So I got my 8350 @4.3 Ghz right now with my memory in dual channel @ 1866Mhz. I can tell you, THIS THING FLYS!!! My Gawd does it fly. From a phenom Owner, to this CPU. This thing kicks the Phenom II in the nuts and steals its lunch. All that crap you heard about the 8350 not supporting dual channel 1866Mhz is bull crap. I think you just have to have the right brand of RAM and the right ammount. I think anything above 16Gb you'll run into dual channel issues. But I see the rest of what people say about the FX 8350 as false.
I look forward to buying a APU for my mATX rig. I think the biggest issue on single core performance is the lack of L3.
BUT BUT BUT 20% MOAR speed
Someone how this has to be Intel's fault probably those patents
at least the aida64 bench debunks the patent claim. it always seemed fishy to me, but i didn't have any proof. this means that quite a bit of the untappable potential in the earlier cores have opened up in steamroller. that's good news.
however, the memory bw benchmark sank my heart a bit. it's turning out the way i feared - the igpu may be more starved than ever before. i am still blaming it on lack of software optimization, or hoping may be linux benches will show otherwise (steamos) but amd seemingly chose to feed the cpu cores more instead of the igpu. i read earlier somwhere, xbitlabs may be, that the igpu can write in cpu's L2 cache. i am waiting to see how that turns out.
Hmm horrible CPU performance, ¡t looks like in the end Kaveri is not worth the upgrade for me anymore... oh well, ill just stick to my current CPU (980BE at 3.9Ghz) and get a R7 270X instead and a new HDD and more Ram instead of a new FM2+ system with Kaveri.
There are no Single Core benchmarks in there, but the Muti Threaded is horrible, ill be the first to declare i was wrong, i was expecting more, and kaveri just does not cut it for me... from what i understand in those benchmarks Kaveri barely beats Trinity, were not even talking Richland here.
noob2222 :
Adjusting clock speed thats a 6% IPC improvement. Cinebench is fairly independent of cache performance so this is mostly the "core" improvement.
The game benches look more promising hitting closer to the 20% IPC mark in a couple of games but still hitting the 6% mark on others.
The cache memory scores are a good improvement, but the external memory latency sucks compared to the 5800k.
Games have the advantage of the stronger GPU, which you may note I've never really made a prediction on. Still, note we're still seeing single digit IPC increases, now across TWO sites, for some of the CPU tests. Who would have thought I might have been optimistic?
And the cache/memory latency looks horrific. Increasing bandwidth is good for the GPU, but increasing latency is bad for the CPU. Might explain the CPU scores somewhat.
We'll know in a few days. But right now, the early benches are not looking good for AMD's per-core performance.
AMD Really needed Kaveri to be awesome, but its not even so-so... how the hell could they screw it so much? from those new benchmarks it would mean that Richland and Kaveri may be on par.
That`s just sad for AMD, no wonder why they been so quiet on the CPU part... bragging so much in the GPU part, Kaveri future looks really really bad, bulldozer 2.0.
If the GPU is good enough it may be a nice great platform for Casual Gamers or people on a really strain Budget, but asking 175 USD looks quite bad with the extremely horrible performance.
Well i can still get some performance out of the 980 at 3.9Ghz if i get a R9 270X and MANTLE, i really wanted to upgrade but i am not moving to AM3+, thats a dead socket most probably up this far.
AMD would had better performance by just using Piledriver in Kaveri and just added more speed with less energy, i see a heck ton of raging fanboys and possibly bad press heading towards AMD after release.
Hmm horrible CPU performance, ¡t looks like in the end Kaveri is not worth the upgrade for me anymore... oh well, ill just stick to my current CPU (980BE at 3.9Ghz) and get a R7 270X instead and a new HDD and more Ram instead of a new FM2+ system with Kaveri.
There are no Single Core benchmarks in there, but the Muti Threaded is horrible, ill be the first to declare i was wrong, i was expecting more, and kaveri just does not cut it for me... from what i understand in those benchmarks Kaveri barely beats Trinity, were not even talking Richland here.
We all have many reasons to to skeptical about Kaveri because of AMD's past promises but I wouldn't take these pre-launch benchmarks seriously tbh. Have you ever considered that Kaveri may be under-performing because compatible catalyst drivers for the APU aren't even out yet? I think these benchmarks are premature. We should wait for proper software support and drivers that officially support Kaveri and then properly benchmark it.
