AMD CPU speculation... and expert conjecture

[lilcinw]

If you read the german article that was posted it explains the difference between dual rank and dual sided. Some one sided sticks are still dual rank and vice versa.

 

[-Fran-]

Since gTranslation didn't cut it, I had to use my Google Powers and voilá:

http://en.wikipedia.org/wiki/Memory_rank

Cheers!

 

[etayorius]

@juan

Corrected by someone who never even verified anything isnt correcting, just making up things from thin air isnt correcting.

my tests with cosmology can be 100% verified. You saying "your wrong" cant be verified.

Passmark ... real world ... omg in the same sentence ... rofl. Passmark is 100% synthetic measuring only fpu and alu preformance. It doesnt even test the memory subsystem in a "cpu" benchmark.

Not to mention the submitted "baseline" results arent even at stock clocks.

http://www.passmark.com/baselines/V8/display.php?id=18879541076

Lets at least look at stock speeds.

http://www.passmark.com/baselines/V8/display.php?id=18936243269

wow. Almost 5000 pts, 1000 pts below 6000.

So, basically the 7850k OCed to 4.5GHz = i5 2500k... well not bad, but that would be 800MHz (20%+) faster than Stock, but i am not sure if every single Kaveri can hit that much OC.

 

[wh3resmycar]

Nothing wrong with a a10 for gaming, if you understand it's purpose.low to mid range fps in all games on a one chip solution.

Dropping down to a low end Pentium and discrete is quickly being nullified by rising gpu prices and game developers taking advantage of more threads. Kaveri's biggest competition does not come from intel but from itself in the form of athlon and fx + descrete.

i have yet to meet a sane person, who buys his games legit, and looks forward to play games from low to medium fps/settings. (eg. "oh bf4? yes please, let me pay $60 to play that game at mediocre visual settings). given the circumstance, a $399 ps4 will make more sense.

there's a reason we call desktop solutions, solutions. "dropping down to a low end pentium" opens up the possibilities for future upgrades, because game developers "are still trying to take" advantage of more threads (how many years has it been since dx11/opencl pundits talked about multicore this and that? how many games in existence did actually scale perfectly across all threads?). as long as the radeon hdx750/260x series is around, these APUs will never have any place in the desktop.

mantle could prove us wrong though, hopefully. 8 more days.

 

[etayorius]

Nothing wrong with a a10 for gaming, if you understand it's purpose.low to mid range fps in all games on a one chip solution.

Dropping down to a low end Pentium and discrete is quickly being nullified by rising gpu prices and game developers taking advantage of more threads. Kaveri's biggest competition does not come from intel but from itself in the form of athlon and fx + descrete.

i have yet to meet a sane person, who buys his games legit, and looks forward to play games from low to medium fps/settings. (eg. "oh bf4? yes please, let me pay $60 to play that game at mediocre visual settings). given the circumstance, a $399 ps4 will make more sense.

there's a reason we call desktop solutions, solutions. "dropping down to a low end pentium" opens up the possibilities for future upgrades, because game developers "are still trying to take" advantage of more threads (how many years has it been since dx11/opencl pundits talked about multicore this and that? how many games in existence did actually scale perfectly across all threads?). as long as the radeon hdx750/260x series is around, these APUs will never have any place in the desktop.

mantle could prove us wrong though, hopefully. 8 more days.

I actually play at 1600x1200 (max monitor Res), but in high or ultra settings... i`ll trade little resolution for higher detail on games any time.

I can play any game if i ramp up to High or Ultra detail with the GTX470 and a 980BE, as long as i don`t go above 1600x1200 Resolution, my GPU is barely a mid range GPU now a days, it is probably better than some Xbone Games that are set to 720p.

CPU OCed from 3.7GHz to 3.91GHz and GPU OCed from 607MHz to 720MHz.

