haswell or piledriver

[vishalaestro]

i need a recommendation for the cpu since today's games are using more no of cores is it wise to stay with amd or with intel

 

[whyso]

The synthetic test is good to use because it shows the maximum practical bandwidth (assuming sandra works which it does). In actual tasks the bandwidth is always less than or greater to. So the i3 may have less bandwidth in practice but the a10 will never have more. Give me a graph that shows trinity memory scaling. The pentium I linked gets 17 GB/s using 1333 mhz ram.

Sony can say whatever they want but its vapourware until we see it. yes ps4 will compete with probably gtx 570 level computers (cause of the gpu). And yes they would have devs code on a 8 core FX platform because their chip isn't finished yet (probably is now but when they gave the dev kits out the chips were still in progress, the same thing happened with the xbox 360 gpu and wii u gpu).

Jaguar is 15% IPC increase over bobcat. Bobcat is terrifically slow. Jaguar is completely different from FX series (the core size is much smaller).

Its going to be about five times that speed. (E-350 is two bobcat cores + IPC boost). A 3570k is about four times faster (11.3 sec) on the same test, making it about 70% faster.

http://www.anandtech.com/bench/Product/328?vs=701

Multiply the e-350 by five. The thing is pretty slow no matter how you cut it.

Yeah, no one is going to be developing for 8 cores anytime soon for the majority of games. How many games are still dual core? (And the xbox 360 cpu is tri-core).

Things in my country arn't representative of the rest of the world but neither are they in yours.

Im not sure where you are getting that ram speed but here http://www.tomshardware.com/reviews/a10-5800k-a8-5600k-trinity-apu,3241-2.html

G.Skill 16 GB (4 x 4 GB) DDR3-1600, F3-12800CL9Q2-32GBZL @ 9-9-9-24 and 1.5 V

There are no motherboard problems in this review. I'd say margin of error is around +/- 5%. And no trinity scales almost linearly in games for two reasons. It needs more bandwidth (much stronger than hd4000) and it gets less bandwidth not because amd has a great memory controller.

1) The synthetic test attempts to measure theoretical performance not real-world performance. In any case trinity has a bandwidth similar to i3 (sandy) and FX-8350 has a bandwidth similar to i7 (ivy) in that synthetic test as I showed to you before.

2) What a Sony representative, with name and surname, says at Game Developers Conference 2013 is more realistic than what an anonymous poster (who have no access to the real thing) writes at a forum.

3) The Photoshop figure that you link is from a preview done on a ES platform that used a lower clocked DDR3 memory that worked at 1066Mhz.

4) The AMD chip used in the PS4 is an innovative unified design. You cannot compare it with an older design as the i5-3570k in the way that you are doing it. Precisely this point was also remarked in the recent GDC 2013:

It's not just about an x86 solution, but it's about that Jaguar APU where it's a combination of the graphics and CPU together and being able to create something that's greater than just putting an x86 PC-like architecture together

[...]

For us, really by looking at that APU that we designed, you can't pull out individual components off it and hold it up and say, 'Yeah, this compares to X or Y.'

It's that integration of the two, and especially with the amount of shared memory [8GB of GDDR5, 176GB/s raw memory bandwidth] that Sony has chosen to put on that machine, then you're going to be able to do so much more moving and sharing that data that you can address by both sides.

5) Nobody said that eight-cores games will be ready in two months, but it is evident that eight cores games will be developed for both the PS4 and the eight-cores PCs.

Your comparison with the old PC gaming ecosystem is not valid, because that ecosystem was not targeted towards the eight-core PCs.

The old Xbox used a tri-core cpu based in the Cell chip cores and Cell was a complete nightmare for game developers. E.g. quake developers blamed Microsoft for their choice of the hardware on the Xbox 360. Those days, quake developers are praising Sony for their choice of an AMD eight-cores APU.

6) The "2 x 4GB GSkill DDR-3 1866 @ 9.10.9.27" where used for both the trinity and the i3 in the Sandra sp5 test that gives virtually the same memory bandwith for both chips. The memory modules that you cite now where used in an older software version in another comparison, one where the AMD trinity was run with underclocked memory.

