作者: vsp 時間: 2013-10-26 18:45 標題: AMD Kaveri 幻想曲 (推測有17%效能增長)
本帖最後由 vsp 於 2013-10-26 18:46 編輯
AMD FANS好安慰
http://juanrga.com/en/AMD-kaveri-benchmark.html
AMD Kaveri benchmark
Abstract: Chip-maker AMD will release a new family of APUs by the end of 2013. This new family will be named Kaveri and will substitute the current Richland family of APUs. In this article I will consider Kaveri performance using all the information that AMD has disclosed as well as my best guesses on the parts that remain unknown. I estimate that the the top Kaveri GPU will have a CPU clocked at 4 GHz with a maximum performance of 128 GFLOP and a GPU clocked at 900 MHz with 922 GFLOP, giving a total of 1050 GFLOP for the whole APU. Combining all this data, I predict that the CPU of the top Kaveri APU will be about 26% faster than top Trinity APU and about 17% faster than top Richland APU. This would put the multi-threaded performance of the CPU of the new quad core Kaveri APU at the same level than an Intel quad core i5 or a six-core AMD FX with traditional software. The new Kaveri APU will show its real strength with HSA software, which will exploit the performance of both the CPU and the GPU. With HSA enabled software, Kaveri has the potential to be much faster than an Intel i7 or an octo-core AMD FX. Some developers are finding accelerations of up to 500% when enabling HSA. A collection of APU and CPU benchmarks and scores is given.
Chip-maker AMD will release a new family of APUs by the end of 2013. This new family will be named Kaveri and will substitute the current Richland family of APUs. Kaveri will combine two to four 28nm CPU cores, based on the Steamroller architecture plus HSA improvements, with a HSA enabled Graphics Core Next GPU. This HSA Graphics Core Next architecture can work as a traditional graphics card (rendering graphics) or as co-processor (computing parallel tasks). Kaveri will also introduce a uniform memory model, dubbed hUMA, that will allow both the CPU and the GPU to access to a common memory pool.
Many sites report some of the specs of the upcoming Kaveri. In this article I will consider Kaveri performance using all the information that AMD has disclosed as well as my best guesses on the parts that remain unknown.
作者: Puff 時間: 2013-10-26 19:33
本帖最後由 Puff 於 2013-10-26 19:42 編輯
呢位兄台... 我都唔知俾乜 comment 好。睇下就算。無記錯佢應該算係調低左佢既 expectation.
17% 我個人覺得如果係講 100W top pin 係頗樂觀,當然可能 uncore 改進多線程會有 so... 但係... emmm... 除非你望 4.1+ Base 囉。反正 AMD 悲觀 d 睇實無死,好過高開低收...

作者: Schaft 時間: 2013-10-26 19:58
其實。⋯amd無進步,對大家都無好處,睇下atv幾咁好,tvb又㸃

作者: vsp 時間: 2013-10-26 20:16
有17%效能增長好好

INTEL HASWELL都係10%
作者: Puff 時間: 2013-10-26 20:24
本帖最後由 Puff 於 2013-10-26 20:39 編輯
單位呢唔該
作者: Puff 時間: 2013-10-26 20:25
本帖最後由 Puff 於 2013-10-26 20:27 編輯
大佬唔同 case 啦。呢邊係死剩三間 + patent wall + tech. advantage;果邊係從頭到尾得兩間 + 死唔發牌。
作者: 喬芬‧費曼高 時間: 2013-10-27 02:37
提示: 作者被禁止或刪除 內容自動屏蔽
作者: Puff 時間: 2013-10-27 03:29
... 但「專玩」果粒野本來就係 derived from 呢個 so-called "放棄'既市場。另外 Intel 會升呢,AMD 唔會升呢咩?
作者: ironmonkey 時間: 2013-10-27 13:35
今年R&D budget 只有上年既1/3.....
除非是快6800k 17%和65w tdp
作者: vsp 時間: 2013-10-27 18:41
kaveri頂級型號 對比 6800K有17%效能增長
haswell 對比 IVY Bridge唔知有冇10%效能增長
作者: Puff 時間: 2013-10-27 22:16
Hi performance = IPC * clock, in general
而且係呢位兄台「估計」,佢又唔係爆料。
作者: B687 時間: 2013-10-27 23:44
CPU/GPU/製程/插座一齊改
17%太少了吧
應該要有71%
作者: lcska02nkc 時間: 2013-10-27 23:46
提示: 作者被禁止或刪除 內容自動屏蔽
作者: wanch 時間: 2013-10-27 23:47
amd speed >1 17%>1.17
intel speed >10 1%>11
作者: win150 時間: 2013-10-28 16:53
AMD FANS好安慰
AMD Kaveri benchmark
Abstract: Chip-maker AMD will release a new family of APUs by the end of 2013.
vsp 發表於 2013-10-26 18:45
此post已完
amd自己講左係Feb-2014
作者: wilsonkf 時間: 2013-10-29 09:38
提示: 作者被禁止或刪除 內容自動屏蔽
作者: dom 時間: 2013-10-29 10:52
http://www.anandtech.com/show/74 ... motherboards-kaveri
自己睇
作者: dom 時間: 2013-10-29 10:54
賣OEM 優先
但係依家 Intel Ivy 都未出清庫存
就算上代 Trinity / Richland 都唔見 OEM 賣街 (Notebook 反而 一面倒 Intel .....)
其實 想搵籍口話 DELAY 到 Q1 2014 者( 改期...好聽d 既話)
作者: wilsonkf 時間: 2013-10-29 11:02
提示: 作者被禁止或刪除 內容自動屏蔽
作者: dom 時間: 2013-10-29 11:47
Kaveri OEM 無人收咪派街貨law
richland trinity 係新APU出後再減價清貨
100-110 USD 6790K 好未? Intel ...
wilsonkf 發表於 29-10-2013 11:02 AM
駛咩用 i3-4xxx 劈
Intel 用 IVY Bridge i3 劈價都夠打你
唔劈佢一樣咁賣到
你望下街賣緊既 "Richland" APU 廠機/Notebook 有幾多部先講啦
i3 就成街都係

