Processador AMD ZEN2 7nm (ler primeiro post)

E boards "baratas" e acessíveis em formato mini-itx? :D

Versões embedded serão mais usadas no mercado empresarial e para aplicações mais "especificas".

Seja como for, a Asrock apresentou uma board com as dimensões do NUC e pode ser que apareça em algum MiniPC deles, no mercado consumidor. :)

LUZg3Ew.jpg


nMK6sId.jpg


nbIDh6D.jpg


http://linuxgizmos.com/ryzen-v2000-shows-up-on-sbcs-mini-pcs-and-signage-player/
 
Eram ofertas relativamente comuns da AMD, até dos "normais" OEM, como Asrock, Gigabyte, etc; neste momento a oferta é só da Asrock IPC, Gigabyte IPC e por aí fora, mas isso também as Kontron, Congatec, Advantech e por aí fora têm...

Lá está, essas 4x4V2000 são Asrock Industrial.
 
Lá está, essas 4x4V2000 são Asrock Industrial.

Yep, a asrock já tem esta serie industrial à algum tempo, mas existem sites que vendem ao consumidor, os preços é que sao um pouco puxados e atenção que isto é alimentação 24v. Acho que existe suporte para trabalhar em 12v mas não sei que limitações implica.
 
Material para o mercado industrial será sempre mais caro, mas normalmente, se derem importância a qualidade de construção, estão bastante acima do que se vende no mercado comercial.

Dito isto, supostamente a Asrock vende estas Motherboards em separado e têm versões da mesma motherboard, com os 4800U, 4500U e 4300U, em vez das versões embedded dos "Renoir".
O nome que lhes dão é 4X4, mas é por provavelmente não puderem usar o nome NUC. As dimensões são as mesmas. Há caixas à venda para as dimensões das boards dos NUC.
 
GIGABYTE Releases GPU-Centric Servers for 8+ Accelerators and 160 PCIe Lanes

Improved Accelerator Performance:

All three new servers can run all their PCIe Gen 4 lanes to function at x16, which means there can be as many as 160 PCIe lanes available, and they all have a theoretical bandwidth of 32 GB/s for fast communication between CPU and GPU. This is made possible by AMD EPYC 7002 architecture that allows a single socket-to-socket Infinity Fabric link (one of four) to reallocate the bandwidth to PCIe lanes for other purposes. This is possible because a dual AMD EPYC 7002 system uses 64 PCIe lanes from each CPU to link to the adjacent CPU. By freeing up one of these four links, 16 lanes per CPU can be redirected for fast PCIe Gen 4 speeds in storage, accelerators, or networking.
G482-series-768x340.png

Incredibly, the G482-Z53 and G482-Z54 are built for 8 x GPUs, such as the AMD Instinct MI50, slotted next to each other and linked via the AMD Infinity Fabric link. These servers have been tested and designed for AMD Instinct MI50 and other accelerators. The servers utilize dual AMD EPYC 7002 processors operating with 128+ PCIe Gen 4 lanes. Recognizing the full potential of high bandwidth PCIe Gen 4 lanes, the 2nd Gen AMD EPYC processors are capable of reallocating bandwidth dedicated to CPU to CPU connectivity to be redirected for accelerators and networking. What is even more impressive is that all GPUs have their own dedicated PCIe Gen 4 x16 lanes, without the use of a PCIe switch to share a single lane with two GPUs.
https://www.hpcwire.com/off-the-wir...ervers-for-8-accelerators-and-160-pcie-lanes/
 
Top500 November 2020 Our New Systems Analysis

SC20-Top500-November-2020-New-Systems-by-CPU-Architecture.jpg

This is extremely interesting. On this list, we see a “resurgence” of 2016-era CPUs with a single Intel Xeon E5 V4 “Broadwell” system. Something that is very interesting here is that we also have 6 of Intel’s 29 systems using first-generation Intel Xeon Scalable processors codenamed Skylake. That means a total of seven “new” systems are using older generations of processors, and Cascade Lake is not feeling modern at this point. In contrast, AMD did not have a new EPYC 7001 system but did have 12 AMD EPYC 7002 systems.

