Processador AMD ZEN4 (Ryzen 7000 series)

Sim, acho que é as 15H (Hora Europa Central) cá deve ser ás 14H

Mas a Puget é uma empresa que vende workstations, por isso a review deles não devia de estar abrangida pelo NDA, a malta quer é saber dos FPS

EDIT: entretanto parece que a AMD já divulgou o iGPU no IOdie dos Ryzen 7000

AMD-RYZEN-7000-RDNA2-768x432.jpg

AMD-RYZEN-7000-RDNA2-2-768x432.jpg
 
Sim, acho que é as 15H (Hora Europa Central) cá deve ser ás 14H

Mas a Puget é uma empresa que vende workstations, por isso a review deles não devia de estar abrangida pelo NDA, a malta quer é saber dos FPS

EDIT: entretanto parece que a AMD já divulgou o iGPU no IOdie dos Ryzen 7000

AMD-RYZEN-7000-RDNA2-768x432.jpg

AMD-RYZEN-7000-RDNA2-2-768x432.jpg

Até este iGPUs suportam Display port 2.0 e as novas RTX 4000 não. Lol
 
Mas mesmo os monitores até alguns já anunciados para fins de 2022 inícios de 2023 é tudo 1.4.
Mas sim, faz sentido já ter 2.0.



Quanto à discussão anterior, estou curioso pelos 3D destes Zen4, mas o preço deve ser upa upa.
 
Para gaming, os 7600x e 7700x têm consumos razoáveis. Talvez dê para fazer um tweaking engraçado a nível de power limit e PBO undervolt mas até agora este lançamento é um bocado meh.
 
É assim a vida, mas está dificil o pessoal capacitar-se que todas as marcas, vão ter este tipo de consumos e temps, seja cpu ou gráficas...

"This round, however, it feels like AMD did an Intel. Where energy efficiency and heat levels mattered up to Ryzen series 5000/6000, that picture is painted black with Ryzen 7000 when it comes to Wattage and thermals"

A sério, estas conclusões é só rir. Quem quiser fazer melhor, entre na faculdade, vá estudar engenharia e depois falamos ;)
 
Última edição:
A meu ver só quando saírem as boards B com preços minimamente razoáveis (e não os preços das X) e as versões com 3D-vcache é que vai valer a pena dar upgrade para quem estava à espera. Isto sem ter em conta o que a intel vai trazer.
 
Já vi umas quantas de reviews e ele há coisas que, enfim...


Test Bed and Setup​


As per our processor testing policy, we take a premium category motherboard suitable for the socket, and equip the system with a suitable amount of memory running at the manufacturer's maximum supported frequency. This is also typically run at JEDEC subtimings where possible. It is noted that some users are not keen on this policy, stating that sometimes the maximum supported frequency is quite low, or faster memory is available at a similar price, or that the JEDEC speeds can be prohibitive for performance.

While these comments make sense, ultimately very few users apply memory profiles (either XMP or other) as they require interaction with the BIOS, and most users will fall back on JEDEC-supported speeds - this includes home users as well as industry who might want to shave off a cent or two from the cost or stay within the margins set by the manufacturer. Where possible, we will extend out testing to include faster memory modules either at the same time as the review or a later date.
Screenshot-2022-09-26-at-14-44-12-AMD-Zen-4-Ryzen-9-7950-X-and-Ryzen-5-7600-X-Review-Retaking-The-High.png

https://www.anandtech.com/show/1758...-ryzen-5-7600x-review-retaking-the-high-end/9


O Andreas Schilling testou os 7950X e 7700X com DDR5 a 5200 e 6000, e apresentou os resultados dos testes com essa referência e ainda teve tempo de testar ambos os cpu com limite de voltagem

memory benchmarks​


First, let's look at a few synthetic memory benchmarks. We carried out this for the Ryzen 7000 processors without a corresponding marking with DDR5-5200 - for the Core i9-12900K with DDR5-4800. Then we each switched to DDR5-6000, which should be the interesting area for us.
Screenshot-2022-09-26-at-14-54-36-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png

As expected, increasing the memory take and increasing the transfer rate to 6,000 MT/s has a positive effect on the throughput rates - that's to be expected. The Ryzen 7 7700X struggles with the fact that its read throughput is somewhat lower due to the connection of the CCD to the IOD. But it becomes clear here that DDR5-5200 is not the desired memory to be used with the new processors.