Hmm horrible CPU performance, ¡t looks like in the end Kaveri is not worth the upgrade for me anymore... oh well, ill just stick to my current CPU (980BE at 3.9Ghz) and get a R7 270X instead and a new HDD and more Ram instead of a new FM2+ system with Kaveri.
There are no Single Core benchmarks in there, but the Muti Threaded is horrible, ill be the first to declare i was wrong, i was expecting more, and kaveri just does not cut it for me... from what i understand in those benchmarks Kaveri barely beats Trinity, were not even talking Richland here.
We all have many reasons to to skeptical about Kaveri because of AMD's past promises but I wouldn't take these pre-launch benchmarks seriously tbh. Have you ever considered that Kaveri may be under-performing because compatible catalyst drivers for the APU aren't even out yet? I think these benchmarks are premature. We should wait for proper software support and drivers that officially support Kaveri and then properly benchmark it.
I understand, but those new benchmarks were not aimed at GPU but at the CPU with a Discrete GPU on lowest Settings, and it looks horrible... i would say Kaveri has a CPU as strong as a Richland but with a better GPU, after release with new Drivers the GPU will receive a massive boost, i can understand the Pudget benchmarks GPU performance, but their CPU benchmarks were true, Kaveri improved very little in that regard.
Maybe AMD PR is not lying, Clock for Clock Kaveri may be 20% faster IPC than Richland but the less speed does negate their gains, maybe its the reason why AMD is avoiding CPU performance in all their slides and presentations, so noob and Gamerk were right all along... ill give you that.
This is all starting to remind me of when Bulldozer was released. In the run up, you had a heap of expectations, much of it fuelled by people arguing some performance metrics based on some very thin or mistaken assumptions. BD was going to crush Intel, bwahahaha, etc.
Then the early leaks started to come out, suggesting that it wasn't going to be all that great. Of course, then you had people saying, "wait and see" or maybe "they're obviously using rubbish RAM" or suchlike, or even, "they're obviously biassed".
Then the reviews came out, and it turned the leaks were right, and not only was it not an Intel beater, it couldn't even beat a Phenom II unequivocally. And for the AMD fans, you went through the stages of grief:
Denial: Obviously the review sites are all biassed. All the benchmarks were using "Intel optimised" software, etc.
Anger: Lots of nerd rage, accusations of fanboy-ism, trolling from both sides
Bargaining: Ah, but if you disable one core per module, then you end up with a GREAT proc
Depression: AMD really messed up by dropping Phenom II
Acceptance: BD's rubbish, here's looking forward to Piledriver
^^ 20% over previous gen (a10 6800k) isn't enough? bulldozer has a Lot of room for improvement.
Those 20% would have been sweet, but i don`t see them in most leaked bechmarks... Just like Pudget showed Kaveri is Behind Richland by 5% and in the new Benchmarks it seems Kaveri barely beats Trinity by 6%.
AMD PR is right but they are not telling the whole story, Kaveri is up to 20% faster clock for clock but the gains are useless because of the lack of matching or greater speed compared to the previous generation.
I could get a much stronger GPU now that i plan to skip also on Kaveri, but my CPU may be a bottleneck to even a R9 270X, and i don´t see a CPU from AMD that gives a nice boost... if i jump to AM3+ amd will probably dump the Socket and If i jump to FM2+ with Kaveri i would not see that much of a increase from a 980BE at 3.9Ghz, i would probably win 15% more performance... and i want at least 30% more performance, i could go Intel but i decided never to support their products... so i am clueless as to what to what platform should i upgrade or if i should only upgrade other parts of my current System like a new HDD, more Ram or a GPU.
I was hoping Kaveri to be 30% or more faster than my current CPU so that the upgrade would be worth it, im left out in the cold i guess, i been using the 980 since 2010, decided to skip Bulldozer because it was such a fail... then Piledriver came and i decided to wait for Steamroller, then it got delayed and finally cancelled, then decided to just grab Kaveri but starting to feel Kaveri will be a massive disappointment.
colinp :
This is all starting to remind me of when Bulldozer was released. In the run up, you had a heap of expectations, much of it fuelled by people arguing some performance metrics based on some very thin or mistaken assumptions. BD was going to crush Intel, bwahahaha, etc.