When we remove pre-2018 x86 “new” systems as we did in previous versions of this analysis, we get a different picture:

SC20-Top500-November-2020-New-Systems-by-CPU-Vendor-Excluding-pre-2018-x86.jpg

Intel still has a commanding share here, but if we look at current architectures that share is slightly more muted. In the June 2020 list Intel has around 76% share with the older x86 CPUs removed. In the November 2020 list, Intel is down to 59% in this metric. AMD has moved up to 27% share overall but 32% share when removing legacy chips. As we noted in previous analysis pieces, we will track in the future if this delta continues to increase if customers continue to use older generations of non-Intel CPUs.

SC20-Top500-November-2020-New-Systems-by-CPU-SKU.jpg

We do see a single Intel Xeon Platinum 9242 system, but not the higher-end Intel Xeon Platinum 9282 56-core part. Interestingly the nine 64-core parts were all AMD EPYC 7002 series and none were the AMD EPYC 7H12 we reviewed.
https://www.servethehome.com/top500-november-2020-our-new-systems-analysis/

- Top500: Fugaku Keeps Crown, Nvidia’s Selene Climbs to #5
Top500-top12-Nov2020_560x.png

https://www.hpcwire.com/2020/11/16/top500-fugaku-keeps-crown-nvidias-selene-moves-up-to-5/

A AMD tem 2 sistemas no TOP 10... por causa das Nvidia A100 :berlusca:
 
Gigabyte Brix com os Renoir

AMD Renoir enters GIGABYTE BRIX S
Gigabyte unveiled its new Mini-PCs featuring AMD Ryzen 4000U series processors. The APUs feature AMD Zen2 core architecture and AMD Vega graphics. In total there are eight BRIX S models announced today. The Mini-PC is available in two form factors: 34.7 mm tall and 46.8 mm tall. The taller version features support for a 2.5-slot SSD drive and offers an optional daughter board with 2 COM and 1 LAN port.

Both models offer a single M.2 2280 slot for PCIe SATA drive, offer 2.5G Realtek 8125 Ethernet port, and have dual DDR4-3200 MHz memory support.
Gigabyte-BRIX-S-Tall-Ryzen-4000U-6-1200x764.jpg
Gigabyte-BRIX-S-Thin-Ryzen-4000U-2-1200x764.jpg

The BRIX S will be offered with four Ryzen 4000U CPUs: Ryzen 7 4800U, 4700U, Ryzen 5 4500U and Ryzen 3 4300U:
https://videocardz.com/newz/gigabyte-unveils-brix-s-with-amd-ryzen-4000u-zen2-renoir-processors

E já agora... score Geekbench do 5500u, suposto Lucienne e refresh do Renoir (Zen2+Vega7),
AMD Ryzen 5 5500U with Radeon Graphics 1 Processor, 6 Cores, 12 Threads
AuthenticAMD Family 23 Model 104 Stepping 1
Base Frequency 2.10 GHz Maximum Frequency 4.04 GHz
https://twitter.com/TUM_APISAK/status/1329394813725069315
TIMI-Laptop-Geekbench-Browser.png

https://browser.geekbench.com/v5/cpu/4840713
 
Mais uma adição ao Coreboot: Renoir e Lucienne

[Coreboot] include/device/pci_ids: add PCI IDs for new AMD SoCs https://github.com/coreboot/coreboot/commit/07c85ad11c763a3179ca88e8c0e6d5ae4168c09f
FAM17H_MODEL60H_GPU = 0x1636 = RN iGFX DeviceID.
FAM17H_MODEL68H_GPU = 0x164C = LCN iGFX DeviceID.
https://twitter.com/KOMACHI_ENSAKA/status/1330282355852972032