Oh, and i own 96 Games on Steam legally http://steamcommunity.com/id/etayorius/

 

[8350rocks]

personally i like kaveri in the uk it isnt half bad granted i could get cheaper i3 and a 7750 and have best of both worlds but these tests are done without amds mantle drivers which are delayed ( amds habit unfortunately)

people forget that kaveri isnt really aimed at gamers and they dont really control the prices either newegg etc can sell it for whatever personally to make it competive i uk its just got to drop its 95w power consumption down to 65w which a10 7800 if it comes out will. which is usually same performance just not unlocked.

plus the prices will go down eventually once they retailers notice not many are buying them.

To me though im a eco freak so i like to limit myself delibrately

if the a10 can merge with a 7750 or low end r7 card thats fine for me.

memory costs of 2133 and 2400 arent that much diffrent in uk though memory has gone up in uk quite a bit

e.g bought my vengence for round 45 pounds for 8gb of 1600mhz
now the price is 65

point is hsa may take off and if mantle driver works better it will benefit apu kaveri as well.

personally steamroller has issues due to the cpu cores are crammed in there with 50 percent of the die practically for the gpu ( which is a great feat) unfortunately its the manufacturing process that reduced the clock speed so i expected a performance hit. and i dont believe it ocs well you really cant compared steamroller kaveri to fx chip as its not really designed to be no more then a media part.

though i say its above a i3 and is between i3 and i5

if retailers dropped the horrific tag in usa and uk dropped it by 10-20 it look more attractive

I agree that CPU is between i3 and i5. In Spain 2133MHz memory is already cheaper than 1600MHz. And the 2400MHz is only a bit more expensive.

Kaveri will not break OC records, but can hit 4.6GHz on air (average), showing better OC than Trinity.

It also shows remarkably less OC potential than Richland...which often hit close to 5 GHz on air cooling.

That was a mature process...sure...but it was also SOI, which is notoriously better for heat and overclocking. Honestly, given the density of transistors on die in the Kaveri APUs, I would be shocked if they actually turn out able to hit even 4.6 GHz average OC on air cooling. Look how bad Intel OCs these days and Kaveri is far more dense at 28nm than Haswell is at 22nm. Impressive feats from AMD engineering, but bad for OCing enthusiasts.

 

[palladin9479]

i have yet to meet a sane person, who buys his games legit, and looks forward to play games from low to medium fps/settings. (eg. "oh bf4? yes please, let me pay $60 to play that game at mediocre visual settings). given the circumstance, a $399 ps4 will make more sense.

Let me know if you find any Scotsmen while your at it. Your entire statement is loaded with self qualifiers.

I happen to use an APU (6800K + 2133 memory) inside my living room for light gaming, I've even posted pictures of that same box on here. I also have a 900D loaded up with components that cost several thousand USD inside my computer room / office.

Two different solutions for two different problems. There is no way in hell I'd ever run something that resembles my rig inside my living room, entirely different atmosphere.

 

[jdwii]

@juan

Rofl, still stuck on that mystery ES on cosmology? If anyone had a kaveri cpu, could they run the bionic benchmark and show this tool that the ES was not kaveri. @1.8 ghz it scored 3300 pts roughly so at 3.6 ghz if that ES is actual kaveri should only score ~ 6600 pts.

Your the one saying its kaveri not me. LEARN TO READ. only thing I pointed out is that the blog post claining "ermago 37% more ipc" was noting but 100% fake.

Aside from that, AMD is abandoning x86 servers right?

http://www.xbitlabs.com/news/cpu/display/20140117215942_AMD_Develops_Behemoth_Chip_with_Sixteen_Cores_on_Single_Die.html

oops, got that one wrong too didn't ya. ERMAGO its not on the roadmap, must be fake.

The only mystery is why you repeat the same flawed argument which was corrected a dozen of times before.

Also that recent 16-core rumor was debunked here at least two times.

Passmark is not even a accurate benchmark

I don't know what you mean by accuracy (this is a term usually misinterpreted and I don't know if it is your case or if you are using it properly).* It represent real-world performance much better than SuperPi, CB, Sysmark...

Passmark scores have been discussed in the past in this thread and by more than one poster. I note that you only question its accuracy now, when the Passmark score measured for Kaveri agrees (within a 2% error) with the value predicted. My bet is that if the error was of 63% many of my 'fans' here would be posting about how wrong I was and nobody would question accuracy.