And now some games:
Screenshot-2022-09-26-at-15-00-49-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png
Screenshot-2022-09-26-at-15-00-58-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png

Both the Core i9-12900K and the two Ryzen models benefit in some cases extremely from the faster memory. Of course, this also depends a bit on the game and the selected settings, but an FPS plus of 10 to 15% is possible if you are not yet completely at the GPU limit.
https://www-hardwareluxx-de.transla...sl=de&_x_tr_tl=en&_x_tr_hl=pt-PT&_x_tr_pto=sc

We use the following test system configurations to test the processors:


AMD AM5 (Ryzen 7000):


  • Gigabyte X670 AORUS Elite AX
  • G.Skill Trident Z5 Neo 2x 16GB DDR5-6000 30-38-38-96
  • be quiet! Dark Rock Pro 4
  • Corsair HX1000 power supply
  • GeForce RTX 3090 Ti

Intel LGA1700 (Alder Lake) for DDR5:


  • ASUS ROG Maximus Z690 Hero
  • Kingston Fury Beast 2x 16GB DDR5-5200 40-39-39-76
  • be quiet! Dark Rock Pro 4
  • Corsair HX1000 power supply
  • GeForce RTX 3090 Ti

Second system for DDR4:


  • ASUS TUF Z690-Plus WIFI D4
  • Corsair Vengeance 4x 8GB DDR4-3600 18-19-19-39
  • be quiet! Dark Rock Pro 4
  • Corsair HX1000 power supply
  • GeForce RTX 3090 Ti
Screenshot-2022-09-26-at-15-06-20-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png

https://www-hardwareluxx-de.transla...sl=de&_x_tr_tl=en&_x_tr_hl=pt-PT&_x_tr_pto=sc


e agora a parte do powerlimit, fez os testes do 7950X com os 170W, 115w e 65w

Screenshot-2022-09-26-at-15-10-16-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png
Screenshot-2022-09-26-at-15-11-04-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png
Screenshot-2022-09-26-at-15-15-59-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png

https://www-hardwareluxx-de.transla...sl=de&_x_tr_tl=en&_x_tr_hl=pt-PT&_x_tr_pto=sc
não se preocupem também fez os testes em jogos :berlusca:


Page 22: iGPU benchmarks​

Even if the integrated RDNA 2 GPU should only be used by gamers in exceptional cases, we ran a few benchmarks with it. But first, let's take a look at the GPU-Z screenshot.

amd-ryzen-7950x-gpuz_1920px.png

Two compute units and therefore 128 shader units do not uproot any trees. The GPU clock is 2,200 MHz. The main memory is used as graphics memory and therefore there is a certain dependency on its speed.
Screenshot-2022-09-26-at-15-20-59-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png
Screenshot-2022-09-26-at-15-21-06-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png

Screenshot-2022-09-26-at-15-21-14-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png
Screenshot-2022-09-26-at-15-24-11-Zen-4-mit-DDR5-Ryzen-9-7950-X-und-Ryzen-7-7700-X-im-Test-Hardwarel.png

What the technical data have already indicated: The integrated graphics unit of the Ryzen 7000 processors is sufficient for a rudimentary display of 2D and 3D content. Games only achieve adequate FPS at 1080p and below and at low details, if at all. Increasing the memory speed from DDR5-5200 to DDR5-6000 results in a performance increase of about 7%.

The RDNA 2 GPU supports AV1 decoding in 10 and 8 bits. In addition, decoding in VP9, H.265 and H.264 (all also in 10 and 8 bit) is supported. However, encoding is only possible in H.265 and H.264. At least in 1080p and 1440p we were able to decode an AV1 video on YouTube in the Chrome browser, as shown in the following screenshot:
ryzen-7950x-av1-1_1920px.png

But as soon as we switched to a 4K or even 8K video, this was no longer possible or the video jerked, frames were dropped and you certainly don't want to watch a video in this form. Due to the low time budget for this test, we were unable to find out whether the driver was to blame or whether the allocation of the graphics memory was not sufficient. We will look at the topic again.
https://www-hardwareluxx-de.transla...sl=de&_x_tr_tl=en&_x_tr_hl=pt-PT&_x_tr_pto=sc
 
Tal como já disse referente à Nvidia, aplica-se o mesmo aqui.

Precisam mostrar números, e um increase de performance que de algum modo tente convencer malta a fazer o upgrade.

Como tal, há que puxar consumos para conseguir mostrar um salto de performance. Conseguir um jump de performance mantendo o consumo para mim seria uma boa evolução.

Ao subir os consumos sinto como sendo uma "evolução falsa". Tem mais performance, sim. Mas se a ideia for chegarmos ao dia de cpus com tdp de 250W só para conseguirem mostrar números, estão claramente com as prioridades todas trocadas.
 
Temps e consumos gigantes pros top de gama, pro 7600X nada de especial, vi 100w com PBO ligado, tenho 110w no 5600X com PBO ligado e motherboards limits e uso a 90w pro dia-a-dia porque o ganho para 110w é miserável.

Não me compensa, um 5800X 3D ganha-se gaming performance com menos gasto.

Agora o 7950X indo a 250w é um absurdo, do que adianta ter 50% de ganho se consome 70% a mais? o 5950X consome 120-150w...

a sério, não percebo. A AMD do 3950X pro 5950X teve um ganho de performance expressivo, com o MESMO CONSUMO, o mesmo socket e processo de fabrico!!! Agora temos socket novo, memória mais rápida, 5nm e vem isso?

Vamos ver como será com UV e curve optimizer. Acho que um 7950X limitado ai a 5.0 - 5.2 ghz devidamente optimizado deve voltar aos 150w ai e com performance jeitosa.

Alguma review testou isso a sério? a do techpower up curiosamente tem testes com -0,05v, mas não mostraram consumo e temps...
 
Última edição:
Back
Topo