Nadolazeći AMD socketi, ploče i procesori - Stranica 345

The Exiled · 29.03.2025., 15:12

Citiraj:

AMD Ryzen 9000G APUs for AM5 socket expected launch in Q4 2025

Citiraj:

AMD may be planning to introduce its next-gen APU series around Q4 2025. The obvious candidate for this upgrade is Strix Point, which would shift the Ryzen G-Series (Desktop APUs) from Zen 4 and RDNA 3 to Zen 5 and RDNA 3.5. Assuming it’s the Strix Point, these APUs would support the AM5 socket and would likely also enable AMD to launch the Ryzen 9000F series, budget-oriented models without integrated graphics. Such variants are popular among budget gaming prebuilt PCs. AMD's Ryzen 9000G "Strix" APUs family is supposedly the Strix Point variant of the laptop series, which was launched in the second half of 2024. Going by the specs' comparison of the Ryzen 8000 Phoenix Point mobile APUs and the Ryzen 8000G series for desktops, it appears that AMD might release up to a 12-core Ryzen 9000G processor.

Izvor: VideoCardz

The Exiled · 15.04.2025., 09:27

Citiraj:

AMD achieves first TSMC N2 product silicon milestone

Citiraj:

AMD today announced its next-generation AMD EPYC processor, codenamed "Venice," is the first HPC product in the industry to be taped out and brought up on the TSMC advanced 2 nm (N2) process technology. This highlights the strength of AMD and TSMC semiconductor manufacturing partnership to co-optimize new design architectures with leading-edge process technology. It also marks a major step forward in the execution of the AMD data center CPU roadmap, with "Venice" on track to launch next year. AMD also announced the successful bring up and validation of its 5th Gen AMD EPYC CPU products at TSMC's new fabrication facility in Arizona, underscoring its commitment to U.S. manufacturing.

"TSMC has been a key partner for many years and our deep collaboration with their R&D and manufacturing teams has enabled AMD to consistently deliver leadership products that push the limits of high-performance computing," said Dr. Lisa Su, chair and CEO, AMD. "Being a lead HPC customer for TSMC's N2 process and for TSMC Arizona Fab 21 are great examples of how we are working closely together to drive innovation and deliver the advanced technologies that will power the future of computing." "We are proud to have AMD be a lead HPC customer for our advanced 2 nm (N2) process technology and TSMC Arizona fab," said TSMC Chairman and CEO Dr. C.C. Wei. "By working together, we are driving significant technology scaling resulting in better performance, power efficiency and yields for high-performance silicon. We look forward to continuing to work closely with AMD to enable the next era of computing."

AMD's 6th Generation EPYC 'Venice' is expected to be based on the company's Zen 6 microarchitecture and is expected to be launched sometime in 2026. The CPU will rely on CCDs to be made on TSMC's N2 (2nm-class) fabrication process, so it is about time for the company to get the first Venice CCDs out of the fab. Yet, the fact that AMD already has chips it can talk about highlights the long-standing collaboration between AMD and TSMC as well as the culmination of joint efforts to build chips on one the most advanced process technologies that TSMC has ever developed to date. TSMC's N2 is the foundry's first process technology that relies on gate-all-around (GAA) nanosheet transistors. The company expects its manufacturing technology to offer either a 24% to 35% reduction in power consumption or a 15% increase in performance at constant voltage, along with a 1.15X boost in transistor density compared to the previous N3 (3nm-class) generation. These gains are primarily driven by the new type of transistors and the N2 NanoFlex design-technology co-optimization framework.

AMD's announcement comes after its arch-rival Intel delayed the release of its next-generation Xeon 'Clearwater Forest' processor made on its 18A manufacturing technology (which is set to rival TSMC's N2) to the first half of next year.