Then the early leaks started to come out, suggesting that it wasn't going to be all that great. Of course, then you had people saying, "wait and see" or maybe "they're obviously using rubbish RAM" or suchlike, or even, "they're obviously biassed".
Then the reviews came out, and it turned the leaks were right, and not only was it not an Intel beater, it couldn't even beat a Phenom II unequivocally. And for the AMD fans, you went through the stages of grief:
Denial: Obviously the review sites are all biassed. All the benchmarks were using "Intel optimised" software, etc.
Anger: Lots of nerd rage, accusations of fanboy-ism, trolling from both sides
Bargaining: Ah, but if you disable one core per module, then you end up with a GREAT proc
Depression: AMD really messed up by dropping Phenom II
Acceptance: BD's rubbish, here's looking forward to Piledriver
Pretty much.
There are several aspects of Kaveri that are still very worth it like HUMA, HSA, MANTLE and the GPU which is very promising, not all is pure doom, is just that if the benchmarks are saying Kaver is between Trinity and Richland in regards IPC and i have little use for it as an upgrade.
AMD went from moar cores!! to moar GPU Cores!!, well at least they managed to match Richland with less power and less speed, Steamroller arch is very good... On Paper, comes to my mind the Slide that claims 20% over Richland and in the very bottom a very very small yellow text that says:
*Pre-Silicon projections and not based on actual measured data
APU's will never compete with dGPU / dCPU's due to cost and space / heat limits, physics is a hard master.
No. The laws of physics says otherwise. That is the reason that both Nvidia and AMD engineers are designing their future supercomputers around APUs, and not around dCPU+dGPU anymore. Nvidia 40TFLOP design is scheduled for ~2018.
Hmm horrible CPU performance, ¡t looks like in the end Kaveri is not worth the upgrade for me anymore... oh well, ill just stick to my current CPU (980BE at 3.9Ghz) and get a R7 270X instead and a new HDD and more Ram instead of a new FM2+ system with Kaveri.
There are no Single Core benchmarks in there, but the Muti Threaded is horrible, ill be the first to declare i was wrong, i was expecting more, and kaveri just does not cut it for me... from what i understand in those benchmarks Kaveri barely beats Trinity, were not even talking Richland here.
We must be reading a different review, because the one that I see shows that the Steamroller CPU in Kaveri can be up to 30% faster than Piledriver, which agrees very much with previously leaked benchmarks of the ES
APU's will never compete with dGPU / dCPU's due to cost and space / heat limits, physics is a hard master.
No. The laws of physics says otherwise. That is the reason that both Nvidia and AMD engineers are designing their future supercomputers around APUs, and not around dCPU+dGPU anymore. Nvidia 40TFLOP design is scheduled for ~2018.
Hmm horrible CPU performance, ¡t looks like in the end Kaveri is not worth the upgrade for me anymore... oh well, ill just stick to my current CPU (980BE at 3.9Ghz) and get a R7 270X instead and a new HDD and more Ram instead of a new FM2+ system with Kaveri.
There are no Single Core benchmarks in there, but the Muti Threaded is horrible, ill be the first to declare i was wrong, i was expecting more, and kaveri just does not cut it for me... from what i understand in those benchmarks Kaveri barely beats Trinity, were not even talking Richland here.
We must be reading a different review, because the one that I see shows that the Steamroller CPU in Kaveri can be up to 30% faster than Piledriver, which agrees very much with previously leaked benchmarks of the ES
I actually expected Kaveri to be almost on par with the i5 2500k, but i thought it would be like 3-5% behind it, but it will still be good enough.
But you said it, up to 30% faster... but usually only 6% faster, that`s what i am seeing and avg of 6% better MultiThreaded CPU benchmarks, i assume who ever is doing the benchmarks is avoiding completely Single Core performance.
When i saw your predictions they just made sense... but i guess at the end of the day they are just only predictions and the actual benchmarks show a different picture, no one is at fault since AMD delivered 20% better IPC on their new Arch, which is great but the bad news is they took 15% less speed which would put Kaveri on par with Richland.
I guess the only one to blame is GloFo, they been hitting AMD in the balls over and over... they can`t even compete against TSMC, still wondering why AMD does not get rid of them for once, maybe TSMC is too busy and can`t really give AMD the full attention they need, so they stick to GloFo whatever TSMC cannot do due to their agenda which ends up really bad for AMD in the end.