Moreover, the figure that you link contains trinity/llano data which was copied and pasted from a previous preview

which was done with a motherboard could not boot at some speeds and with memory modules not accepting some manual configurations. I already said this to you.

In any case this is rather irrelevant per previous comments made about the relationship between this synthetic benchmark and real-word performance where trinity can up to twice the number of FPS with faster ram.

Look this is not going anywhere. I'd give no more reason to trust a manufacturer over someone else (other than hard claims such as 128 bit interfact). Claims of "up to 20% gain" are worthless. Look at bulldozer and what it was promised to be and what it delivered. There is a strong link between theoretical and actual performance in many cases. Its not an absolute but there is s strong relationship (ex the link between a synthetic such as cinebench and real world-adobe cs encoding is fairly linear).

The sandra sp5c results are one set of results that show a sandy bridge i3 (not ivy) still doing better than the a10. The sandy bridge cpu is using RAM that it is not rated for (only 1333 mhz for sandy bridge) and is a clear outlier among the other data. Its using 'overclocked' RAM (as you keep putting it) and so is not expected to show good scaling.

The e-350 will not be ram bottlenecked in any way by slower ram.

i7-975 (about much faster shows little or no difference in ram speed in that test). This is triple channel ram but 800 mhz triple channel is comparable to 1333 dual channel. If a much more powerful processor shows no difference then a puny e-350 is not going to care. (Pentium 4 beats the e-350 in that test and it has less bandwidth). None of the other cpu's have any faster ram there. e-350 needs bandwidth for the igp which is not used in the test.

 

[juanrga]

Look this is not going anywhere. I'd give no more reason to trust a manufacturer over someone else (other than hard claims such as 128 bit interfact). Claims of "up to 20% gain" are worthless. Look at bulldozer and what it was promised to be and what it delivered. There is a strong link between theoretical and actual performance in many cases. Its not an absolute but there is s strong relationship (ex the link between a synthetic such as cinebench and real world-adobe cs encoding is fairly linear).

The sandra sp5c results are one set of results that show a sandy bridge i3 (not ivy) still doing better than the a10. The sandy bridge cpu is using RAM that it is not rated for (only 1333 mhz for sandy bridge) and is a clear outlier among the other data. Its using 'overclocked' RAM (as you keep putting it) and so is not expected to show good scaling.

The e-350 will not be ram bottlenecked in any way by slower ram.

i7-975 (about much faster shows little or no difference in ram speed in that test). This is triple channel ram but 800 mhz triple channel is comparable to 1333 dual channel. If a much more powerful processor shows no difference then a puny e-350 is not going to care. (Pentium 4 beats the e-350 in that test and it has less bandwidth). None of the other cpu's have any faster ram there. e-350 needs bandwidth for the igp which is not used in the test.

1) Yes, Bulldozer was a kind of disappointment for many people (Intel also makes mistakes). But I again wonder how you select a special case and take it as the general rule: as-this-happened-once-it-will-happen-now, as-this-game-run-bad-many-games-will-run-bad... You also miss that there are more than one company involved in the development of the PS4.

2) The Sandra sp5 results show that trinity A10 performs on the same league than the sandy i3 regarding the memory subsystem:

The difference of about 2.6 Gb/s is really minimal for everyone except you. It is specially minimal when one notices that the A10 was run at stock speeds whereas the i3 was run with memory overclocked by a factor of about 40% (on MHz) on a Z77 motherboard. The immense majority of owners of an i3 will be using stock speed memory and cheaper chipset motherboards and their bandwithds will be smaller. Next is an older test on an ivy i7 that gives an idea of the sensibility of this test to changes on the memory speed:

The bandwidth increases about 7.5 Gb/s when memory is overcloked by a factor of 40% (MHz). Even if you take a half of that value for the sandy i3, this closes the above 2.6 Gb/s gap.

Therefore, it is fair to say that trinity at stock is on pair with an i3 (sandy) at stock. Your claims against AMD were a complete exaggeration.

 

[whyso]

Look this is not going anywhere. I'd give no more reason to trust a manufacturer over someone else (other than hard claims such as 128 bit interfact). Claims of "up to 20% gain" are worthless. Look at bulldozer and what it was promised to be and what it delivered. There is a strong link between theoretical and actual performance in many cases. Its not an absolute but there is s strong relationship (ex the link between a synthetic such as cinebench and real world-adobe cs encoding is fairly linear).

