process nodes the current will burn the CPU cores out and having a backup wouldn't be a bad idea.

Intel CTO announced that Intel want mix (big) icore and (little) Atom cores on same die for future versions of the Phi. And there are strong rumors that Skylake will have both big and little cores on same die, with big x86 cores being used for the CPU and the little x86 cores being used for graphics (Larrabe return!).

 

[juanrga]

Some new FX chips:

- AMD FX-7500 Radeon R7, 10 Compute Cores 4C+6G

- AMD FX-7600P Radeon R7, 12 Compute Cores 4C+8G

 

[Lessthannil]

Still waiting for a Steamroller Athlon or the cheapest quad core Steamroller APU I can get other than the A10-7700K and the 7850K.

 

[8350rocks]

Juan, the discussions I had with someone at AMD regarding bulk vs FD-SOI were what gave me the impression I had. Their engineers told me going to bulk would be "suicide", because the properties of the substrate would be "catastrophic" for the BD uarch design. It NEEDS high clocks, it was DESIGNED FROM ONSET to have high clocks to offset the single thread performance deficit in the name of parallel threading.

Now look, we have Kaveri, and the highest clocked part does not even hit the stock clock of the flagship part introduced 2 YEARS AGO. Was my prediction that Kaveri on bulk would compromise clock speed so greatly that any gains in performance would be negated by loss of clockspeed wrong? No. Thank you for playing *bows*

Yes, the 'engineer' supposedly said you wrong stuff, which you parroted here, but was replied then. I recall explaining to you during many months that Kaveri was bulk no SOI and you pretending otherwise again and again.

I also recall your FUD about clocks and how the world come to an end. I replied you that Kaveri could hit 4.5Ghz on air. Then the first leaks of an engineering sample appeared and you posted FUD about those. Your crazy argument was that AMD benchmarking an ES at only 1.8GHz was implying that Kaveri couldn't be clocked much higher than that. I still laugh today.

I already explained to you why clocks are lower. Kaveri CPU is clocked at 3.7GHz because (i) it has lower TDP than Richland/Trinity and (ii) it has a bigger and powerful iGPU sharing that lower TDP.

If you increase the TDP to 100W you can hit the 4GHz without any problem. In fact despite having lower clocks at stock, Kaveri has higher OC factor.

Richland has an OC factor of 19.7% on air.
Kaveri has an OC factor of 20.9% on air.

3.) If that is the case, then why did AMD just release a new workstation GPU with 12 GB VRAM? Because their APU is going to scale so well it would outrun a R9-290X on steroids right? No. Wrong.

Are you really asking me to explain you why AMD releases an dGPU today, when they didn't select any for their future exascale supercomputer? Interesting.

Then you would be also very confused on why Nvidia Research Team claims that dGPUs will be replaced by APUs about 2018 but the same Team is releasing a Titan Z dGPU those days.

I suppose that both AMD and Nvidia engineers know how to use a calendar.

Juan, I said it was "comparatively low" for an early ES.

Amazingly you manage to twist words time and again. As I recall, at that time you said it was not an issue and clockspeed should be 4.0 GHz as your "predictions" stated.

Now, as I recall, when Kaveri was revealed, someone had to "fix" his predictions to accomodate for the loss in clockspeed...correct?

Now maybe you should go fix your crystal ball.

 

[jdwii]

I find it funny he says higher OC factor although its only by 1%? While its clocked way lower than the other chip.

 

[Cazalan]

Intel CTO announced that Intel want mix (big) icore and (little) Atom cores on same die for future versions of the Phi. And there are strong rumors that Skylake will have both big and little cores on same die, with big x86 cores being used for the CPU and the little x86 cores being used for graphics (Larrabe return!).

"strong rumors"? LoL. Just because you perpetuate it despite being shut down on the S|A forums repeatedly doesn't make it so. Even ARM doesn't make those kind of absurd statements with their little cores, because they have a dedicated Mali graphic core.

Intel has already spoken about the plans for their graphic cores. It has nothing to do with little cores.
http://blogs.intel.com/intellabs/2014/02/09/what-would-you-do-with-extra-40/

 

[Ags1]

When are we going to finally see some steamroller laptops?

 

[juanrga]

Juan, I said it was "comparatively low" for an early ES.

Amazingly you manage to twist words time and again. As I recall, at that time you said it was not an issue and clockspeed should be 4.0 GHz as your "predictions" stated.

Now, as I recall, when Kaveri was revealed, someone had to "fix" his predictions to accomodate for the loss in clockspeed...correct?

Now maybe you should go fix your crystal ball.