Hopefully for AMD, GloFo gets their act together in the second half of 2014 and manages to release a Kaveri with at least the same speed as Richland, that would be game changer... Kaveri would massively benefit with a max speed of 4.4Ghz or better.
I know about Skyrim, been modding that game since release... just in case anyone wonders i am the author of SkyTEST - Realistic Animals and Predators mod, which is on the top 100 all time mods #68 with around 8,300 Endorsements, more than a Million views and Half-Million Downloads... i know Skyrim engine and been in communication with the guys that forced Bethesda to enable more compilers optimizations on PC with "SKSE TESVAL" and "SKYBoost" back in Dec of 2011.
The thing is, Kaveri is making a shameful 6% improvement over Trinity (we are not even talking Richland), i know Skyrim mostly uses 2 Cores and a Third one in a small matter, if you disable 1 Core on Skyrim you lose no FPS, if you disable a second Core you lose around 2-4 fps and Disabling a Third core you lose half your Frames, Skyrim does not even use 4 Cores apart from just spreading CPU Load to all 4 Cores, but if Kaveri would in fact be 20% faster than Richland it would show a massive boost compared to the shameful 6% it´s showing in those benchmarks, it seems to me that the 30% IPC are shown in few cases where MultiCore performance comes into Play.
Just looking at the Skyrim benchmark tells me how horrible final Performance Kaveri have in Single Core.
Just as extra, i have a friend with a i3 3220 paired with an old GTX260, we were benchmarking both out systems and right in the entrance of inside Markarth he gets 44 FPS and an average of 49 in other areas of Markarth while i with my 980 and a GTX470 get 38 FPS in the entrance and an avg of 60 in other spots, seems he got a massive Minimum while his Max speed is rather low... that´s Skyrim engine playing favorite with a stronger SingleCore performance.
Whiterun looking down from DragonReach to the Gildergreen tree:
I get 40 he gets 52, even when my GPU is twice as strong as his, but i get a constant 60 in other places while his GPU struggles to go pass 55 in areas where i could easily hit the 120 fps due to the game being capped at 60 in the ini settings.
So yeah... i know about Skyrim scores in those benchmarks are shameful for Kaveri.
Those 20% would have been sweet, but i don`t see them in most leaked bechmarks... Just like Pudget showed Kaveri is Behind Richland by 5% and in the new Benchmarks it seems Kaveri barely beats Trinity by 6%.
fudis: the 20% gain could easily be amd quoting peak gain or a "optimistic nominal" gain.
with that out of the way, are you refering to pugetsystem's gaming benches? in those, the apus appeared bottlenecked iirc. kaveri is a denser apu with similar die area as richland. so while it is 5% more power-friendly, the heat generation may be higher... the silicon choice may contribute as well. what i'm trying to say is that with kaveri, you might see more throttling at peak load with stock cooling. experts will be able to explain/correct me. i think the puget review showed some of that.
etayorius :
AMD PR is right but they are not telling the whole story, Kaveri is up to 20% faster clock for clock but the gains are useless because of the lack of matching or greater speed compared to the previous generation.
PR never tells the whole story, just the truth that helps selling the product.
er.. piledriver's turbo was never up to snuff. combined with throttling, and kaveri's inherent improvments, kaveri will likely hold consistent lead over pd. how much, will depend on the system.
etayorius :
I could get a much stronger GPU now that i plan to skip also on Kaveri, but my CPU may be a bottleneck to even a R9 270X, and i don´t see a CPU from AMD that gives a nice boost... if i jump to AM3+ amd will probably dump the Socket and If i jump to FM2+ with Kaveri i would not see that much of a increase from a 980BE at 3.9Ghz, i would probably win 15% more performance... and i want at least 30% more performance, i could go Intel but i decided never to support their products... so i am clueless as to what to what platform should i upgrade or if i should only upgrade other parts of my current System like a new HDD, more Ram or a GPU.
I was hoping Kaveri to be 30% or more faster than my current CPU so that the upgrade would be worth it, im left out in the cold i guess, i been using the 980 since 2010, decided to skip Bulldozer because it was such a fail... then Piledriver came and i decided to wait for Steamroller, then it got delayed and finally cancelled, then decided to just grab Kaveri but starting to feel Kaveri will be a massive disappointment.