Izvor: TechPowerUp i Tom's Hardware

EDIT:

Citiraj:

TSMC unveils next-generation A14 process at North America Technology Symposium

Citiraj:

TSMC today unveiled its next cutting-edge logic process technology, A14, at the Company's North America Technology Symposium. Representing a significant advancement from TSMC's industry-leading N2 process, A14 is designed to drive AI transformation forward by delivering faster computing and greater power efficiency. Compared with the N2 process, which is about to enter volume production later this year, A14 will offer up to 15% speed improvement at the same power, or up to 30% power reduction at the same speed, along with more than 20% increase in logic density. Leveraging the Company's experience in design-technology co-optimization for nanosheet transistor, TSMC is also evolving its TSMC NanoFlex standard cell architecture to NanoFlex Pro, enabling greater performance, power efficiency and design flexibility.

The first iteration of 14A does not have backside power delivery. It was the same with N2 which was followed by A16 with Super Power Rail (SPR). SPR for A14 is expected in 2029. The TSMC 16A specs were updated as well. 16A is the first version of SPR for reduced IR drop and improved logic density. This has the transistor connection on the back. SPR is targeted at AI/HPC designs with improved signal routing and power delivery. A16 is on track for production in the second half of 2026. In comparison to N2P, A16 provides an 8-10% speed improvement at the same power, 15-20% power reduction at the same speed.

Izvor: TechPowerUp, SemiWiki, Bloomberg i Reuters

Citiraj:

TSMC is currently building at 24 factories worldwide + Next-Gen Packaging: TSMC to CoWoS, SoIC, SoW, HBM Base Dies, Optics and more

Citiraj:

There are figures and statistics that are unparalleled. The company illustrates how large TSMC is through the construction of 24 Fabs. This will enable the Group to consolidate its predominance in the field of contract manufacturing for other companies for decades. Years ago, every single factory construction was celebrated and TSMC's new construction project in the USA with three new factories also became a political event, but today individual fabs fade away when the overall picture is viewed at TSMC: 24 plants are currently under construction. 15 plants are currently being built in Taiwan, eleven wafer factories and four packaging facilities. The USA has caught up as a location through the last major investment with six factory buildings, all of which are phases of the Fab 21 in Arizona – the first already work – as well as two packaging facilities. A research campus is also located in the USA. In addition, there is a second factory in Japan and the plant in Germany.

By 2020, TSMC has built an average of three factories per year, and from 2021 to 2024, the average was raised to 5 fables per year. This year, the number has already climbed to nine falleries – eight wafer fabs and a packaging facility. Since the plants are built on average over three to four years, the large number of currently 24 simultaneous buildings come together. Packaging was another important topic at TSMC's 2025 Symposium. CoWoS-L, SoIC Gen 3 and innovations are coming to the fore. For increasingly complex types of chips, continuous improvements are also needed. These are realised together with the suitable new factory buildings. In production since 2021, CoWoS (Chips on Wafer on Substrate) has become a cash register hit. No modern larger chip is available without this technology. In the future, many of the possibilities mentioned will not be limited to servers, PCs, communication systems and smartphones, but will be extended in all directions.

Izvor: ComputerBase

Citiraj:

TSMC starts building US chip factory within hours of receiving permit – AMD CEO says 2nm HPC chips in play

Citiraj:

The Taiwan Semiconductor Manufacturing Company (TSMC) quietly broke ground at a third chip fabrication facility in Arizona yesterday, shared a press release by the Commerce Department. TSMC received the permit to start building the site yesterday and had construction equipment start working at the site within hours of receiving government approval. The ground breaking of a new factory follows a $100 billion commitment by the firm to invest in the US announced by President Trump earlier this year and is part of TSMC's Arizona campus, which kicked off construction under a previous $65 billion investment by the firm in the US.

TSMC's Arizona campus is the largest operating presence of the firm in the country. The firm's $65 billion initial investment had outlined three fabs, with the third final fab intended to produce advanced products on technologies such as the 2-nanometer node. TSMC broke ground on the third fab at the Arizona site yesterday, with Commerce Secretary Howard Lutnick visiting the site as soon as construction started. The primary goal of the Arizona fab is to provide US companies with their chip needs domestically instead of having to source the products from overseas, primarily Taiwan. Apple, nVidia and AMD's CEOs all shared their prepared remarks for the event.

