Nadolazeći AMD socketi, ploče i procesori

[Manuel Calavera]

Autor Drug Brko

Pretpostavljamo onda da to bude i službeno predstavljanje i launch ili još jedan demo "evo samo što nije"? Nadam se ovo prvo
Epyc nam je uglavnom nebitan no svakako bi voljeli vidjeti konkurentnu GPU i taktove novog Ryzena.

Za ryzen i epyc bi trebao biti launch, odnosno predstavljanje pred sam launch (tipa izlazi za tjedan-dva, max mjesec dana). Znači sve živo bi se trebalo znati - modeli, taktovi, cijene itd. a za Navi bi trebalo biti samo prvo službeno predstavljanje (kao za ryzen u 1. mjesecu), a sam izlazak pitaj boga kada (do kraja god. vjerojatno)

[The Exiled]

Ma taman da apsolutno ništa ne lansiraju, već zaspamaju medije PR materijalom - dobro bude, jer Intel i ionak već zadnje dvije godine trči ko kokoš bez glave po dvorištu.

[stef]

Ak će predstavit negdje, logično da će biti na Computexu jer je to zadnja prilika prije jeseni, bar što se tiče velikih evenata. Availability ne mora biti taj dan.

[The Exiled]

Ma je, ovo prije je bila zajebancija s moje strane, jer cijela stvar podsjeća na 28-core Xeon W čiler edišon, koji je pokazan na drž-ne daj, jer je sljedeći dan AMD službeno predstavil Threadripper 2990WX.

U svakom slučaju, izgleda da se AMD lagano priprema.

[Manuel Calavera]

Vidim i da im je dionica opet 29$, ovaj put vjerojatno da ostane tako (i na više), bez velikih padanja


GPU prodaja im je u ogromnom padu naspram prošle god. jer nema mininga

AMD expects its first-quarter 2019 revenues to slip 12% sequentially and 24% on year due to seasonality and weakening graphics card shipments. Although the company launched its latest 7nm Radeon VII GPU in February, the product alone was still not sufficient to offset the sharp decline created by the dissipation of the cryptocurrency mining fad.
...
At the moment, AMD's profits from the GPU business is mainly contributed by demand from the datacenter market, while Google's new cloud stream gaming platform Stadia is also adopting AMD's GPUs for the system.

[The Exiled]

TSMC makes progress on 5nm with complete infrastructure design and risk production

TSMC has announced that it has completed infrastructure design on its 5nm process node that will leverage the company's second generation of extreme ultraviolet (EUV), as well as deep ultraviolet (DUV) lithography. TSMC's 5nm chips will be aimed at SoC designs, 5G mobile applications, AI, and high performance computing. TSMC has also completed development of its 5nm Design Rule Manual (DRM), SPICE model (simulation program with integrated circuit emphasis), and process design kits (PDK). The entirety of the 5nm design infrastructure is available for TSMC customers via TSMC Online. TSMC also notes it has chips in risk production with its alpha customers, which likely indicates volume production could ramp in 2020. That would align with TSMC's previous plans to hit 5nm by 2020 and 3nm by 2022. Their 5 nm node continues to use bulk silicon FinFET transistors. Leveraging their experience from 7nm+, 5nm is expected to utilize EUV for more critical layers in order to reduce the multi-patterning complexity. The 5nm node is expected to deliver a 15% improvement in performance at constant power or a 20% reduction in power at constant performance. TSMC is targeting both low-power mobile and high-performance compute with this node.

Why AMD EPYC Rome 2P will have 128-160 PCIe Gen4 lanes and a bonus

With AMD EPYC “Rome”, there are two I/O die with x86 cores that hang off the I/O die. When we look externally, AMD has a set number of pins that need to carry data to and from the socket. That includes pins for eight memory channels along with eight x16 links. We are going to leave the memory channels alone, and discuss the x16 links. With AMD EPYC 7002 or “Rome” already has PCIe Gen4 support announced, while the 2nd Gen Intel Xeon Scalable CPUs still use 14nm and Intel’s PCIe Gen3 IP. In a single socket, that means that AMD EPYC Rome will have 128 PCIe lanes using eight x16 links for PCIe. We expect AMD will use the Rome generation to add another PCIe lane, making 129 PCIe lanes total, and we are going to discuss that in our “Bonus Lanes” section later. Unlike Intel’s PCH, AMD’s chips are designed to have all I/O on package. When looking at 1st and 2nd Gen Intel Xeon Scalable systems versus AMD EPYC, one has to remember that Intel systems include a Lewisburg PCH that uses power and adds cost. AMD had a goal of using flexible I/O and its Server Controller Hub (SCH) to remove the need for the PCH.