The sandra sp5c results are one set of results that show a sandy bridge i3 (not ivy) still doing better than the a10. The sandy bridge cpu is using RAM that it is not rated for (only 1333 mhz for sandy bridge) and is a clear outlier among the other data. Its using 'overclocked' RAM (as you keep putting it) and so is not expected to show good scaling.

The e-350 will not be ram bottlenecked in any way by slower ram.

i7-975 (about much faster shows little or no difference in ram speed in that test). This is triple channel ram but 800 mhz triple channel is comparable to 1333 dual channel. If a much more powerful processor shows no difference then a puny e-350 is not going to care. (Pentium 4 beats the e-350 in that test and it has less bandwidth). None of the other cpu's have any faster ram there. e-350 needs bandwidth for the igp which is not used in the test.

1) Yes, Bulldozer was a kind of disappointment for many people (Intel also makes mistakes). But I again wonder how you select a special case and take it as the general rule: as-this-happened-once-it-will-happen-now, as-this-game-run-bad-many-games-will-run-bad... You also miss that there are more than one company involved in the development of the PS4.

2) The Sandra sp5 results show that trinity A10 performs on the same league than the sandy i3 regarding the memory subsystem:

The difference of about 2.6 Gb/s is really minimal for everyone except you. It is specially minimal when one notices that the A10 was run at stock speeds whereas the i3 was run with memory overclocked by a factor of about 40% (on MHz) on a Z77 motherboard. The immense majority of owners of an i3 will be using stock speed memory and cheaper chipset motherboards and their bandwithds will be smaller. Next is an older test on an ivy i7 that gives an idea of the sensibility of this test to changes on the memory speed:

The bandwidth increases about 7.5 Gb/s when memory is overcloked by a factor of 40% (MHz). Even if you take a half of that value for the sandy i3, this closes the above 2.6 Gb/s gap.

Therefore, it is fair to say that trinity at stock is on pair with an i3 (sandy) at stock. Your claims against AMD were a complete exaggeration.

AMD also claimed "up to 56% performance gain with trinity vs llano" for mobile. Anandtech article for a10-4600m. In reality trinity increased games performance by 19% average and 3d mark performance by 35%. 19 vs 56 seems a massive exxageration but all companies to this (nvidia is especially bad).

Yes the bandwidth increases by 7.5 GB/s that is however an increase in bandwidth of 7.5/17.1=44%. That is impressive scaling.

That cpu is sandy bridge not ivy bridge.

That same graph shows trinity with less bandwidth than llano, stepping back from your previous generation is kinda a fail.

If we assume that trinity is strongly bandwidth bound (which scaling seems to suggest) we would expect trinity to perform (17.8-15.2)/15.2=17% better with i3 level bandwidth. That is basically the improvement of richland on the high end level. (If you look at the gpu clocks the richland a10 is only about 6% faster than the a10 trinity).

I don't think you also got the part where I said the immense majority of owners of trinity will not be using 1866 mhz ram (I don't think any non-customizable storebought trinity system shipping in volume uses ram faster than 1600 mhz). I don't think systems with a i7 or 8350 are using 1866 mhz ram. Basically the only people who are using trinity with faster ram are those who are building it themself, which is a small section of the market.

Anyway here is the memory bandwidth of my system using the newest version of sandra (released march 24). i7-3630qm 8 GB CL 11 DDR3 1600. Approximately 20 GB/s. Which matches to the i7-3770k (its a desktop using lower latency ram + margin of error). This is an i7 but its a mobile version using mobile RAM at a higher latency.

 

[utroz]

You're neglecting the fact that the GPU cores on the Richland chip are GCN cores, not last gen technology like Trinity...

Richland is trinity 2.0, just a new silicon revison, SAME GPU AS TRINITY... NO GCN for A series APU's till kaveri...
Source: http://techreport.com/news/24277/leaked-richland-specs-reveal-higher-clock-speeds

 

[8350rocks]

"Some fanboys must be worried because their high-end four-core PCs will be unable to play those games..."

There is no way you typed that with a straight face.