* And not only the general public! In the below link you can find someone using the term accuracy incorrectly to refer to benchmarks. The correct term for what he is reporting is "imprecise".

http://www.engadget.com/2011/07/15/engadget-primed-using-benchmarks/#accurate

I only do it now since it's relevant, i never thought this piece of software was any good in the real world
http://www.cpubenchmark.net/cpu.php?cpu=AMD+FX-8350+Eight-Core

That for example proves what i mean.

 

[jdwii]

"63% many of my 'fans' here would be posting about how wrong I was and nobody would question accuracy."
lol so i guess in real world programs a A10 does equal a I5 funny in all the programs i would use it wouldn't even come 25% close. I guess like the BD launch you will have those people looking for the rarest and crappiest benchmarks.

 

[griptwister]

Nothing wrong with a a10 for gaming, if you understand it's purpose.low to mid range fps in all games on a one chip solution.

Dropping down to a low end Pentium and discrete is quickly being nullified by rising gpu prices and game developers taking advantage of more threads. Kaveri's biggest competition does not come from intel but from itself in the form of athlon and fx + descrete.

i have yet to meet a sane person, who buys his games legit, and looks forward to play games from low to medium fps/settings. (eg. "oh bf4? yes please, let me pay $60 to play that game at mediocre visual settings). given the circumstance, a $399 ps4 will make more sense.

there's a reason we call desktop solutions, solutions. "dropping down to a low end pentium" opens up the possibilities for future upgrades, because game developers "are still trying to take" advantage of more threads (how many years has it been since dx11/opencl pundits talked about multicore this and that? how many games in existence did actually scale perfectly across all threads?). as long as the radeon hdx750/260x series is around, these APUs will never have any place in the desktop.

mantle could prove us wrong though, hopefully. 8 more days.

Hey! Don't bash console players! They can't help that most of their machinery is out dated quickly and they can't upgrade except by going out and buying a new system.

 

[con635]

@juan

Corrected by someone who never even verified anything isnt correcting, just making up things from thin air isnt correcting.

my tests with cosmology can be 100% verified. You saying "your wrong" cant be verified.

Passmark ... real world ... omg in the same sentence ... rofl. Passmark is 100% synthetic measuring only fpu and alu preformance. It doesnt even test the memory subsystem in a "cpu" benchmark.

Not to mention the submitted "baseline" results arent even at stock clocks.

http://www.passmark.com/baselines/V8/display.php?id=18879541076

Lets at least look at stock speeds.

http://www.passmark.com/baselines/V8/display.php?id=18936243269

wow. Almost 5000 pts, 1000 pts below 6000.

@juan

Corrected by someone who never even verified anything isnt correcting, just making up things from thin air isnt correcting.

my tests with cosmology can be 100% verified. You saying "your wrong" cant be verified.

Passmark ... real world ... omg in the same sentence ... rofl. Passmark is 100% synthetic measuring only fpu and alu preformance. It doesnt even test the memory subsystem in a "cpu" benchmark.

Not to mention the submitted "baseline" results arent even at stock clocks.

http://www.passmark.com/baselines/V8/display.php?id=18879541076

Lets at least look at stock speeds.

http://www.passmark.com/baselines/V8/display.php?id=18936243269

wow. Almost 5000 pts, 1000 pts below 6000.

So, basically the 7850k OCed to 4.5GHz = i5 2500k... well not bad, but that would be 800MHz (20%+) faster than Stock, but i am not sure if every single Kaveri can hit that much OC.

First test using 6gb secondtest using 7gb ram. NO dual channel benches

They're both dual channel, 2x4gb, first uses 2gb for gpu, 2nd uses 1gb for gpu, its adjustable in bios.

