Processador AMD Ryzen / Threadripper

Não acredito que a amd passe a intel em IPC, especialmente os novos skylake, mas basta que venham ao nível de um ivy bridge ou algo assim, com 8/16? cores e com consumos contidos que podem vir a ser excelentes opções.

Deixa ver, futurologia é complicado, ainda mais com processadores que só vão sair no próximo ano, de acordo com o roadmap.
 
ObscureAngel disse:
Tem de vir bem melhor que os Phenom a termos de IPC, senão ficamos na mesma situação que os FX.

Por agora só acredito quando ver, criei demasiado Hype com os FX para ver um fail enorme na altura, não quero voltar a ter de ficar desiludido de igual forma.
Clicar para expandir...

Se voltares atrás ao período em que os cpu's Intel perdiam e bem para os cpu's da AMD, a Intel recuperou e começou a ganhar terreno à AMD graças a ter dado um pontapé no que era o Pentium IV e ao regressar à base do Pentium III para a partir daí evoluir para o Pentium M e Core Duo... e o resto é história.


Não é que este 'retrocesso' da AMD seja sinónimo de algo muito bom, mas se o que tinha de mais recente era mau e se reconhecidamente a arquitectura Bulldozer para o mercado dos pc's no geral é má - a ideia até pode ser boa, mas só para certos sectores/tipos de software e é dependente de boa optimização de software [cuja presença/força da AMD nesse mercado é diminuta para conseguir convencer devs a fazer essa optimização]... nada como retroceder, partindo de algo que com melhorias e evolução pode ser bem melhor e até competitivo.


E quer me parecer que algo do tipo Phenom II a 14nm + optimizações, novas instruções e [ao que parece] HT, é suficiente para bater e bem os cpu's actuais da AMD.
 
Última edição:
attachment.php
 
Em complemento ao que escreveu o Roberto:

http://wccftech.com/amd-2016-14nm-cpu-apu-zen-k12-product-roadmap-leaked/

AMD 2016 Roadmap Leaked – Zen And K12 Based 14nm CPU And APU Products Inbound
HardwareLeak 3 seconds ago by Khalid Moammer
AMD’s roadmap for 2016 has been leaked, featuring desktop and mobile products based on the new high performance Zen and K12 cores. As well as APUs that feature AMD’s upcoming “next generation” graphics core next architecture. Likely referring to what we have come to know by now as the Arctic Islands graphics architecture featuring 14nm and the second generation of the HBM stacked memory standard.



Two months ago we published an exclusive article detailing the future technologies that AMD will be discussing in its 2016 roadmap unveiling on May 6th at the Financial Analyst Day event. Including its upcoming Zen, K12 and next generation graphics cores. And as we approach this upcoming event in May, we’ve been bearing witness to a new flood of leaks that have come out to shed light on what’s to come.

Two days ago the CPU block diagram for AMD’s upcoming high performance x86 CPU code named Zen was leaked. It gave us a rough idea of what to expect from Zen compared to AMD’s previous Bulldozer family of cores. And what we saw from the high level layout was quite promising.
Yesterday another leak surfaced, one that detailed how each Zen core would share L3 cache with three other CPU cores arranged in a group of four. This is to form what’s described as a quad core Zen unit.

AMD 2016 Roadmap Leaked – Zen And K12 Based 14nm CPU And APU Products Inbound
However in today’s leak we get to look at how AMD will be its future cores, the “crown jewels” as AMD likes to calls them, to make a myriad of high performance and low power CPUs and APUs. http://cdn3.wccftech.com/wp-content/uploads/2015/04/mobile.jpg
So let’s take a look at the 2015-2016 desktop roadmap.


In 2015 the performance segment will see a continuation of the Vishera FX series which are based on the Piledriver CPU core. In the mainstream segment we’re looking at a Kaveri refresh dubbed Godavari. This refresh represents what Richland was to Trinity. The same piece of silicon design but with an optimized manufacturing process and higher clock speeds. In the low power segment we see that the AM1 socket will also get an update from current Kabini APUs to Beema APUs featuring Puma+ cores and higher clock speeds.