Izvor: TechSpot, TechPowerUp, WikiChip, RetiredEngineer i ServeTheHome

[Omega 3]

Koliko sam shvatio Zen 2 dizajn svaki taj čiplet sadrži do 8 jezgri? Hoće li modeli do 8 jezgri ići na 1 chiplet ili će raspodijeliti na 2? Pitam zbog latencija, ako bi išli na 1 chiplet onda bi latencije trebale biti vrlo niske između jezgri?

[Manuel Calavera]

8 core je prikazan kao 1 čip, no Škot je siguran da će 8 core biti 4 jezgre u jednom čipu, a 4 u drugome (vjerojatno zbog yieldova na novom nodeu), a 8 core čipovi vjerojatno za epyc i eventualno APU

Što se tiče latencija nema smisla uspoređivati sa postojećim zenovima jer je zen2 nešto posve drugo - ima I/O die, navodno postoje veze i između čipova sa jezgrama + navodno je infinity fabric duplo brži odnosno veći bandwidth

Uglavnom sve te brojke i teorija su nebitne, kad dođe gotov proizvod vidjet će se koliko je točno brži u konretnim stvarima

[Blue Spirit]

MSI Betrays AMD's Socket AM4 Longevity Promise: No Zen2 for 300-series?

MSI will support Zen 2 processors only on X470 amd B450 chipsets

https://www.techpowerup.com/254634/m...for-300-series


Greedy motherboard vendors such as MSI want you to buy a new motherboard every two generations of processor for no sound reason at all.

MSI is reportedly blocking support for 3rd generation Ryzen "Matisse" processors on its AMD 300-series chipset motherboards, including those based on high-end AMD X370 and OC-capable B350 chipsets.

This would also put those who own $300 motherboards such as the X370 XPower out of luck. To recap, AMD announced on numerous occasions that it doesn't want to be a greedy clique like its competitor, by forcing motherboard upgrades and promised that socket AM4 motherboards will be backwards and forwards compatible with at least four generations of Ryzen processors, running all the way up to 2020.

This normally should mean that any 300-series motherboard must support 4th generation Ryzen processors with a simple BIOS update.

Most 300-series motherboards, including from MSI, even ship with USB BIOS Flashback feature to help with forwards compatibility.

Unfortunately, motherboard companies such as MSI care more about their bottom-lines than the consumer.
In a support e-mail to an X370 XPower Titanium owner, MSI confirmed that it will not extend Zen 2 support to AMD 300-series.
Other motherboard vendors could follow MSI's suit as a representative of another motherboard vendor, on condition of anonymity, told TechPowerUp that "Zen 2" processors have steeper electrical requirements that 300-series motherboards don't meet.
This is an excuse similar to the one Intel gave for the planned obsolescence of its 100-series and 200-series chipsets, even as it was repeatedly proven that those motherboards can run and overclock 9th generation processors with custom firmware just fine.
Would MSI care to explain whether a B450M PRO-M2 has a stronger VRM than an X370 XPower Titanium to warrant "Zen 2" support? Will all "Zen 2" processor SKUs have steep electrical requirements?
Will there not be any SKUs with double-digit-Watt TDP ratings?

[Tajl3r]

Sta sad triba isti rumor na 5 thredova razvlacit.

Sent from my Redmi Note 5 using Tapatalk

[Manuel Calavera]

Očekivano je demantirano (tema o matičnim pločama)


Nego:

AMD 7nm chips for next-generation PlayStation to be ready in 3Q20

Ako je ovaj vremenski okvir Q3 točan (prije se spominjala 2020) - PS5 bi možda mogao imati zen 3 mikroarhitekturu na 7nm+ (od Navija ionako očekujem refresh u drugoj generaciji)

[The Exiled]

U svakom slučaju, AMD za PS5 sprema čudovište i kroji ga prema Sonyjevim željama.

[Manuel Calavera]

AMD Zen 3 Based Ryzen, Threadripper and EPYC Processors Utilizing TSMC’s 7nm+ Process Node Could Feature Up To 20% More Transistors Than Zen 2 CPUs

[Manuel Calavera]

Što se tiče APU-a izgleda da se događa ono što se već spominjalo kao opcija - 3xxx series apu je current gen (zen+, 12nm i vega)

klik

APU sa zen2 i Navi očito onda neće još godinu dana (4xxx series apu). Uglavnom potvrđuje se da je APU uvijek jedna generacija iza običnih ryzena - prve generacija APU-a je došla dosta kasnije, (uoči zen+) sad tek izlazi zen+ uoči zen2 cpu itd.