 

[de5_Roy]

a bit more on the socketed kabini apus(socs)
http://techreport.com/news/25948/specs-leak-out-for-desktop-kabini-processors
this is begining to look more and more interesting to me. before, all we got were bga cpus/apus/socs with fixed cooler and hsf. with amd paving the way to socketed socs in entry level, this'll make for a cheap low end diy pc. if amd makes it upgradable, we might even get beema/mullins apus on it.... and who knows... may be one of those new arm socs on dt. the boards need proper cooler mounts though. the ones i saw had two holes like the ones for mobo chipsets and cheap gpu hs.

edit: it needs:
4 mounting holes for soc cooling,
so-dimm instead of standard dimm sockets to make room for more (angled, if possible.. near the top edge)sata and/or usb 3.0 ports,
3 cooling fan headers, preferably 4 pin, in case users want to use passive cooling on the soc,
and lastly, overclocking options - in hardware as well as in bios.

 

[colinp]

Also that recent 16-core rumor was debunked here at least two times.

Oops:

http://www.zdnet.com/amd-adds-new-12-and-16-core-parts-to-the-opteron-6300-lineup-7000025448/

 

[jed]

Also that recent 16-core rumor was debunked here at least two times.

Oops:

http://www.zdnet.com/amd-adds-new-12-and-16-core-parts-to-the-opteron-6300-lineup-7000025448/

(juanrga)
You can't debunk this the AMD Piledriver 6300 series 12 to 16 core cpu's are for real. lol


http://www.eweek.com/servers/amd-launches-16-12-core-opteron-server-chips.html
http://www.pcmag.com/article2/0,2817,2429795,00.asp
http://www.tweaktown.com/pressrelease/8603/amd-debuts-new-12-and-16-core-opteron-6300-series-processors/index.html
http://www.maximumpc.com/amd_adds_more_12-core_and_16-core_processors_opteron_6300_series

 

[jacobian]

This is one of the firsts: "It’s time for AMD to take a page from Intel and dump Steamroller"

http://www.extremetech.com/computing/174980-its-time-for-amd-to-take-a-page-from-intel-and-dump-steamroller

 

[gamerk316]

Is Nintendo working on a new console already?
http://www.fudzilla.com/home/item/33722-is-nintendo-working-on-a-new-console-already?

If so, Nintendo is now SEGA. If true, the CPU is still a cripple:

CPU: IBM 64-Bit Custom POWER 8-Based IBM 8-Core Processor CODENAME JUMPMAN (2.2 GHz, Shared 6 MB L4 cache)
- Co-CPU: IBM PowerPC 750-based 1.24 GHz Tri-Core Co-Processor CODENAME HAMMER

The main CPU, based on clock, looks to be slightly weaker then the 360/PS3 main CPU [Moving from Power7 to Power8 isn't overcoming a 1GHz clock speed disadvantage]. Second is clearly for backward comparability, since the GC, Wii, and Wii U CPU's are all PPC 750 descendants [and why I CORRECTLY called the Wii U CPU weaker then the PS3/360].

 

[de5_Roy]

This is one of the firsts: "It’s time for AMD to take a page from Intel and dump Steamroller"
http://www.extremetech.com/computing/174980-its-time-for-amd-to-take-a-page-from-intel-and-dump-steamroller

the hell? dumping sr will worsen amd's situation with high margin products. while kaveri itself is a moderately low margin, it is still part of amd's own high-mid lineup. amd needs a core for the alternative high end pc owners. if anything, amd should dump excavator for a standalone cpu and use 20nm for sr cpu and excavator-running carrizo. last thing amd needs is a rumor of dumping cpu cores amidst their cpu lineups taking hit for 2 quarters. another way for amd is to aggressively get into ultramobile and pursue tablet design wins. not like intel's money dumping, but with performance, especially igpu performance.

 

[pangolin_user]

AMD kaveri A10 7850 K with configurable TDP, I think Its very interesting feature

http://www.phoronix.com/scan.php?page=article&item=amd_kaveri_tdp&num=1

 

[pangolin_user]

Is Nintendo working on a new console already?
http://www.fudzilla.com/home/item/33722-is-nintendo-working-on-a-new-console-already?