You can call me a fanboy all you want, but my 2-year-old i5-2500k + cf 5850s is not going to struggle with any "next-gen" console port running at the same settings. These consoles aren't going to revolutionize anything.

Ok, we'll see what happens when they arrive.

If slow 4-core / 8-thread CPUs are so much better than fast 4-core / 4-thread CPUs, why doesn't AMD just make even slower 16 or 32-thread CPUs? Because it doesn't work that way.

You can't say 8-threads on chip A are better than 4-threads on a different-architecture chip B any more than you can say 3.5GHz is always faster than 3.0GHz. There is much more that goes into CPU speed than thread counts and clock speed.

And to clarify, AMD is not making 8-CORE chips any more than Intel's i7s are 8-cores. They are making 4-core / 8-thread chips. If a "core" is sharing logic resources with another "core", it's not an individual core -- it's a processing thread.

Actually, in intel terms you're right...because all intel threads share resources with their cores.

In AMD terms, you're wrong...here's why:

AMD Modules have 4 times the L2 cache size being shared...so if you cut each modules cache in half, it's still twice the size of the intel chips cache. Plus, when you run threads on intel chips, you're sapping resources from a "real core". In an AMD processor, the architecture is designed so 8 cores can run up to 3 threads each, and the FPUs can run 2 threads each...so if you're running 8 or less threads in HSA architecture, you can actually not tap resources from one of your other cores. In intel, once you go over 4 threads, you're splitting up resources. So, the architecture differences, (as you pointed out above) mean that you cannot generically say that 8 cores are hyperthreading, because it's not the same thing.

Now, when steamroller comes (I realize this is slightly putting the cart before the horse, but bare with me...)...each core can run 4 threads and each FPU can 3 threads. The instructions will be more streamlined internally and a 30% single thread performance gain is expected...as well as a 15-20% performance gain in multi-threaded applications.

 

[thdarkshadow]

I would say just wait till both series's come out and then when we have a better idea then we can all make educated decisions.

 

[whyso]

The good thing about hyperthreading is that it is very power efficient in heavily threaded workloads.

Take a real world benchmark. The i7 is 34% faster than the i5 when it only uses a maximum of 13% more power for the system (idle usage is 67 watts so using just cpu performance power increase is 31%-- 34 vs 31 is an increase in efficiency). Considering that power vs performance does not increase linearly this represents a good gain.

AMD's method is good too in that they get much more performance than hyperthreading, unfortunately the module architecture only stretches its legs when at full load. Some tests show that there is a performance gain by deactivating every core of the bulldozer/piledriver module in single thread loads.

 

[juanrga]

AMD also claimed "up to 56% performance gain with trinity vs llano" for mobile. Anandtech article for a10-4600m. In reality trinity increased games performance by 19% average and 3d mark performance by 35%. 19 vs 56 seems a massive exxageration but all companies to this (nvidia is especially bad).

Yes the bandwidth increases by 7.5 GB/s that is however an increase in bandwidth of 7.5/17.1=44%. That is impressive scaling.

That cpu is sandy bridge not ivy bridge.

That same graph shows trinity with less bandwidth than llano, stepping back from your previous generation is kinda a fail.

If we assume that trinity is strongly bandwidth bound (which scaling seems to suggest) we would expect trinity to perform (17.8-15.2)/15.2=17% better with i3 level bandwidth. That is basically the improvement of richland on the high end level. (If you look at the gpu clocks the richland a10 is only about 6% faster than the a10 trinity).

I don't think you also got the part where I said the immense majority of owners of trinity will not be using 1866 mhz ram (I don't think any non-customizable storebought trinity system shipping in volume uses ram faster than 1600 mhz). I don't think systems with a i7 or 8350 are using 1866 mhz ram. Basically the only people who are using trinity with faster ram are those who are building it themself, which is a small section of the market.

Anyway here is the memory bandwidth of my system using the newest version of sandra (released march 24). i7-3630qm 8 GB CL 11 DDR3 1600. Approximately 20 GB/s. Which matches to the i7-3770k (its a desktop using lower latency ram + margin of error). This is an i7 but its a mobile version using mobile RAM at a higher latency.

1) In the first place I find truly amazing how you pick a 56% from Anandtech, then a 19% and a 35% from two arbitrary tests and you decide to compare the 56 against the 19, but not against the 35 in your new attempt to bash AMD. In any case, this is unimportant, because once again you and Anandtech are plain wrong.