.... so you're saying that your options are dead socket, >30% better cpu while intel is a no-no for priciple...? none of those sound logical. since you got phenom ii, apus were never your suitable market, it was the fx cpus or intel core i5 and higher. or kaveri if hsa gains momentum.
don't write off kaveri without seeing reliable independent reviews. keeping your expectation low is good, but being depressed isn't, lol.
colinp :
This is all starting to remind me of when Bulldozer was released. In the run up, you had a heap of expectations, much of it fuelled by people arguing some performance metrics based on some very thin or mistaken assumptions. BD was going to crush Intel, bwahahaha, etc.
Then the early leaks started to come out, suggesting that it wasn't going to be all that great. Of course, then you had people saying, "wait and see" or maybe "they're obviously using rubbish RAM" or suchlike, or even, "they're obviously biassed".
Then the reviews came out, and it turned the leaks were right, and not only was it not an Intel beater, it couldn't even beat a Phenom II unequivocally. And for the AMD fans, you went through the stages of grief:
Denial: Obviously the review sites are all biassed. All the benchmarks were using "Intel optimised" software, etc.
Anger: Lots of nerd rage, accusations of fanboy-ism, trolling from both sides
Bargaining: Ah, but if you disable one core per module, then you end up with a GREAT proc
Depression: AMD really messed up by dropping Phenom II
Acceptance: BD's rubbish, here's looking forward to Piledriver
it happens every launch. i went through stages 1 and 5 with bd, and 2 because cougar point's degrading sata bug. set me back a couple of years. it's usually the passionate supporters that have the biggest meltdowns, trolls and shills just hide away until the next launch.
etayorius :
AMD went from moar cores!! to moar GPU Cores!!, well at least they managed to match Richland with less power and less speed, Steamroller arch is very good... On Paper, comes to my mind the Slide that claims 20% over Richland and in the very bottom a very very small yellow text that says:
*Pre-Silicon projections and not based on actual measured data
gotta watch out for disclaimers.
bd was about moar cores, then went to moar gigglehurtz. kaveri is about moar cores but be careful what you wish for. unlike pd pver bd, though, kaveri is shaping up to be a legitimate/solid upgrade at entry level. whether it came out too late or not, will be decided by users.
Those 20% would have been sweet, but i don`t see them in most leaked bechmarks... Just like Pudget showed Kaveri is Behind Richland by 5% and in the new Benchmarks it seems Kaveri barely beats Trinity by 6%.
fudis: the 20% gain could easily be amd quoting peak gain or a "optimistic nominal" gain.
with that out of the way, are you refering to pugetsystem's gaming benches? in those, the apus appeared bottlenecked iirc. kaveri is a denser apu with similar die area as richland. so while it is 5% more power-friendly, the heat generation may be higher... the silicon choice may contribute as well. what i'm trying to say is that with kaveri, you might see more throttling at peak load with stock cooling. experts will be able to explain/correct me. i think the puget review showed some of that.
etayorius :
AMD PR is right but they are not telling the whole story, Kaveri is up to 20% faster clock for clock but the gains are useless because of the lack of matching or greater speed compared to the previous generation.
PR never tells the whole story, just the truth that helps selling the product.
er.. piledriver's turbo was never up to snuff. combined with throttling, and kaveri's inherent improvments, kaveri will likely hold consistent lead over pd. how much, will depend on the system.
etayorius :
I could get a much stronger GPU now that i plan to skip also on Kaveri, but my CPU may be a bottleneck to even a R9 270X, and i don´t see a CPU from AMD that gives a nice boost... if i jump to AM3+ amd will probably dump the Socket and If i jump to FM2+ with Kaveri i would not see that much of a increase from a 980BE at 3.9Ghz, i would probably win 15% more performance... and i want at least 30% more performance, i could go Intel but i decided never to support their products... so i am clueless as to what to what platform should i upgrade or if i should only upgrade other parts of my current System like a new HDD, more Ram or a GPU.
I was hoping Kaveri to be 30% or more faster than my current CPU so that the upgrade would be worth it, im left out in the cold i guess, i been using the 980 since 2010, decided to skip Bulldozer because it was such a fail... then Piledriver came and i decided to wait for Steamroller, then it got delayed and finally cancelled, then decided to just grab Kaveri but starting to feel Kaveri will be a massive disappointment.