However what’s really exciting is what lies ahead in 2016. As AMD plans to update the entire product range with its brand new Zen CPU core and its next generation Arctic Islands graphics architecture.
In the performance segment we see AMD finally introducing a successor to the Piledriver based Vishera high performance chips.

Code named “Summit Ridge” this product family will feature eight core CPUs with DDR4 support, an entirely new chipset and a new FM3 socket. An eight core Zen based CPU should be considerably faster than a current eight core FX CPU. In fact according to the Zen block diagram, each Zen core should be comparable to a dual core Excavator module in total throughput but significantly faster in single threaded performance. So an 8 core Zen CPU will be more in the league of AMD’s current 16 core Opteron CPUs. But because Zen packs a comparable amount of performance in half the number of cores, the single threaded / per core performance will be significantly up compared to the previous generation.

What’s even more interesting is that these high end Summit Ridge CPU only parts will share the same FM3 socket with their APU siblings. So users can have a seamless continuum of product options on a single platform. This is in contrast to how the market is segregated today between the FM2+ and AM3+ sockets on the AMD side and the LGA 1150 and LGA 2011-3 on the Intel side. Having both APUs and high performance CPUs share the same socket will not only reduce the cost of developing motherboards and thus make them more affordable. It also enables the motherboard makers to be a lot more flexible with their designs. Allowing them to deliver a greater variety of motherboards to end-users to choose from.

This brings us to “Bristol Ridge” which is a family of APUs featuring up to four Zen CPU cores, DDR4 memory support and a next generation GCN graphics architecture. Again on the same FM3 socket. However as Bristol Ridge APUs are SOCs, i.e. they feature an integrated southbridge it is inevitable that we will see AMD introduce different chipset tiers to cater for the enthusiast and mainstream segments. Similar to how today we have the 990FX, 990X and 970 chipsets on the AM3+ socket. And a similar scheme on the FM2+ socket with A88X, A78, A68 and A58.

Also because Bristol Ridge APUs have an integrated southbirdge they can boast area and power savings in the mobile segment. Due to not requiring a dedicated southrbidge chipset at all. It should be noted that Carrizo will be the first product to feature an integrated southbridge in the same die. So Bristol Ridge merely keeps the ball rolling in 2016.
Finally in the low power segment AMD will introduce the “Basilisk” APU which is just a miniature version of Bristol Ridge. Featuring up to 2 Zen cores and a smaller integrated GPU all inside a lower power envelope.


Moving on to the mobility roadmap, we see both Bristol Ridge and Basilisk APUs making a reappearance thanks to their highly integrated SOC design. However we also get a glimpse into AMD’s first ARM consumer SOC. Code named Amur, this APU will launch later this year. Featuring four ARM A57 CPU cores, an integrated GCN GPU and a security processor all inside a frugal 2W SDP.

It’s quite interesting to see that AMD managed to get its GCN graphics architecture inside such a small power envelope. This is perhaps an indication of what we may see on the GPU front from the company in the future. Succeeding Amur will be “Styx” the first consumer product to feature AMD’s custom ARM core code named K12. It’s also interesting to note that all of AMD’s upcoming APUs in 2016 will be fully HSA 1.0 compliant. 2016 is shaping up to be a very busy and exciting year for PC hardware.


Read more: http://wccftech.com/amd-2016-14nm-cpu-apu-zen-k12-product-roadmap-leaked/#ixzz3Yj2aggky
Clicar para expandir...
 
Uma nova arquitetura utilizando a mesma socket? Isso assusta-me um bocado, normalmente quando isso acontece o salto não é tão grande.
 
Sim e o interessante é que o primeiro produto vai ser um FX e só depois saem os APUs. Além disso o socket é o mesmo para os FX e APUs.
 
Não é preciso apanhar a Intel, basta chegar perto para se posicionar como alternativa nos segmentos que não sejam performance de topo.
É questão de jogar com preços e funcionalidades/possibilidades para ser competitiva.

Como está actualmente é que é distância demasiado grande.
Muito também se deve ao actual processo de fabrico, mas com este Zen esta distância vai encurtar seguramente.
 
