***ROZ3 - Republic Of Zen #EP-3 : Society for All AMD Citizens***

[Parona]

ลุงหมูลงไปโลก Crypto DeFi หมดแล้วมั้ง โคขวิดน่าจะไม่ค่อยได้กินแกหรอกนะถ้าไม่บ้าไปเชื่อพวกรอวัคซีน

ยกเว้นโดนวัคซีนเล่นซะเอง

[PatRaceTin]

ลุงสบายดี นะ ตกลงอัพ่ค่ายไหนดีครับ 55

ดีขึ้นครับแต่ยังไม่คืนสภาพปกติ ยังไม่อัพครับ รอดู ศึก cpu amd + ddr5 ของฝั่งแดงก่อน

[ssk]

Microsoft ได้ของจริงไปลองแระ บอกว่ายัด V-cache มาเฉยๆก็แรงขึ้นเยอะแล้ว

Microsoft Posts In-Depth AMD EPYC Milan-X Benchmarks


https://www.tomshardware.com/news/mi...n-x-benchmarks

Microsoft Posts In-Depth AMD EPYC Milan-X Benchmarks


By Zhiye Liu about 10 hours ago



AMD 3D V-Cache (Image credit: AMD)

Milan-X features up to 768MB of L3 cache (L3 + 3D V-Cache) per chip, so a dual-socket configuration delivers up to 1.5GB of L3 cache per system, or in Microsoft's case, per VM. Logically, the L3 allocation will depend on the setup. For example, the 16-core VM has access to 96MB per core, whereas the 32-core setup drops to 48MB per core. At any rate, Milan-X's L3 capacity represents a 3x upgrade over current Milan chips, or a 6x improvement over the previous Rome processors.

Microsoft noted that a large cache obviously boosts effective memory bandwidth and latency. Workloads, such as computational fluid dynamics (CFD), explicit finite element analysis (FEA), weather simulation, and EDA RTL simulation will benefit from Milan-X's generous helping of L3 cache. On the contrary, workloads that are dependant on peak FLOPS, clock speeds, or memory capacity are immune to large L3 caches. These include molecular dynamics, EDA full-chip design, EDA parasitic extraction, and implicit finite element



Microsoft noted that a large cache obviously boosts effective memory bandwidth and latency. Workloads, such as computational fluid dynamics (CFD), explicit finite element analysis (FEA), weather simulation, and EDA RTL simulation will benefit from Milan-X's generous helping of L3 cache. On the contrary, workloads that are dependant on peak FLOPS, clock speeds, or memory capacity are immune to large L3 caches. These include molecular dynamics, EDA full-chip design, EDA parasitic extraction, and implicit finite element analysis.




Microsoft put the EPYC 7V73X through its paces and compared the Milan-X chip to the company's Azure HBv3 VMs with EPYC Milan, EPYC Rome and Xeon Platinum (Skylake) processors. Needless to say, Milan-X's performance is nothing short of amazing.

At the 64 VM configuration, Milan-X delivered up to 77% higher performance than Milan and was up to 257% faster than Skylake with the f1_racecar_140 model on Ansys Fluent 2021 R1. With the combustor_830m model, Milan-X posted 16% and 131% higher performance numbers than Milan and Skylake, respectively, with the 128 VM arrangement.

Under the OpenFOAM Motorbike benchmark, Milan-X was up to 60% faster than Milan and 305% faster than Skylake at the 8 VM setup. The trend was clear as Milan-X boasted double-digit performance improvements over its predecessor and triple-digit enhancements over Skylake.

[ssk]

ZEN-5000 จะขาดตลาด จะแพงก็ไม่แปลกล่ะ ป้าแกเล่นทำส่ง SERVER-OEM เป็นหลัก
พอดังก็ไม่เห็นหัวสาวกโลย..................................

AMD Stock Surges as EPYC CPUs Land in Facebook, Microsoft Servers

https://www.tomshardware.com/news/am...osoft-use-epyc


By Aaron Klotz about 18 hours ago

A good rebound from Alder Lake Comments (0)(Image credit: AMD)
AMD stock rose by 11.2% today as Facebook announced the use of AMD-based EPYC processors in its data centers. With the inclusion of Facebook, AMD now supplies numerous top companies with large data centers, giving it additional ammunition to fight Intel for market share.

On top of this, Microsoft will use EYPC processors in its Azure cloud service, with AMD making a custom SKU specifically for Microsoft with extra memory.

All of this combined has allowed AMD's stock to rise by over 75% in the past six months and we suspect it'll continue to climb if AMD can gain more market share over Intel in the server space.

The CPU manufacturer just announced its new Milan-X EPYC CPUs that will be equipped with 765MB of L3 cache, thanks to the introduction of V-Cache. According to AMD, the performance of the new Milan-X is expected to be around 66% faster than the current Milan chips in server biased workloads such as Synopsys VCS, This is a huge leap in CPU performance for any company that can take advantage of all that L3 cache.

AMD also released an official Zen 4 roadmap update today, with the announcement of a new 96-core EPYC codenamed "Genoa" that will be coming next year. Genoa will run on the brand new Zen 4 architecture that will utilize TSMC's new 5nm process node, and integrate new standards such as DDR5 and PCIe Gen 5.0.

For 2023, AMD is upping the bar even further with a 128-core EPYC chip codenamed "Bergamo". These chips will be designed for high-performance cloud computing running on what AMD calls "Zen 4 C" cores.