[toshiro]

AMD Ryzen 3000-Series APUs Reportedly Feature Solder, Overclock Better Too

https://www.tomshardware.com/news/am...ock,39154.html

Zanimljivo.

[Manuel Calavera]

AMD-ov fanzin javlja da su yieldovi 70% za zen2 die od oko 88mm2

https://www.bitsandchips.it/english-...-but-very-good

Dobar je za zen, no jasno je zašto Navi neće imati high end (pretpostavljam da yieldovi ni za čip od 200-300mm2 nisu bajni)

[Manuel Calavera]

AMD Ryzen 3000 CPUs Allegedly Feature 15% Better IPC, Up To 4.5 GHz – X570 Rumored To Have 40 PCIe Lanes, No 3rd Gen Ryzen Support on A320 Boards

[Brko]

Po tome bi Zen2 4.5GHz bio ekvivalent Zen/Zen+ na 5+ GHz. Tj. Zen2 na 4 GHz ekvivalent Zen/Zen+ na 4.5+ GHz.

A također, ako je 15% rast, onda to ide i iznad Coffee Lake Refresh procesora.

No i dalje sve uzimati sa lopatom soli.

Sent from my SM-G965F using Tapatalk

[Manuel Calavera]

Ja sam siguran da prebacuju intel sa IPC (relativan pojam jer ovisi u čemu, no recimo singlecore cinebench, treba im više od 7% sa zen+), mediji tvrde da je u rasponu od 10% do 15%, a za clockove ćemo još vidjeti. Za mene 4.5ghz zvuči fenomenalno (ako je to ekvivalent 4.2ghz od zen+ ili 3.9-4ghz od zena), jer AMD čitavo vrijeme ima spiku - napredniji node = niži clockovi, tako da je bilo kakav rast plus.

These CPUs were early samples which were only supplied for internal testing and not necessarily indicate final retail performance or clock speeds.

It is also reported that the memory controller has got an upgrade but it isn’t as significant as we had expected.

[The Exiled]

Fifty Years of AMD, a Tech Leader

• AMD was founded the year after Intel. As with Intel, AMD’s founders, including CEO Jerry Sanders, were alumni of Fairchild Semiconductor. While Intel’s founders came from the semiconductor pioneer’s engineering and management ranks, Sanders came from the sales department at Fairchild, where he was a top salesman. AMD eventually began selling MSI logic products, including the Am9300 shift register and the Am2501 binary/hex up/down counter. AMD had a solid, but a bit more expensive, selection of logic products. A major AMD differentiator was that it offered an enhanced military-grade reliability standard, Mil-Std-883B, for no additional cost. AMD built a number of second-source products, but during the 1980s the company transitioned to more innovative chips based on AMD research and development. AMD built one of the earliest Ethernet controllers (the Am7990) and an early 802.11b controller (before if became known as Wi-Fi). In addition, AMD built networking controllers for bleeding-edge fiber distributed data interface (FDDI), along with Ethernet switches and hubs. Other unique products included a video compressor/decompressor processor (VCEP) for image compression, a scalable 2D graphics chip (QPDM), ISDN communications chips, and programmable array logic (PAL) devices, including the ultimate PAL — the 22V10.

• The company had signed a second-source agreement for the 8086/8088 and subsequent x86 processor in support of Intel’s design win for the IBM PC. Back in those days, second-source agreements were common (as were reverse-engineered products). The rationale for having a second source was that the semiconductor manufacturing processes were still fickle and the semiconductor companies only had a single manufacturing plant. A semiconductor manufacturing plant (commonly called a “fab”) could suddenly “lose the recipe” and product yields would plummet. For IBM, to rely on a sole-source partner (Intel) was deemed too risky for the new PC product. IBM and Intel thus approached AMD to manufacture an exact copy of the Intel processors: such that, if Intel’s fab had problems, IBM could then source chips from AMD while Intel fixed the problem. As the IBM PC compatible business became more and more lucrative, Intel determined it didn’t want to share the business and designs with AMD. AMD had even produced faster 80286 and 80386 processors than Intel.