.... so you're saying that your options are dead socket, >30% better cpu while intel is a no-no for priciple...? none of those sound logical. since you got phenom ii, apus were never your suitable market, it was the fx cpus or intel core i5 and higher. or kaveri if hsa gains momentum.
don't write off kaveri without seeing reliable independent reviews. keeping your expectation low is good, but being depressed isn't, lol.
colinp :
This is all starting to remind me of when Bulldozer was released. In the run up, you had a heap of expectations, much of it fuelled by people arguing some performance metrics based on some very thin or mistaken assumptions. BD was going to crush Intel, bwahahaha, etc.
Then the early leaks started to come out, suggesting that it wasn't going to be all that great. Of course, then you had people saying, "wait and see" or maybe "they're obviously using rubbish RAM" or suchlike, or even, "they're obviously biassed".
Then the reviews came out, and it turned the leaks were right, and not only was it not an Intel beater, it couldn't even beat a Phenom II unequivocally. And for the AMD fans, you went through the stages of grief:
Denial: Obviously the review sites are all biassed. All the benchmarks were using "Intel optimised" software, etc.
Anger: Lots of nerd rage, accusations of fanboy-ism, trolling from both sides
Bargaining: Ah, but if you disable one core per module, then you end up with a GREAT proc
Depression: AMD really messed up by dropping Phenom II
Acceptance: BD's rubbish, here's looking forward to Piledriver
it happens every launch. i went through stages 1 and 5 with bd, and 2 because cougar point's degrading sata bug. set me back a couple of years. it's usually the passionate supporters that have the biggest meltdowns, trolls and shills just hide away until the next launch.
etayorius :
AMD went from moar cores!! to moar GPU Cores!!, well at least they managed to match Richland with less power and less speed, Steamroller arch is very good... On Paper, comes to my mind the Slide that claims 20% over Richland and in the very bottom a very very small yellow text that says:
*Pre-Silicon projections and not based on actual measured data
gotta watch out for disclaimers.
bd was about moar cores, then went to moar gigglehurtz. kaveri is about moar cores but be careful what you wish for. unlike pd pver bd, though, kaveri is shaping up to be a lagitimate upgrade. whether it came out too late or not, will be decided by users.
I am not really disappointed, i am more amused that Kaveri has such low IPC compared to Trinity (let alone Richland), My motherboard does not accept AM3+ CPUs only AM3 CPUs... so i would need a new AM3+ board plus a new FX CPU and i really fear AMD will dump AM3+ very soon... so mostly i am left out with no CPU upgrade path... jumping to Kaveri wont be such a boost and i would be spending about 450 USD in total for the new Ram, APU and Motherboard, jumping to AM3+ may be the only choice if i really want to upgrade but i would be spending another 400 USD max... and getting just a new R9 270 would be 230USD the most, so i guess ill just stick to the 980 for another year and just upgrade my GPU and pray it does not Bottlenecks the 270x... well if MANTLE comes into play i will be safe i guess.
Kaveri should be up to 10% faster than my 980, there is definitely more performance compared to my current CPU, but i dont want to spend 450 or more just to get 15% upgrade... it`s just not worth it, if would had been at least 20% i would not even consider it... not only do i need new to spend in the CPU but in the fast Ram which is very very costly and a new FM2+ MoBo.
I think my final options are either getting AM3+ Mobo with a FX-8320 and OC the heck out of it or just upgrade my GPU, buy 4 GB more of Ram and a new HDD.
Kaveri needs to be kickass... and then some. I hope that it isn't another 6800K where its marginally better than the old one with a $30-40 price increase. Some leaked prices are suggesting the A10-7850K to be around $170 which is right up there with the locked 3rd and 4th generation i5s. With the pricing, AMD must be very confident in these APUs or are trying to collect a return from investing in HSA and other technology.
I say it should at least come around Sandy Bridge or Nehalem performance through hell or high water. I hope with the whole 8 core debacle that AMD has learned that its better to improve what you have rather than introduce something radically different and hope that the industry adopts it.
Edit: I just don't want to use the AM3+ platform; its growing long in the tooth and for my next build that is uATX I have no trust in the uATX AM3+ boards at all. The FM2+ platform is newer, has more features, and just seems more refined.