AMD is poised to grab even more market share in the server, enterprise, and cloud sectors over the upcoming few years. It'll be exciting to see how many more companies are willing to select AMD over Intel for their needs.

[micronz]

ZEN-5000 จะขาดตลาด จะแพงก็ไม่แปลกล่ะ ป้าแกเล่นทำส่ง SERVER-OEM เป็นหลัก
พอดังก็ไม่เห็นหัวสาวกโลย..................................

สำนักไทยบางจ้าวก็บอกว่าไม่ลดราคาของมันกำลังขาดอยู่เลย ลดทำไมวะ ถ้าขายไม่ออกนี่สิ

[micronz]

12 VS 12 เหมาะเล่นเกมส์ กว่าใหม่ ช่วยกันดูหน่อยไม้ต่างกันเยอะไม้ คลิปมาเรื่อยๆๆ นะ

ตัวรอง 12 อินเทลตบเป็น I7 เลยเหรอเนี่ย

[micronz]

win 11 มันบัค ทั้งคู่สิเนี่ย เทสตืใหม่เลยอ่ะ

[micronz]

[Parona]

https://www.overclockzone.com/view-6...Mab92TbJpG97TE

AMD ขายดีอีกละ

แต่ขายดีแบบนี้ไม่เอานะครับ เดี๋ยวเหมือน Chia

[micronz]

3 D วีแคช จบเลยครับ 555

[Parona]

3 D วีแคช จบเลยครับ 555

คือแบบตะโกนว่าม่ายยยย งานกำลังเข้า ที่ใช้นี่ก็ซดแรมเท่าไหร่ก็ไม่พอ น่าจะขึ้นไปกิน L3 กับเขาด้วยมั้งหนะ

[Sumomo]

แช้งให้rtmเจ๊งดีกว่า

[micronz]

แช้งให้rtmเจ๊งดีกว่า

RTX หรือ RX รึเปล่าท่าน 55
ลำคาญใจนะพวกเหมืองเดี่ยวมาแย่งโน่นนี่ เดะเการ์ดจอ เดะเอาฮาร์ดดิส ล่าสุดจะมาเอาไรเซ่น
เดิมๆๆของก็แพงอยุ่แล้ว

[Sukhoi47]

RTX หรือ RX รึเปล่าท่าน 55
ลำคาญใจนะพวกเหมืองเดี่ยวมาแย่งโน่นนี่ เดะเการ์ดจอ เดะเอาฮาร์ดดิส ล่าสุดจะมาเอาไรเซ่น
เดิมๆๆของก็แพงอยุ่แล้ว

เดียวมีเหรียญที่ใช้ไฟ RGB ขุดอีกนี่จบเลยนะ คอขวดกันกระจายแน่ๆ

[ssk]

RTX หรือ RX รึเปล่าท่าน 55
ลำคาญใจนะพวกเหมืองเดี่ยวมาแย่งโน่นนี่ เดะเการ์ดจอ เดะเอาฮาร์ดดิส ล่าสุดจะมาเอาไรเซ่น
เดิมๆๆของก็แพงอยุ่แล้ว

มู๊บออน...............................................
ไป SAMSUNG

Samsung Exynos 2200’s AMD GPU exhibits 34% peak performance improvement over predecessor

By
Zohaib Ahmed
-
Nov 11, 2021



ReddIt
Samsung Exynos 2200 official GPU benchmark results are here and they show some significant performance improvements over its predecessor. The peak performance is something to put one’s hands together for while sustained performance is more of a mixed bag.

Notable leakster Tron has posted some figures, reportedly given by none other than Samsung itself. These show that AMD’s RDNA 2-powered 6-core GPU on the Exynos 2200 has a peak performance improvement of +31 to 34% over the Exynos 2100, which is quite a bit considering we typically see +20 to 25% performance gains between generations of CPU or GPU.

EXCLUSIVE (Official Benchmark)

Compared to E2100, E2200 GPU is…
Sustained +17~20%
Peak +31~34%

Compared to SD888, E2200 GPU has…
Big difference in 3DMark performance (Wild Life)

ARMv9

* Performance results based on pre-release hardware and software
* Subject to change pic.twitter.com/m6BKqWcgKj

However, the sustained performance, or the performance metrics over a given time duration, is less satisfactory with a difference of +17% to +20%. This could indicate throttling, but Tron says that the performance results are based on pre-release hardware and software. We may hence see improvements over time.

Tron has also shared the Exynos 2200 GPU has a big difference in 3DMark performance (Wild Life) over the Snapdragon 888 chip. Of course, the SD888 is a chip from last year so the comparison isn’t exactly fair. But GPUs on flagship Snapdragon processors are known to vastly outperform the competition, so even reaching up to their mark is quite an achievement.

The tweet further reveals that the Samsung Exynos 2200 will be making use of the new ARMv9 architecture that was launched this year. This one focuses big on AI, machine learning, and security, meaning that the E2200 will be benefitting from it all as well.

It remains to be seen how well the Exynos 2200 will be performing against the likes of the Apple A15 and Snapdragon 898, which are its true rivals. The chip had, after all, already beaten the A14 Bionic in one case. But while its peak GPU performance is indeed impressive, most gamers swear by sustained performance since it’s more meaningful, and Samsung still needs to work a bit on that.