• AMD eventually built its own Am486 silicon design and gained access to the Intel 80486 microcode — though not the Pentium processor. But AMD had already embarked on an independent processor development path. The first clean-slate x86 processor by AMD was the AMD-K5. While AMD’s K5 was an engineering triumph, the chip took longer to develop than anticipated and it lagged behind the Pentium processor in clock speeds. To keep pace with Intel, AMD needed a second design team and a new CPU design. Jerry Sanders found the x86 processor and the design team the company needed in a startup named Nexgen. Sanders took a gamble by buying Nexgen and porting their next-generation design to AMD’s semiconductor process. That product became the AMD-K6 and was followed by the AMD-K6-2, with a new set of instructions for PC games called AMD-3DNow!

• The next breakthrough for AMD PC processors came with the Athlon processor (also known as K7). AMD had hired some leading engineers from the DEC Alpha microprocessor team, including future CEO, Dirk Meyer, and had also adopted a new high-performance bus from DEC, called the EV6 bus. The Athlon processor was fabricated using AMD’s advanced process technology: it included copper interconnects (instead of aluminum) for the first time in a PC processor, and it was the first PC processor to reach 1 GHz (beating out Intel’s Pentium III by a few days).

• Unfortunately, the Athlon design and the EV6 bus could only take AMD so far: it needed another breakthrough to really step up its competition with Intel. A team was assembled under Jim Keller to build a next-generation 64-bit “K8” processor that was still backward compatible with the 32-bit x86 instruction set and with a more scalable processor interconnect. The code name of the chip was SledgeHammer. Although Keller left during the design, a Nexgen engineer named Fred Weber took over the project. Under Weber, AMD developed the AMD-64 instruction set (also known as x86-64) and built the Opteron server processors and Athlon 64 PC processors. These design elements became the foundation for the x86 processors we have today. The K8 processor caught Intel by surprise. Intel’s internal politics had stopped the release of Xeon processors with 64-bit instruction extensions because the company was committed to the Itanium processor, codeveloped with HP, for its 64-bit roadmap. This proved to be a strategic error for Intel and opened a product gap for AMD’s Opteron to enter the mainstream server market as well as the high-end PC market. The K8 system architecture was more modern than Intel’s Itanium and Xeon.

• While AMD was absorbing an expensive ATi merger, Intel struck back with a new CPU design, code-named Conroe, from its Israeli design team. Conroe was 64-bit, lower power, had higher performance, and had adopted the AMD SledgeHammer-style platform design. This new competition, along with the burdensome debt it had taken on, and the recession, put AMD in a financial bind. Intel had also been able to integrate 3D graphics into its microprocessors, negating some of the Fusion APU advantage (AMD still had superior graphics). This led to AMD selling its handheld graphics group to Qualcomm and spinning off its semiconductor manufacturing plants under a new company, called GlobalFoundries, to help raise cash and to lower R&D and operating costs. Ultimately, AMD’s decision to become a fabless semiconductor company became an asset — it was able to use multiple best-in-class foundries including GlobalFoundries — which can be seen today in AMD’s ability to use TSMC’s advanced 7-nm processor for the latest graphics and forthcoming server and PC processors.

Izvor: EE Times | Electronic Engineering Times

TechSpot: The rise and fall of AMD + ArsTechnica: The rise and fall of AMD

CPU AMD Ryzen 9 3850X 4.3 GHz
https://nguyencongpc.vn/cpu-amd-ryzen-9-3850x-4-3-ghz/



CPU AMD Ryzen 9 3800X 3.8 GHz
https://nguyencongpc.vn/cpu-amd-ryzen-9-3800x-3-8-ghz/



CPU AMD Ryzen 7 3700X 4.2 GHz
https://nguyencongpc.vn/cpu-amd-ryzen-7-3700x-4-2-ghz/



CPU AMD Ryzen 5 3600x 4.0GHz
https://nguyencongpc.vn/cpu-amd-ryzen-5-3600x-4-0ghz/



CPU AMD Ryzen 5 3600 3.6 GHz
https://nguyencongpc.vn/cpu-amd-ryzen-5-3600-36-ghz/

[thunderbird]

A boosta do 4,7 Ghz

[Brko]

4.7 GHz × 1.15% IPCa = 5.4GHz na Zen1 uArch
Fino

Ako ću neš uzimati, to bude 3600X ukoliko ne bude 300+ kn skuplji od 3600.