Anandtech is omitting important info... What AMD said was that trinity A10-4600M would give up to up to a 56% increase over Llano A8-3500M in visual performance based on the figures of 3D Mark Vantage.

http://ir.amd.com/phoenix.zhtml?c=74093&p=irol-newsArticle&ID=1695793

Let us then see to what finally measured

The measured increase in performance is of 55%, which is close to the promised 56%.

Moreover, trinity was found to be better than promised

http://news.softpedia.com/news/AMD-Trinity-A8-4500M-APU-Benchmarked-Better-Than-Expected-262701.shtml

2) In the past, you said that AMD trinity was a fail because i3 sandy was better on Sandra. I showed you this was not true and now you change the claim to AMD trinity is a fail because llano was better on Sandra. Well if you were to use your logic ("stepping back from your previous generation is kinda a fail") with ivy Bridge you would claim that ivy Bridge is a fail because an i7-3770k (ivy) gives less bandwidth than an i5-2550k (Sandy) in the same test

 

[juanrga]

"Some fanboys must be worried because their high-end four-core PCs will be unable to play those games..."

There is no way you typed that with a straight face.

You can call me a fanboy all you want, but my 2-year-old i5-2500k + cf 5850s is not going to struggle with any "next-gen" console port running at the same settings. These consoles aren't going to revolutionize anything.

Ok, we'll see what happens when they arrive.

Yes. Meanwhile, we can read what Sony said some days ago at Game Developers Conference:

Using DDR5 RAM inside the PS4 will mean that the system will be able to produce a higher quality of games than some high-end systems that are only equipped with DRR3 based memory.

or read enthusiasts

http://gamingbolt.com/ps4-will-will-out-power-most-pcs-for-years-to-come-avalanche-studios-cto

 

[hapkido]

"Some fanboys must be worried because their high-end four-core PCs will be unable to play those games..."

There is no way you typed that with a straight face.

You can call me a fanboy all you want, but my 2-year-old i5-2500k + cf 5850s is not going to struggle with any "next-gen" console port running at the same settings. These consoles aren't going to revolutionize anything.

Ok, we'll see what happens when they arrive.

Yes. Meanwhile, we can read what Sony said some days ago at Game Developers Conference:

Using DDR5 RAM inside the PS4 will mean that the system will be able to produce a higher quality of games than some high-end systems that are only equipped with DRR3 based memory.

or read enthusiasts

http://gamingbolt.com/ps4-will-will-out-power-most-pcs-for-years-to-come-avalanche-studios-cto

Some console developer says consoles are better than PCs... big deal. That's like a car salesman telling you a new SUV gets better gas mileage than a two-year old compact.

GPUs have been using GDDR5 for years. Everyone's trying to act like these consoles are going to show us tons of new hardware we haven't seen before. They're not. The chips themselves may be new models, but they're not new technology.

Anyone who spent $1000 in parts (spent wisely) on a gaming PC with the last 1-2 years will have something much faster than a next gen console. If you think a console with an APU is going to outperform a CPU and dedicated GPU using hundreds of more watts of power, you're drinking the marketing kool-aid.

If AMD had that technology, why aren't they putting it in their own parts? So they have this <150W APU they can manufacture for < $200 that outperforms an 8350 and 7970 put together... but they're saving it for Sony? Get out of here with that bull-ess. Are you guys even thinking about this garbage before you pass it off?

 

[8350rocks]

Some console developer says consoles are better than PCs... big deal. That's like a car salesman telling you a new SUV gets better gas mileage than a two-year old compact.

GPUs have been using GDDR5 for years. Everyone's trying to act like these consoles are going to show us tons of new hardware we haven't seen before. They're not. The chips themselves may be new models, but they're not new technology.

Anyone who spent $1000 in parts (spent wisely) on a gaming PC with the last 1-2 years will have something much faster than a next gen console. If you think a console with an APU is going to outperform a CPU and dedicated GPU using hundreds of more watts of power, you're drinking the marketing kool-aid.