EDIT: pois parece que vêm aí mais Cromobooks

AMDGPU Linux Driver Seeing A Lot Of Power Saving Optimization Work
With the many different DC display core patch series this month and other patches floating around, there has been seemingly a lot of work on optimizing the graphics power usage. And in particular a lot of work on the optimizations from the DCN3 (Display Core Next 3) perspective with Van Gogh in particular being a focus.
Seeing all of this power optimization work for new and forthcoming hardware is exciting. It wouldn't surprise me the least if this fresh round of recent power optimizations on Linux were at least pushed along in part by more AMD-powered Chromebooks on the way. Similar to HDCP support and other features worked on for the AMDGPU Linux driver as a result of Chromebook AMD APU wins, it's quite possible gearing up for better power efficiency is also part of this effort.
https://www.phoronix.com/scan.php?page=news_item&px=AMDGPU-Nov-2020-Power-Optimize

Não deixa de ser estranho que com o Coreboot tenha adicionado já o Cezanne,
FAM19H_MODEL51H_GPU = 0x1638 = CZN iGFX DeviceID.
mas tenha deixado de fora o Van Gogh, que é o único cujo iGPU não é Vega mas a novel Navi2

cezanne-637x500.jpg
 
Última edição:
Sim, o padrão parece-me ser o mesmo que conduziu à apresentação dos actuais, não apenas o Pixelbook, também a Lenovo entre outros apresentaram as suas versões, mas a adição destes indica que provavelmente daqui a uns meses veremos uns baseados nestes SoC Zen2.
 
Azure Scaled to Record 86,400 Cores for Molecular Dynamics
A new record for MPI-based HPC scaling on the public cloud has been achieved on Microsoft Azure. Led by Dr. Jer-Ming Chia, the cloud provider partnered with the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign (UIUC) to test the limits of Azure’s scalability on a key molecular dynamics tool. The result: successful scaling across a staggering 86,400 CPU cores — a virtual cluster with nearly 5 peak petaflops of performance.
The code at hand was Nanoscale Molecular Dynamics (NAMD), developed in-house at UIUC. NAMD, nicknamed “the computational microscope,” has played a major role in COVID-19 research this year and was at the heart of the massive simulations of the coronavirus that yesterday won the 2020 ACM Gordon Bell Special Prize in HPC-Based COVID-19 Research. Azure in fact hosted similar simulations of SARS-CoV-2’s viral envelope and spike proteins as part of a project funded by the COVID-19 HPC Consortium.
The efforts focused on scaling Azure’s HBv2 virtual machines, each of which was equipped with 120 AMD Epyc 7002-series CPU cores, 480GB of memory and – crucially – 200 Gbps Nvidia HDR InfiniBand.
https://www.hpcwire.com/2020/11/20/azure-scaled-to-record-86400-cores-for-molecular-dynamics/
 
Ya, o cooling parece ser interessante, com heatpipes jeitosos. uma vantagem que vejo que tem mais "clearance" para o airflow

Também gosto do "pacotão" de RAM :D

A parte da frente ta toda vazia, e assim mesmo?

As placas atras que fiquei curioso.

Que grafica é aquela? portas DVI?
 
Também gosto do "pacotão" de RAM :D

Pelas specs das instâncias, são 512 GB de RAM ao todo. Muitos DIMMs, mas a quantidade de RAM não é assim tão grande. Penso que o Epyc pode levar até 4 TB de RAM por socket.

A parte da frente ta toda vazia, e assim mesmo?

Não sei qual é a parte da frente. Há servidores onde o IO está na parte frontal.
Aquela parte vazia parece-me que é para ter espaço para as fontes (Devem ser 2 em redundância) e ao mesmo tempo deixar aquele espaço vazio, para o ar chegar o melhor possível ao dissipador dos heatpipes.

As placas atras que fiquei curioso.

Canto inferior direito parecem-me ser 4 M.2 com dissipadores. No meio, penso que é uma placa Infiniband 200 Gbits, para ligar um transceiver com conector QSFP56:
du2FrcB.png


Que grafica é aquela? portas DVI?

O autocolante parece-me ser da Gigabyte(?). Não faço ideia o que é que uma placa gráfica está a fazer naquele servidor. Não é nada comum, porque bastava o servidor ter uma BMC integrada, para fazer a função de uma gráfica e muito mais. Ainda mais bizarro é ter cooling activo.
 
Back
Topo