No. The laws of physics says otherwise. That is the reason that both Nvidia and AMD engineers are designing their future supercomputers around APUs, and not around dCPU+dGPU anymore. Nvidia 40TFLOP design is scheduled for ~2018.
I really want to see what you say the physics are saying, because afaik: More transistors = More heat = less performance | Bigger diesize = fewer yields = higher cost
Also the problem I see with seperating the chip into 2 dies, is again afaik: space
Though there are most certainly more problems/examples than that I just used what I know so..
^^ 20% over previous gen (a10 6800k) isn't enough? bulldozer has a Lot of room for improvement.
It is if you have to reduce clocks 5% to get there. Yes, we're seeing some tasks hit the 20%. But others less then 5%. On average, we're looking 10%ish, same as "Hasfail", as some here call it.
No. The laws of physics says otherwise. That is the reason that both Nvidia and AMD engineers are designing their future supercomputers around APUs, and not around dCPU+dGPU anymore. Nvidia 40TFLOP design is scheduled for ~2018.
I really want to see what you say the physics are saying, because afaik: More transistors = More heat = less performance | Bigger diesize = fewer yields = higher cost
Also the problem I see with seperating the chip into 2 dies, is again afaik: space
Though there are most certainly more problems/examples than that I just used what I know so..
But the reason both AMD and Intel are focusing so much attention on the GPU: X86 performance has basically stalled. So either you need to move to a new arch (Good luck, considering 99% of the consumer market it X86 based), or find a way to increase performance. Moving processing to the GPU, then buffing that 30% per generation is attractive then, doesn't it?
No. The laws of physics says otherwise. That is the reason that both Nvidia and AMD engineers are designing their future supercomputers around APUs, and not around dCPU+dGPU anymore. Nvidia 40TFLOP design is scheduled for ~2018.
I really want to see what you say the physics are saying, because afaik: More transistors = More heat = less performance | Bigger diesize = fewer yields = higher cost
Also the problem I see with seperating the chip into 2 dies, is again afaik: space
Though there are most certainly more problems/examples than that I just used what I know so..
But the reason both AMD and Intel are focusing so much attention on the GPU: X86 performance has basically stalled. So either you need to move to a new arch (Good luck, considering 99% of the consumer market it X86 based), or find a way to increase performance. Moving processing to the GPU, then buffing that 30% per generation is attractive then, doesn't it?
Then we can safely call it Failveri or just Bulldozer Reloaded.
Understanding Juanrga is a lot more easier if you realize that he's basically the first real ARM fanboy to defend ARM so heavily.
The footnotes in the Nvidia slides where they're making all these claims are laughable at best. Lines like "estimates" and "20 percent margin of error" and "PS3 only using one core and no SPUs".
Nvidia always does this with Tegra. They hype it up like none-other. In fact, they even boasted Tegra 2 was on 360 levels and you'd have the power of a console in your pocket. And look how that turned out.
The truth is Tegra K1 final silicon is a long ways away, at least 10 months. We will be talking about Excavator at that point, and the other ARM companies wilil probably be with ARM64 then too.
Nvidia always starts hyping Tegra way before it's actually due. They're very good at marketing, and if Qualcomm or Samsung were out here hyping their ARM products that would ber coming out around the time of Tegra, people would be going crazy too.
You have used my comments about Denver core made in another part to launch here an anti-Nvidia rant... and you call "fanboy" to others! LOL
But at least this time you are spelling my nickname correctly. Thanks!
Haha, I'm learning
But Juan, the thing is that I don't think it's necessarily bad to be a fanboy to a certain extent. I'm arguably a fanboy for HEDT platforms with serious power, and you're more for low power ARM. I don't know, maybe fanboy is not the right word for what I'm describing, since we both are more aligned with platforms as opposed to brands.
But Nvidia has a massive history of lying about Tegra and making it sound far better than it really is. I would give it a few weeks or so until Nvidia's lies start to show up. We're already seeing them with the 2w figure being used when playing regular games, like Angry Birds and Bejeweled clones, which are all 2d games that use sprites and probably aren't even GPU accelerated.
I'm always very pessimistic when Nvidia talks about ARM. If this was Qualcomm or Samsung, or even Apple (which I despise), I'd be a little more optimistic about what they're claiming. I just feel like this is basically the 5th time Nvidia has gotten on stage and talked about how amazing Tegra was, how it would beat the x86 CPU of the day, etc.