If AMD had that technology, why aren't they putting it in their own parts? So they have this <150W APU they can manufacture for < $200 that outperforms an 8350 and 7970 put together... but they're saving it for Sony? Get out of here with that bull-ess. Are you guys even thinking about this garbage before you pass it off?

DDR5 is coming to desktops...and it's onboard on the consoles(in the form of shared DDR5), so they have access to RAM that you don't which means the CPU/GPU will be communicating data back and forth about 2-3x faster than your PC could ever dream of...so unless you're running something like a HD 7990, you won't be in the same ballpark.

 

[hapkido]

So now consoles are going to be as powerful as a PC with a 7990? AMD themselves can't make a single GPU chip as powerful as their own 7990, but you think they're going to make a low power APU that's faster?

You are so full of it -- stop misleading people.

 

[8350rocks]

So now consoles are going to be as powerful as a PC with a 7990? AMD themselves can't make a single GPU chip as powerful as their own 7990, but you think they're going to make a low power APU that's faster?

You are so full of it -- stop misleading people.

I am talking about data transfer rates...follow along...stop falling behind.

The 8 GB of DDR5 RAM on the console systems is MORE RAM than you have on a HD 7990, and data transfer rates are the same, so speed of data transfer is going to be even higher.

Plus, in the new architecture, CPU and GPU will be able to share memory addresses for specific things so they can work on the same threads/applications simultaneously.

THAT'S why Sony/Microsoft chose AMD...AMD already supports this technology in their GPUs and the Richland APUs are going to get it integrated. Plus, the FX series CPUs are going to be receiving a similar instruction update in the Steamroller iteration.

 

[hapkido]

Go back under your bridge.

 

[8350rocks]

Go back under your bridge.

Why did you wander so far away from yours in the first place?

 

[whyso]

AMD also claimed "up to 56% performance gain with trinity vs llano" for mobile. Anandtech article for a10-4600m. In reality trinity increased games performance by 19% average and 3d mark performance by 35%. 19 vs 56 seems a massive exxageration but all companies to this (nvidia is especially bad).

Yes the bandwidth increases by 7.5 GB/s that is however an increase in bandwidth of 7.5/17.1=44%. That is impressive scaling.

That cpu is sandy bridge not ivy bridge.

That same graph shows trinity with less bandwidth than llano, stepping back from your previous generation is kinda a fail.

If we assume that trinity is strongly bandwidth bound (which scaling seems to suggest) we would expect trinity to perform (17.8-15.2)/15.2=17% better with i3 level bandwidth. That is basically the improvement of richland on the high end level. (If you look at the gpu clocks the richland a10 is only about 6% faster than the a10 trinity).

I don't think you also got the part where I said the immense majority of owners of trinity will not be using 1866 mhz ram (I don't think any non-customizable storebought trinity system shipping in volume uses ram faster than 1600 mhz). I don't think systems with a i7 or 8350 are using 1866 mhz ram. Basically the only people who are using trinity with faster ram are those who are building it themself, which is a small section of the market.

Anyway here is the memory bandwidth of my system using the newest version of sandra (released march 24). i7-3630qm 8 GB CL 11 DDR3 1600. Approximately 20 GB/s. Which matches to the i7-3770k (its a desktop using lower latency ram + margin of error). This is an i7 but its a mobile version using mobile RAM at a higher latency.

1) In the first place I find truly amazing how you pick a 56% from Anandtech, then a 19% and a 35% from two arbitrary tests and you decide to compare the 56 against the 19, but not against the 35 in your new attempt to bash AMD. In any case, this is unimportant, because once again you and Anandtech are plain wrong.

Anandtech is omitting important info... What AMD said was that trinity A10-4600M would give up to up to a 56% increase over Llano A8-3500M in visual performance based on the figures of 3D Mark Vantage.

http://ir.amd.com/phoenix.zhtml?c=74093&p=irol-newsArticle&ID=1695793

Let us then see to what finally measured

The measured increase in performance is of 55%, which is close to the promised 56%.