Sent from my SM-G965F using Tapatalk

[warlock25]

3600x znaci oc bi mogao na cca 5ghz ako boosta na 4.8.lol

Poslano sa mog H8266 koristeći Tapatalk

[Manuel Calavera]

Hold your pants, netko je samo prepisao škotove info stare skoro pola godine. Nema ničeg novog, nisam siguran da je legit, jer je prerano da trgovine stavljaju ponude

I dalje ne vjerujem u ikakvo prismrđivanje 5ghz, no očito ću morati čekati službene info da se uvjerim u išta preko 4.5ghz

tako nesto, cine se nerealno visoke boost vrijednosti, kako god nadam se da ce b350 plocice primati 16 jezgri / 125w

[Brko]

Autor Manuel Calavera

Hold your pants, netko je samo prepisao škotove info stare skoro pola godine. Nema ničeg novog, nisam siguran da je legit, jer je prerano da trgovine stavljaju ponude

I dalje ne vjerujem u ikakvo prismrđivanje 5ghz, no očito ću morati čekati službene info da se uvjerim u išta preko 4.5ghz

Možda X budu zabrijavali na 4,7GHz sa 1 jezgrom, ali da, 4,5GHz vjerojatno bude max. 4,4GHz sa pristojnim naponom/temperaturom u gamingu i ako stvarno IPC bude 13-15% od Zen/Zen+, biti će to i više nego dobro

[Manuel Calavera]

Autor Drug Brko

Možda X budu zabrijavali na 4,7GHz sa 1 jezgrom, ali da, 4,5GHz vjerojatno bude max. 4,4GHz sa pristojnim naponom/temperaturom u gamingu i ako stvarno IPC bude 13-15% od Zen/Zen+, biti će to i više nego dobro

Na kraju krajeva sam AMD apsolutno ničim nije dao neki bitan hint o nekim većim clockovima (čak štoviše tvrde da je sad situacija novi node = niži clockovi). Jedino što imamo jesu adoredove tvrdnje i "maybe 5ghz" tvrdnje GloFo čelnika, no više se uopće ne radi o tom nodeu otkupljenom od IBM-a, jer je to has been i nema veze sa TSMC

Mislim da će biti odlična stvar samo zbog IPC-a, a kamoli ako bude nekog povećanja clockova


EDIT

Najbolji živući dokaz TSMC 7nm nodea je i jedini proizvod na istom - Radeon VII. Znači identična arhitektura je sa 14nm spuštena na 7nm, sav power headroom što su dobili sa novim nodeom su iskoristili na clock i dobili - cca 8-9%. Ponavljam prelaskom sa 14nm, ne sa 12nm. Tom logikom - ono što je na Ryzenu 1st gena 4.0ghz to bi na 7nm bilo 4.35ghz, kad nebi bilo nekih promjena arhitekture (i kad bi skaliralo jednako kao sa gpu). Nešto su možda dobili promjenom arhitekture (jer 4.2-4.3 se dobivalo na 12nm) i nekih 4.5ghz je možda baš to.

Autor trpi

tako nesto, cine se nerealno visoke boost vrijednosti, kako god nadam se da ce b350 plocice primati 16 jezgri / 125w

Iskreno nebih se nadao, 16c je too much za 4 phase, a mislim da su sve B ploče 4 phase. Ne radi se tu toliko o tdp (koji je ionako marketing značajka, piše tko šta hoće) nego amperaži koju može povući. Buldzoid je također govorio da je OC-an 2700X neki plafon za B350/4 phase ploče. E sad 7nm će biti efikasniji, no i 8 core će se bolje clockati pa to anulira...tako da 8 core možda ostane plafon, mooožda 12 core zbog bolje efikasnosti

Na kraju krajeva uopće nije sigurno dali AMD želi izbaciti 16 core (više od 8 tj. 12 je sigurna stvar). Bilo je glasina i da bi 16c32t izbacili nešto kasnije

[The Exiled]

65W Ryzen 5 3600G 8C/16T@3.2GHz + iGPU i to je to kaj se mene tiče. Sve više i bolje od toga po pitanju broja jezgri, visokog (OC) takta, IPC-a, potrošnje i popratne skalamerije je dobrodošlo, ali ne i presudno.

Autor Manuel Calavera

Na kraju krajeva uopće nije sigurno dali AMD želi izbaciti 16 core (više od 8 tj. 12 je sigurna stvar). Bilo je glasina i da bi 16c32t izbacili nešto kasnije

16C/32T AM4 Zen 2 Ryzen 3000 bi mogla biti točka na i, jer Intelov marketing već sad priprema teren za njihove 10C/20T mainstream modele.

[Sinac]

https://www.guru3d.com/news-story/am...ne-stores.html

Eto, i drugi prenose navedene vijesti. Kak se kaže, gdje ima dima...