You said that the ES at 1.8GHz did mean that final Kaveri couldn't pass the 3GHz. I said you that Kaveri could hit 4.5Ghz and guess what the average frequency on air is of 4775MHz

4775MHz = 4.775GHz = 4.5GHz.

just as predicted by my broken crystal ball.

As explained before Kaveri CPU is clocked lower because has less TDP and more GPU on die. Your FUD about clocks was so accurate as your "facts" about FX-Steamroller coming to AM3+, Kaveri FD-SOI 28nm, AM4 on store, extra ALU per core, 30% single thread performance increase, Steamroller being faster than Haswell...

I find it funny he says higher OC factor although its only by 1%? While its clocked way lower than the other chip.

What is really funny is the double standard where lower base frequency is attributed to bulk, but higher OC factor is then dismissed.

You would not look at absolute differences, because the difference between 98% and 97%) and between 3% and 2% is of 1%, but in the last case you are adding a 50% more.

If percentages and relative values generate some problem try with frequencies:

The top Kaveri APU hits 4775MHz on air.
The FX-8350 CPU hits 4750MHz on air.

Yes I can see how Kaveri cannot hit 3Ghz. It is a complete disaster.

 

[8350rocks]

Juan, I said it was "comparatively low" for an early ES.

Amazingly you manage to twist words time and again. As I recall, at that time you said it was not an issue and clockspeed should be 4.0 GHz as your "predictions" stated.

Now, as I recall, when Kaveri was revealed, someone had to "fix" his predictions to accomodate for the loss in clockspeed...correct?

Now maybe you should go fix your crystal ball.

You said that the ES at 1.8GHz did mean that final Kaveri couldn't pass the 3GHz. I said you that Kaveri could hit 4.5Ghz and guess what the average frequency on air is of 4775MHz

4775MHz = 4.775GHz = 4.5GHz.

just as predicted by my broken crystal ball.

As explained before Kaveri CPU is clocked lower because has less TDP and more GPU on die. Your FUD about clocks was so accurate as your "facts" about FX-Steamroller coming to AM3+, Kaveri FD-SOI 28nm, AM4 on store, extra ALU per core, 30% single thread performance increase, Steamroller being faster than Haswell...

I find it funny he says higher OC factor although its only by 1%? While its clocked way lower than the other chip.

What is really funny is the double standard where lower base frequency is attributed to bulk, but higher OC factor is then dismissed.

You would not look at absolute differences, because the difference between 98% and 97%) and between 3% and 2% is of 1%, but in the last case you are adding a 50% more.

If percentages and relative values generate some problem try with frequencies:

The top Kaveri APU hits 4775MHz on air.
The FX-8350 CPU hits 4750MHz on air.

Yes I can see how Kaveri cannot hit 3Ghz. It is a complete disaster.

Juan...You are comparing the maxed out A10-7850k to what the 3rd highest clocked part?

Wrong, FLAGSHIP parts are the FX 9590 @ 4.7 GHz base clock/5.0 GHz Turbo Core 3.0.

Now, your part overclocked does not hit the stock clocks of the old flagship part.

Show me a production Kaveri @ 4.5 GHz and we can have your discussion...otherwise...your crystal ball is cracked.

 

[juanrga]

Intel CTO announced that Intel want mix (big) icore and (little) Atom cores on same die for future versions of the Phi. And there are strong rumors that Skylake will have both big and little cores on same die, with big x86 cores being used for the CPU and the little x86 cores being used for graphics (Larrabe return!).

"strong rumors"? LoL. Just because you perpetuate it despite being shut down on the S|A forums repeatedly doesn't make it so. Even ARM doesn't make those kind of absurd statements with their little cores, because they have a dedicated Mali graphic core.

Intel has already spoken about the plans for their graphic cores. It has nothing to do with little cores.
http://blogs.intel.com/intellabs/2014/02/09/what-would-you-do-with-extra-40/

Your link is completely irrelevant. And Skylake CGPU rumors are covered in many places:

http://semiaccurate.com/2012/06/12/what-comes-after-knights-landing/

http://news.softpedia.com/news/Intel-s-2015-and-2016-CPUs-Are-Skylake-and-Skymont-237504.shtml

http://www.brightsideofnews.com/2011/11/28/the-future-of-intel-cpu-architectures-revealed-haswell2c-skylake/

http://www.techradar.com/news/computing-components/graphics-cards/intel-could-kill-performance-pc-graphics-in-2015-1089339

http://www.chw.net/2012/06/futurologia-cpus-intel-sky-lake-tendran-un-igp-basado-en-larrabee/

http://www.scientificcomputing.com/articles/2014/01/mobile-tech-between-rock-and-hard-place