Moreover, trinity was found to be better than promised

http://news.softpedia.com/news/AMD-Trinity-A8-4500M-APU-Benchmarked-Better-Than-Expected-262701.shtml

2) In the past, you said that AMD trinity was a fail because i3 sandy was better on Sandra. I showed you this was not true and now you change the claim to AMD trinity is a fail because llano was better on Sandra. Well if you were to use your logic ("stepping back from your previous generation is kinda a fail") with ivy Bridge you would claim that ivy Bridge is a fail because an i7-3770k (ivy) gives less bandwidth than an i5-2550k (Sandy) in the same test

Regardless, the 56% was far less than the actual performance increase measured in games. Sorry I did not see that quote. AMD intentionally lead people to believe that trinity would be a massive gain in gpu performance when really in REAL WORLD tests the performance gain was about 1/3. The promised 20-40% gain of richland over trinity could be similar.

For an apu memory bandwidth is important. For strict cpu related tasks its significantly less so. I said this plenty of times.
I said many times that you must use a margin of error of around 5-10% when looking at something like memory bandwidth. 0.5 GB/s/20 is pretty insignificant.

Anyway here's another comparison of memory bandwidth between the a10 and the i3.

Here is the other slide from toms hardware that has been posted before. Bandwidth is more than before. Either of way its margin of error.

 

[hapkido]

Go back under your bridge.

Why did you wander so far away from yours in the first place?

Post some data and benchmarks for your outrageous, trolling claims.

 

[8350rocks]

Go back under your bridge.

Why did you wander so far away from yours in the first place?

Post some data and benchmarks for your outrageous, trolling claims.

Benchmarks on PS4? That's a good one...

Maybe while I am benchmarking product that isn't out yet, I can also find a way to turn lead into gold?

You can find the specs...you can read about the system...google PS4 specs...I bet there are no less than 10-15 articles that come up right away talking about the hardware...

Do I have to explain the difference between DDR3 and DDR5 as well, or have you got a grasp on that?

 

[juanrga]

Some console developer says consoles are better than PCs... big deal. That's like a car salesman telling you a new SUV gets better gas mileage than a two-year old compact.

GPUs have been using GDDR5 for years. Everyone's trying to act like these consoles are going to show us tons of new hardware we haven't seen before. They're not. The chips themselves may be new models, but they're not new technology.

Anyone who spent $1000 in parts (spent wisely) on a gaming PC with the last 1-2 years will have something much faster than a next gen console. If you think a console with an APU is going to outperform a CPU and dedicated GPU using hundreds of more watts of power, you're drinking the marketing kool-aid.

If AMD had that technology, why aren't they putting it in their own parts? So they have this <150W APU they can manufacture for < $200 that outperforms an 8350 and 7970 put together... but they're saving it for Sony? Get out of here with that bull-ess. Are you guys even thinking about this garbage before you pass it off?

1) It is not "Some console developer", it is the entire gaming developers ecosystem which is impressed by Sony next console. John Carmack (Id Software) already praised Sony for their choice of hardware; Eurogamer calls the graphics part of the console "impressive"; the console is being considered a "perfect gaming PC"; in fact, Mark Rein (Epic Games) has said yesterday about the PS4:

It's like giving you the world's best PC.

2) Believe it or not, but Sony, AMD, game developers... all they know that GDDR5 has been available on GPUs (yes the people who makes the cards and the games know this trivial stuff). The point is that most graphics cards for PCs have only 1 or 2 GB of GDDR5, with very high-end cards having 4 GB. This console has 8GB GDDR5 (not even an overly-expensive Titan has that!)

3) You also miss that this console is having a unified memory model and that its basic memory subsystem gives about 120 GB/s of bandwidth, whereas your high-end PC will be giving, at best, 25 GB/s.

4) You also miss that this console will be accessible close to bare metal level which is something you cannot even dream in your Windows 7/8 high-end PC.

 

[hapkido]

8350, if you're going to make claims a PS4 is faster than a modern gaming PC, post your sources. It's a stupid claim and I'm not going to do research to prove you wrong -- I don't have to -- the burden of proof is on the one making the stupid claims. I've caught you lying about things multiple times.

juanrga, I understand a PS4 will use shared GDDR5. The CPU portion of the APU will have faster system memory to use -- that should help with some tasks -- but CPUs now aren't memory constrained. The benchmarks you guys are passing around are largely irrelevant because the effects essentially just show up in memory benchmarks. There's a reason there hasn't been a big push for DDR4 on desktops; it's not needed. Also, as a whole, even with faster memory, the CPU is still going to be a lot slower than modern quad core CPUs. You guys try and try to dispute this all you want, but they simply can't overcome physics. It's like expecting a 35W laptop part to keep up with a 130W server part. Or a 5W phone SoC to keep up with a 35W laptop CPU. In order to match a modern desktop CPU, they're going to need to match the transistor count and also the frequency -- that's not happening. To be clear I'm talking about chips manufactured at the same time. Don't bother going back in time to bring up 386s or something.