IPC é só uma parte do puzzle pelo que no global o aumento de performance deverá ser um pouco mais baixo. De qualquer forma deverá resolver um dos pontos onde perde mais com os atuais da Intel.
 
AMD’s 2016-2017 x86 Roadmap: Zen Is In, Skybridge Is Out

AMDZenRoadmap_678x452.jpg


AMD’s CTO Mark Papermaster just left the stage at AMD’s 2015 Financial Analyst Day, and one of the first things he covered was AMD’s CPU technology roadmap for the next couple of years.

The big question on everyone’s mind over the last year has been AMD’s forthcoming x86 Zen CPU, developed by Jim Keller’s group, and Papermaster did not disappoint, opting to address the future of AMD’s x86 plans first and foremost. AMD it not releasing the complete details on Zen until closer to its launch in 2016, but today they are providing some basic details on the CPU’s abilities and their schedule for it.



In terms of features, AMD once again confirmed that they’re aiming for significantly higher performance, on the order of a 40% increase in Instruction Per Clock (IPC) throughput. In a significant shift in threading for AMD’s x86 CPUs, Zen will also shift from Bulldozer’s Clustered Multithreading (CMT) to Simultanious Multithreading (SMT, aka Intel’s Hyperthreading). CMT is the basis for Bulldozer’s unusual combination of multiple integer cores sharing a single FPU within a module, so the move to SMT is a more “traditional” design for improving resource usage, and it means Zen will similarly have a more traditional resource layout. AMD is also labeling Zen’s cache as a “high-bandwidth, low latency cache system,” though at this time they aren’t quantifying just how that differs from the Dozer family’s cache.

Meanwhile AMD has confirmed that Zen will be shipping in 2016, and that it will be produced on a yet-to-be-named FinFET process. Our bet would be that AMD continues to use traditional partner (and spin-off fab) GlobalFoundries, who will be ramping up their 14nm equipment for next year as part of their licensing/partnership with Samsung to implement Samsung’s 14nm FinFET process. Zen at this time is AMD’s priority, to the point where the company is willing to push back the ARM K12 in order to get Zen out the door first.

Desktop users will be happy to know that the first Zen processor out the door will be AMD’s high-end desktop CPU (AMD was very deliberate in this, it’s not an APU). AMD will be aiming high and then cascading Zen down into APUs and lower-end products.

Said Zen CPU will use a new AMD platform – AM4 – which will also support DDR4. Unlike the Dozer family, all of AMD’s desktop CPUs will use the same AM4 platform/socket, so when AMD does ramp up their Zen APU, it too will be on AM4 and not its own socket, simplifying the process.



Finally, AMD’s roadmap for Zen over 2016-2017 calls for further improved Zen cores, “Zen+”, later on. Though the biggest jump for AMD comes from the transition from Dozer to Zen, the company is looking to push the envelope on IPC harder than they did with the Dozer family, projecting a higher increase in IPC over time than what we saw with the Dozer parts.



As for what this means for single-threaded performance, that remains to be seen. After having chased a lower-IPC/higher-clockspeed strategy in an attempt to do something different from Intel – the AMD that designed Bulldozer was convinced they couldn’t beat Intel at their own game – the focus on higher IPC is arguably the correct move to make as the laws of physics have continued to keep 5GHz+ clockspeeds from being efficient enough to be practical. That said, single-threaded performance is a combination of IPC and clockspeed, so to be competitive with Intel AMD needs to deliver higher IPC while maintaining relatively high clockspeeds, and the latter is not something AMD is disclosing this early in the process.

Finally, we’ll note that this roadmap is empty of any mentions of project Skybridge. CEO Lisa Su has commented that AMD has decided to change their focus away from Skybridge based on customer feedback. Customers were telling AMD that they didn't necessarily need socket compatible solutions, so AMD is going to work on differentiated solutions. That said, given that Skybridge was announced last year and planned for 20nm, I suspect that it has also become a victim of the short lifespan (and underperforming nature) of 20nm, leading to it being downplayed in favor of 14nm prodcts.