If so, Nintendo is now SEGA. If true, the CPU is still a cripple:

CPU: IBM 64-Bit Custom POWER 8-Based IBM 8-Core Processor CODENAME JUMPMAN (2.2 GHz, Shared 6 MB L4 cache)
- Co-CPU: IBM PowerPC 750-based 1.24 GHz Tri-Core Co-Processor CODENAME HAMMER

The main CPU, based on clock, looks to be slightly weaker then the 360/PS3 main CPU [Moving from Power7 to Power8 isn't overcoming a 1GHz clock speed disadvantage]. Second is clearly for backward comparability, since the GC, Wii, and Wii U CPU's are all PPC 750 descendants [and why I CORRECTLY called the Wii U CPU weaker then the PS3/360].

I think you got it wrong, 8 core IBM Power 8 at 2.2Ghz is way more powerfull than xbox 360 3 core SMT and PS3 1 core 7 SPU Cell , xenon prosesor is between power 5+ to power 6
Power 5 launched in 2004
Power 6 Launched in 2007
Power 7 in 2010
Power 8 in 2013
and xenon in 2005

 

[Rum]

Is Nintendo working on a new console already?
http://www.fudzilla.com/home/item/33722-is-nintendo-working-on-a-new-console-already?

If so, Nintendo is now SEGA. If true, the CPU is still a cripple:

CPU: IBM 64-Bit Custom POWER 8-Based IBM 8-Core Processor CODENAME JUMPMAN (2.2 GHz, Shared 6 MB L4 cache)
- Co-CPU: IBM PowerPC 750-based 1.24 GHz Tri-Core Co-Processor CODENAME HAMMER

The main CPU, based on clock, looks to be slightly weaker then the 360/PS3 main CPU [Moving from Power7 to Power8 isn't overcoming a 1GHz clock speed disadvantage]. Second is clearly for backward comparability, since the GC, Wii, and Wii U CPU's are all PPC 750 descendants [and why I CORRECTLY called the Wii U CPU weaker then the PS3/360].

Nintendo would have to screw up a lot more in order to become the next Sega, considering they have 18 billion in the bank I don't foresee them going anywhere. Now if they drop the WiiU and start going all crazy like Sega did with their systems at the end of their console making days then I could see that, but I highly doubt they would be that dumb! They will eat the cost and hopefully learn from this fiasco, from the looks of it those rumored "new system specs" are probably fake.

While backwards compatibility is good they should have just gone with the AMD APU and put their own spin on it like the other 2.. this would have made getting multi plats far easier, or they could have used a more powerful IBM CPU in the thing! Only saving grace in the system is the dedicated GPU brings it at least "close" to the current gen consoles and when Nintendo starts releasing their big name titles they will start selling more consoles!

 

[de5_Roy]

Kaveri Versus Richland: A Performance Per Clock Comparison
http://semiaccurate.com/2014/01/23/kaveri-versus-richland-performance-per-clock-comparison/

edit:
New MSI FM2+ GAMING motherboards Available Now
http://www.techpowerup.com/197153/new-msi-fm2-gaming-motherboards-available-now.html

SanDisk's ULLtraDIMM is an SSD on a memory stick
http://techreport.com/news/25951/sandisk-ulltradimm-is-an-ssd-on-a-memory-stick

edit2:

AMD’s total wafer purchases from GlobalFoundries in 2013 were approximately $960 million, lower than the previously estimated $1.15 billion, due to lower fourth quarter purchases. There were no penalties associated with the reduction. AMD is working on its 2014 wafer supply agreement with GlobalFoundries, based on 2014 full year demand expectations, with the goal to manage inventory flat to down year-over-year. In Q1 2014, AMD made the final $200 million payment related to the reduction of the take or pay wafer obligation commitments in 2012. In addition, the company expects wafer purchases of approximately $250 million in Q1 2014.

http://www.xbitlabs.com/news/cpu/display/20140121234035_Thank_You_Consoles_AMD_s_Posts_Good_Results_Amid_Drop_of_Microprocessor_Sales.html

 

[gamerk316]

Is Nintendo working on a new console already?
http://www.fudzilla.com/home/item/33722-is-nintendo-working-on-a-new-console-already?