GPUs already use GDDR5 (and have been for years), so I don't know why anyone is expecting increased memory bandwidth for the component most responsible for graphics quality. Regardless, memory bandwidth is only responsible for part of the GPU's speed. The GPU in consoles is going to be entry-level gaming PC level, that's about how good the games will look on consoles. Again, they're not going to overcome physics. If AMD could make an APU that could compete with their enthusiast gaming cards for less money, they'd already be doing it.

Consoles are going with shared GDDR5 probably because it's faster for the GPU (edit: as opposed to using DDR3 for the whole system like Trinity, for example), and also cheaper to use one set of memory instead of two separate sets. They could have used DDR3 for the system memory and it wouldn't matter from a performance standpoint, but it would be more expensive to manufacture the unit.

 

[hapkido]

Forgot to mention this. Console developers aren't programming in machine language for the same reason PC developers aren't. Not that it's not possible, but because it takes forever (as opposed to using more powerful high-level languages) and you don't get much benefit from it. Program inefficiency is largely due to lazy developing, not poor high-level compilers. Console developers may be more diligent at making their code efficient, but then again, they have to be.

 

[8350rocks]

http://www.pcper.com/news/Graphics-Cards/PlayStation-4-PS4-Orbis-Hardware-Specifications-AMD-APU-or-Discrete

There, those guys got a hold of a developer kit and broke it down...

The developer kit wouldn't be equal because of the lack of DDR5 memory, but that should give you some ideas...the developer kit curiously includes a FX8350

 

[hapkido]

Well at least you tried. This is better than you making wild, baseless claims.

"AMD has told us several times that Kaveri should be able to hit the 1.0 TFLOPs level of performance and if we compare to the current discrete GPUs would enable graphics performance similar to that of an under-clocked Radeon HD 7770."

So the guys from the site you linked (assuming AMD isn't embellishing like they did with Trinity) agree with me that the PS4 will be about as powerful as a current entry-level gaming PC. What have I been saying the entire time?... Next-gen consoles will be comparable to current entry-level gaming desktop PCs -- or relatively high-end gaming notebook PCs.

Development kits are just that. They allow developers to start making games before hardware is finalized. Otherwise there wouldn't be any games for 1-2 years after launch.

 

[8350rocks]

Well at least you tried. This is better than you making wild, baseless claims.

"AMD has told us several times that Kaveri should be able to hit the 1.0 TFLOPs level of performance and if we compare to the current discrete GPUs would enable graphics performance similar to that of an under-clocked Radeon HD 7770."

So the guys from the site you linked (assuming AMD isn't embellishing like they did with Trinity) agree with me that the PS4 will be about as powerful as a current entry-level gaming PC. What have I been saying the entire time?... Next-gen consoles will be comparable to current entry-level gaming desktop PCs -- or relatively high-end gaming notebook PCs.

Development kits are just that. They allow developers to start making games before hardware is finalized. Otherwise there wouldn't be any games for 1-2 years after launch.

Sony has already stated onboard graphics will be 1.84 TFLOPS capable, which is between HD 7870 and HD 7950 performance for graphics. The speculated speed of the processor is ~2.0-2.4 GHz with 8 cores.

Furthermore, if the development kits shipped with the FX8350, ports to PC should go exceptionally well...

So, CPU wise, you'll be on par with a mid-high end 4 core because of data transfer rates that aren't available on desktops, and graphics wise, it's equivalent to roughly a HD 7890 (if such an animal existed)...with 4x the memory.

 

[hapkido]

POST

YOUR

SOURCE

Your speculation means nothing to anyone, anywhere. What you say is contrary to current technology and the laws of thermodynamics. I've already proven you'll lie about facts.

I'm tired of talking to you. Is there a way I can unsubscribe to thread updates?