作者: ironmonkey 時間: 2013-10-29 11:58
始終Trinity/Richland CPU太食電...
作者: wilsonkf 時間: 2013-10-29 12:00
提示: 作者被禁止或刪除 內容自動屏蔽
作者: dom 時間: 2013-10-29 12:05
Intel 照翻之前禁賣 $9xx i3 Ivy 已經夠打你

作者: wilsonkf 時間: 2013-10-29 12:28
提示: 作者被禁止或刪除 內容自動屏蔽
作者: qcmadness 時間: 2013-10-29 23:58
本帖最後由 qcmadness 於 2013-10-30 00:03 編輯
Totally wrong for mobile APU
http://www.anandtech.com/bench/Mobile/139
http://www.anandtech.com/bench/Mobile/140
Trinity / Richland is not that bad compared with similar TDP IvyBridge / Haswell CPUs
作者: ironmonkey 時間: 2013-10-30 00:53
Totally wrong for mobile APU
Trinity / Richland is not that bad compared with similar TDP IvyB ...
qcmadness 發表於 2013-10-29 23:58
I don't understand. How do power consumption of idle and internet use invalidate my point?
作者: qcmadness 時間: 2013-10-30 01:09
I don't understand. How do power consumption of idle and internet use invalidate my point?
ironmonkey 發表於 2013-10-30 00:53
because it shows that the battery life of trinity / richland notebook is not bad compared to IVB / Haswell notebooks
作者: qcmadness 時間: 2013-10-30 01:11
The major problem for Trinity / Richland is the high price instead.
作者: ironmonkey 時間: 2013-10-30 02:10
本帖最後由 ironmonkey 於 2013-10-30 02:19 編輯
because it shows that the battery life of trinity / richland notebook is not bad compared to IVB / ...
qcmadness 發表於 2013-10-30 01:09
Man, thats a pretty weak support. Firstly, there is only one trinity laptop in the list and its a "prototype". Secondly, when compared to higher load benchmarks (which you 'forgotten' to include), its ranked amongst the worst. The consumption pattern also looks similar to its desktop brethren. Thirdly, this all adds up as more costs and less return for OEMs.
http://www.anandtech.com/bench/Mobile/141
http://techreport.com/review/249 ... and-apus-reviewed/3
Of course I could be wrong and the main reasons for not using mobile APUs are profits and branding instead, but I'm pretty sure power consumption is one of them.
Edit: In fact, I'll go as far to say that their new power hungry and poorly designed CPU is the main reason why their APUs aren't flourishing.
作者: qcmadness 時間: 2013-10-30 09:42
Man, thats a pretty weak support. Firstly, there is only one trinity laptop in the list and its a ...
ironmonkey 發表於 2013-10-30 02:10
你無follow開anandtech, 我唔怪你
果個H.264係睇片, 真正壓H.264片既話, 無一部機係去到6個幾鐘的
http://www.anandtech.com/show/58 ... -4600m-a-new-hope/8
Original Review article