If so, Nintendo is now SEGA. If true, the CPU is still a cripple:

CPU: IBM 64-Bit Custom POWER 8-Based IBM 8-Core Processor CODENAME JUMPMAN (2.2 GHz, Shared 6 MB L4 cache)
- Co-CPU: IBM PowerPC 750-based 1.24 GHz Tri-Core Co-Processor CODENAME HAMMER

The main CPU, based on clock, looks to be slightly weaker then the 360/PS3 main CPU [Moving from Power7 to Power8 isn't overcoming a 1GHz clock speed disadvantage]. Second is clearly for backward comparability, since the GC, Wii, and Wii U CPU's are all PPC 750 descendants [and why I CORRECTLY called the Wii U CPU weaker then the PS3/360].

I think you got it wrong, 8 core IBM Power 8 at 2.2Ghz is way more powerfull than xbox 360 3 core SMT and PS3 1 core 7 SPU Cell , xenon prosesor is between power 5+ to power 6
Power 5 launched in 2004
Power 6 Launched in 2007
Power 7 in 2010
Power 8 in 2013
and xenon in 2005

Mainly because I tried to compare three things at once; technically, the Cell PPE [which the Xenon CPU is based off of] uses the Power 2.02 ISA. The Wii U's CPU is just a heavily modified PPC 750 [released back in 2001, FYI].

My point remains unchanged. IBM hasn't done enough with the arch where a POWER CPU is going to be competitive at that low a speed.

Nevermind that Nintendo would burn every bridge it had left with the software Devs if they did this; see SEGA as a prime example.

 

[de5_Roy]

nintendo should either go with 8-16 core high-clocked arm a57 cpu (hopefully boosted by st-micro's tech) with a gcn igpu. or a puma-powered cpu with gcn igpu. both made by amd.
<chant>fullfil the trifecta/hat-trick, nintendo
buy from amd's semi-custom
resistance is futile
in the console space without semicustom amd soc, no one can hear you scream for profit</chant>

 

[etayorius]

nintendo should either go with 8-16 core high-clocked arm a57 cpu (hopefully boosted by st-micro's tech) with a gcn igpu. or a puma-powered cpu with gcn igpu. both made by amd.
<chant>fullfil the trifecta/hat-trick, nintendo
buy from amd's semi-custom
resistance is futile
in the console space without semicustom amd soc, no one can hear you scream for profit</chant>

I rather Nintendo go for an x86 APU, just like PS4 and Xbone so it`s easier for devs to also develop the same games easily, if not, i would like to see Nintendo games on other consoles.

 

[juanrga]

@juan

Corrected by someone who never even verified anything isnt correcting, just making up things from thin air isnt correcting.

my tests with cosmology can be 100% verified. You saying "your wrong" cant be verified.

Passmark ... real world ... omg in the same sentence ... rofl. Passmark is 100% synthetic measuring only fpu and alu preformance. It doesnt even test the memory subsystem in a "cpu" benchmark.

Not to mention the submitted "baseline" results arent even at stock clocks.

http://www.passmark.com/baselines/V8/display.php?id=18879541076

Lets at least look at stock speeds.

http://www.passmark.com/baselines/V8/display.php?id=18936243269

wow. Almost 5000 pts, 1000 pts below 6000.

My explanation made explicit why your "my tests" excuse was invalid.

Passmark scores reflect very accurately real-world performance, because the ordering via Passmark scores agrees very well with the ordering with real-world performance that some of us obtain. It is irrelevant if Passmark achieve that using 100% synthetic tests or using 'magic', as a consequence your argument is invalid.

Moreover, we are comparing average score of Kaveri to average score of Richland. You pretend to compare average score of Richland to your cherry picked Kaveri score.

In fact, you cherry picked the poor score in the list whereas ignored this (above 5800 points at stock speed)

http://www.passmark.com/baselines/V8/display.php?id=18835440199

Check also this new baseline

http://www.passmark.com/baselines/V8/display.php?id=19024964763

Another >5800 points at stock speed.