"Tegra 2 faster than Core 2 Duo":
http://androidcommunity.com/kal-el-nvidia-chipset-benchmarked-against-core2duo-and-tegra-2-20110218/
See the trend? Nvidia has done nothing but cherry pick benchmarks and exaggerate potential of their Tegra chips for nearly as long as they have existed.
You can do your own googling and take a look, I just gave this post a few minutes of my time. But Nvidia has pretty much always lied about their performance of their Tegra chips.
I really don't see why Kaveri is a fail right off the bat looking at everything posted by you, guys.
It improves over Richland (and Trinity), even with lower clocks. Packs GCN + all the features the 7790 supports (AFAIK) and they kept it in FM2 socket. Now, I'll have to get a new MoBo, but those with low end Trinity might as well jump to Kaveri (if they exist, lol). It will be a sweet update from my good'ol Llano 3850 and Athlon II X4s in lower end MoBos (4250 iGPU). Remember they're not aimed to battle the 4770K nor the 4670K (I think) head to head, except in GPU related tasks.
I don't want to get into numerology either, because it blurs the picture a little. 20% when we're talking in large numbers are large gains, but in small numbers they're a small gain. Also, the CPU might be catching to the 2500k, but the iGPU is far ahead from the 4770K in almost every regard, right? Credits due where credits owed.
I don't think I'm being delusional, but I do think numbers just tell half the story so far. I'll wait for Tom's, Anands and Techreport's benchies to really conclude if Kaveri is worthy or not. So far, it doesn't look bad at all IMO.
^ Give up your 3850 ? Only if kaveri makes a big improvement on the cpu side. Not seeing it yet, did you notice the spanish site used a closed loop for testing ? I am starting to think all the variations in benchmarks can be attributed to thermal throttling
I kinda agree with him...although those benches are underwhelming they only tell part of a big story.
That said it's nothing to turn my head toward as an 8350 owner. But when your looking at an upgrade that adds to both CPU related functions and a sizable GPU upgrade as well, it starts to become worth it.
I kinda want a HTPC build...was thinking about a trinity build...I actually may go for it when the prices level out.
APU's will never compete with dGPU / dCPU's due to cost and space / heat limits, physics is a hard master.
No. The laws of physics says otherwise. That is the reason that both Nvidia and AMD engineers are designing their future supercomputers around APUs, and not around dCPU+dGPU anymore. Nvidia 40TFLOP design is scheduled for ~2018.
Hmm horrible CPU performance, ¡t looks like in the end Kaveri is not worth the upgrade for me anymore... oh well, ill just stick to my current CPU (980BE at 3.9Ghz) and get a R7 270X instead and a new HDD and more Ram instead of a new FM2+ system with Kaveri.
There are no Single Core benchmarks in there, but the Muti Threaded is horrible, ill be the first to declare i was wrong, i was expecting more, and kaveri just does not cut it for me... from what i understand in those benchmarks Kaveri barely beats Trinity, were not even talking Richland here.
juanrga :
We must be reading a different review, because the one that I see shows that the Steamroller CPU in Kaveri can be up to 30% faster than Piledriver, which agrees very much with previously leaked benchmarks of the ES
I actually expected Kaveri to be almost on par with the i5 2500k, but i thought it would be like 3-5% behind it, but it will still be good enough.
But you said it, up to 30% faster... but usually only 6% faster, that`s what i am seeing and avg of 6% better MultiThreaded CPU benchmarks, i assume who ever is doing the benchmarks is avoiding completely Single Core performance.
Cinebench scores can be avoided, not only they are optimized for Intel, but the new R15 version cuts down Kaveri and puts it at same level than Trinity, which is rather odd. We know that ICC forces AMD chips to run the slowest possible code to fake benchmarks, but here R15 seems to introduce extra cut-down on Kaveri. Just look at Kaveri performing much better on R11.5
Now about CPU improvements shown in the game benchmarks.
I don't know from where you obtain that Steamroller offers only a 6% IPC gain over Piledriver. The average is more than 20% and in a pair of cases it goes above 30%.
And it looks much better at compute. I can see Kaveri APU is 40% faster than Trinity APU, in Luxmark, 158% faster in Raytrace, and 177% faster in fluid simulation.