Another test that we ran is simulated gaming; we looped the four graphics tests in 3DMark06 at 1366x768 until the battery ran out. We’ve run this same test on quite a few other laptops, and Llano initially looked to be far and away the best solution. Later, we discovered that when we tested Llano we were letting the GPU run in power saving mode—basically half the performance you’d get compared to being plugged in. We retested and measured 98 minutes, so the extra graphics performance comes with a heavy cost. We only tested Trinity (and Ivy Bridge and Sandy Bridge) using higher performance graphics settings, and this is one more area where it scores worse than Llano: Trinity managed just 77 minutes. That’s about the same as Ivy Bridge and Sandy Bridge (79 and 73 minutes, respectively), so if you’re after better gaming performance while running off the mains, you might need to keep looking.
作者: ironmonkey 時間: 2013-10-30 10:40
本帖最後由 ironmonkey 於 2013-10-30 10:47 編輯
你無follow開anandtech, 我唔怪你
果個H.264係睇片, 真正壓H.264片既話, 無一部機係去到6個幾鐘的
Ori ...
qcmadness 發表於 2013-10-30 09:42
I know. Thats why I said "higher load" not "full load".
Also, the passage you quote seems to include the iGPU.
Edit: If I remember correctly, the iGPU sucks very little power 15-30W. It'll be interesting to see if the A10-4600m or 5750m will throttle at CPU full load though.
作者: usei 時間: 2013-10-30 12:30
Ivy 差不多賣完
Haswell i3 GPU 性能更接近 AMD
明年 notebook 轉 boardwell AMD 還有多少 GPU 優勢 ?
作者: Puff 時間: 2013-10-30 14:24
本帖最後由 Puff 於 2013-10-30 14:41 編輯
Ivy 差不多賣完
Haswell i3 GPU 性能更接近 AMD
明年 notebook 轉 boardwell AMD 還有多少 GPU ...
usei 發表於 2013-10-30 12:30
上左 22nm 後,本來就已經無。當初飛甩雞毛係因為起跑線一樣,同埋 focus 既點唔同 (moar CPU vs moar GPU),當然都要多謝 GF 嘅。依家嘅情況,就只係 AMD 食少兩個核心既電係同樣既耗電預算下,搞個大 D 既 graphics core,然後賣平過 Intel。
反正到頭來 Intel 都係食 high premium...。下年頭半年 28nm Kaveri 打 22nm,後半年就高價 Broadwell 空降。後年嘅話,好彩咪 3/4 時間 20nm Excavator vs 14nm Broadwell 囉,不過 14nm 新鮮上陣預佢實賣貴 D 啦。
至於再之後嘅,各大代工廠從 16/14nm 開始o既 hybrid process 可以扳回耗電果一半劣勢嘅。至於 die shrinking... 查實 Intel 係咪次次都係 full node shrink in all aspects 真係唔知。
作者: wilsonkf 時間: 2013-10-30 19:18
提示: 作者被禁止或刪除 內容自動屏蔽
作者: wilsonkf 時間: 2013-10-30 19:23
提示: 作者被禁止或刪除 內容自動屏蔽
作者: hong999 時間: 2013-10-30 23:09
而家intel最勁既核顯去到咩級數?Geforce 9800 GTX/HD4850?定係9600GT/HD 4750?
作者: ironmonkey 時間: 2013-10-31 11:16
You mean using it as intended?
作者: vcfirstben_II 時間: 2013-10-31 16:04
提示: 作者被禁止或刪除 內容自動屏蔽
作者: lcska02nkc 時間: 2013-10-31 16:06
提示: 作者被禁止或刪除 內容自動屏蔽
作者: usei 時間: 2013-10-31 17:14
iris PRO 大約是 4850/640M 級 是 KO 左 現在的 AMD IGP
普通 U 的 HD 4600 630M 級
作者: dom 時間: 2013-11-1 11:52
算吧啦, AMD 唔係實物 都係唔好信佢話多左幾多performance
Faildozer 中伏之後 唔再信佢話加幾多 % 每代了
作者: wilsonkf 時間: 2013-11-1 12:25
提示: 作者被禁止或刪除 內容自動屏蔽
作者: dom 時間: 2013-11-3 11:24
我塊三年前出既 AM3 matx 板( MSI 890GXM) 居然認到 8350...
CPU support 無寫到出黎 ... 好神奇
...
wilsonkf 發表於 1-11-2013 12:25 PM
C4F 由頭到尾(已賣) 都係話 beta support AM3+ CPU (Faildozer 一代)
無話過正式support / 用到
作者: wilsonkf 時間: 2013-11-4 13:04
提示: 作者被禁止或刪除 內容自動屏蔽
作者: dom 時間: 2013-11-4 13:45
我果塊都係 beta support for bulldozer
而而完全無寫用到 piledriver, BIOS release date 都早過 piledr ...
wilsonkf 發表於 4-11-2013 01:04 PM
ASUS 都唔出新版去 AM3+ 啦
你見無出新 ROG 比全部 AMD 平台就知
Intel 就係咁出完大版出細板
作者: wilsonkf 時間: 2013-11-4 15:45
提示: 作者被禁止或刪除 內容自動屏蔽
作者: dom 時間: 2013-11-4 16:52
都無新chip
玩到呢個價位既 都玩 Intel 啦
貪 Faildozer 220W TDP 咩