Meanwhile also absent was any further mention of AMD's Cat cores. Zen would appear to be AMD’s top-to-bottom x86 core, a significant departure from an AMD who previously wanted separate designs for the 10W-100W market and sub-10W markets.

Clicar para expandir...

Aumento de 40% de IPC comparado com cores Excavator (suponho que sejam os do Carrizo?). Quando esses saírem será fácil fazer umas continhas para comparar com os Intel e ver se chega para serem competitivos ou não.
 
A termos de IPC diria que ainda vai ficar longe dos intel, devem ter que combater isso com clocks altos.
Só que chegue ao nível dos i5 fico feliz.
 
Boas!

Vamos deixá-los "pousar", mas gráficos deste género, a mais de um ano do lançamento devem, na minha opinião, ser lidos com uma boa pitada de sal.

Não me acredito em milagres, mas gostava que a AMD, pelo menos, se tornasse mais competitiva no segmento das CPUs.

Cumprimentos.
 
Na minha opinião a AMD não precisa de bater a Intel. Precisa apenas que sejam decentes.

Para recordar produtos mais recentes, dou o exemplo do Phenon II X6. O primeiro Phenom foi uma desilusão, depois houve o TLB bug. O Phenom II já era mais interessante, mas o X6 a meu ver foi um dos melhor processadores da AMD. Tive 2 e com um ssd, para a altura e para quem precisa mais de cores que outra coisa, eram e ainda continuam a ser uma maravilha.
Grande parte do sucesso vinha do preço, de desbloquear cores, plataforma barata, etc.

Com o Zen eles devem apanhar a transição da Intel entre o Skylake e o Cannonlake. Ainda não sabemos sequer o que vale o Skylake, e com a pouca informação do Zen é dificil dizer se terá sucesso.
Mas com um core decente e preços razoáveis, pode ser um sucesso.

Quanto ao IPC não chegar aos niveis de um Intel, como já disseram IPC não é tudo, mas também pelo que dá a perceber, o Zen não me parece que seja um "animal com muito clock para compensar o IPC". Não me parece que seja por aí que a AMD quer ir.

Agora é esperar.
 
ObscureAngel disse:
A termos de IPC diria que ainda vai ficar longe dos intel, devem ter que combater isso com clocks altos.
Só que chegue ao nível dos i5 fico feliz.
Clicar para expandir...

No Cinebench 11.5 single thread, o Kaveri A10-7850K a 3.7Ghz/4Ghz atinge um valor de 1.02. Baixando o TDP (ex. benchmakrs do A8-7600) estes valores não descem muito.
Digamos que o Zen terá performance single-thread de 1.4-1.5 fiando-nos nesse resultado (mais 40% de performance que os Excavator, não que os Steamroller mas como não temos benchmarks de cores Excavator...). Isso será o equivalente á performance de um Intel Core i5 Ivy Bridge ou Haswell em IPC se usarem essas frequências o que não é mau.

Depois se tiver essa performance e for mais eficiente que os Carrizo, tendo em conta os ganhos de eficiência energética que são sugeridos que vão acontecer com os Carrizo pode-se ter uma boa combinação.

Depois resta ver se conseguirão atingir frequências altas no processo de fabrico de 14nm Finfet como não conseguiram no de 28nm com os Kaveri porque tiveram que baixar a frequência e isso anulou os ganhos de frequência comparando com a anterior geração.
Pelo que li, a GF actualmente está com um yield de 30% no processo de fabrico de 14nm e para o ano deverá atingir a eficiência necessária para fornecer a AMD. Resta ver se terá bom yield e maturidade suficiente para atingir frequências elevadas.

Resumindo, com o Zen a AMD vai ter finalmente CPUs competitivos com pelo menos a mid-range da Intel se olharmos para os Haswell e tudo irá ficar dependente da maturidade do processo de fabrico, se vão conseguir atingir frequências elevadas com os CPUs e da eficiência energética para atingir essas frequências.

E definitivamente será um salto de gigante face aos actuais CPUs/APUs da AMD. Claro que para apanhar a Intel talvez precisassem de dois saltos de gigante...
 
Última edição:
Inicie sessão ou registe-se para poder participar.
Back
Topo