AMD CEO Dr. Lisa Su went a step further and shared that her firm plans to use TSMC's N2 manufacturing process for high performance computing (HPC) applications. These chips are used in modeling, research, AI and other applications. Su shared that her firm is "a lead HPC customer for TSMC's N2 process and for TSMC Arizona Fab 21" and praised TSMC's technologies for enabling AMD to deliver advanced products.

Izvor: Bloomberg

The Exiled · 07.05.2025., 10:25

Citiraj:

Zen 5 architecture boosts AMD Client and Data Center sales

Citiraj:

AMD just wrapped up Q1 2025 with a strong showing thanks to its Zen 5 architecture. The company pulled in $7.44 billion in revenue, up 36 percent from this time last year, and net income jumped to $709 million from $123 million. That boost came from two main areas: client CPUs and data center chips. On the client side, Ryzen desktop and notebook processors generated $2.29 billion, a 68 percent increase over Q1 2024’s $1.37 billion. New parts like the Ryzen 7 9800X3D and 9 9950X3D helped, and AMD also scored a lot more OEM deals. In fact, major PC makers such as HP, Lenovo, Dell, and Asus are offering 80 percent more systems with Ryzen Pro than they did a year ago.

Over in the data center world, EPYC “Turin” chips are now shipping with all the big server vendors, and that lifted revenue from $2.36 billion to $3.67 billion. AMD also ran its first test wafers at TSMC’s new Arizona plant, using their N4 process. Those chips will start production in the U.S. later this year. Meanwhile, AMD and TSMC are testing the next-gen “Venice” silicon on the N2 node, aiming for a 2026 launch.

Not everything lit up the scoreboard, though. The gaming graphics group saw revenue slip 30 percent to $647 million, mainly because console-chip royalties dropped – even though the Radeon RX 9000 series launched. On the pro-GPU side, the Instinct accelerators are ready for prime time, but U.S. export controls are holding back the MI308 models bound for China unless AMD secures special licenses. That’s expected to shave about $700 million off Q2 and up to $1.5 billion over the full year. Despite that, AMD still anticipates double-digit growth in Instinct cards once licensing catches up.

Intense competition remains an ever-present challenge. In the crucial data center and AI market, Nvidia maintains a dominant market share, estimated at over 80%. While AMD’s Instinct accelerators like the MI300 series offer competitive hardware, its ROCm software ecosystem continues to work to bridge the gap with Nvidia’s deeply embedded CUDA platform. AMD has acknowledged this, describing its effort on ROCm as being in a “wartime stance” with a new sense of urgency. Intel also competes in data center CPUs and is developing its own AI accelerators. Intel remains a significant competitor in the Client segment, although AMD has made notable market share gains with its Ryzen processors.

In the decade since Su took the top job at AMD, the company has transformed itself into a leading provider of technology across the computing industry.
It now reports more than five times the revenue it had 10 years ago and is no longer in the shadow of Intel.

Izvor: Guru3D, ComputerBase, EE Times i Bloomberg

Dottore · 07.05.2025., 16:45

Trenutačno je jako slaba ponuda matičnih ploča sa 10Gbe LAN i škrtoj ponudi pcie x8. Uglavnom više manje odnosilo na B850m i x870 između 2,5 i 5Gbe. Prije sa AM4 x570 ATX verziji bili dostupni 10Gbe

The Exiled · 07.05.2025., 16:58

Modeli poput ProArt X670E-Creator Wi-Fi nude nativno 10GbE + 2.5GbE, ali ta ploča košta skoro 200€ skuplje, nego da kupiš npr. MSI Tomahawk X670E Wi-Fi i zasebnu PCI-E 10GbE karticu.

The Exiled · 12.06.2025., 23:14

AMD confirms next-gen EPYC Venice “Zen 6” CPUs with 256 cores in 2026, EPYC Verano “Zen 7” CPUs & Instinct MI500 GPUs for 2027
AMD says Instinct MI400X GPU is 10X faster than MI300X, will power Helios rack-scale system with EPYC 'Venice' CPUs
AMD announces MI350X and MI355X AI GPUs, claims up to 4X generational performance gain, 35X faster inference
AMD EPYC Venice boasts 256 cores and bandwidth galore — next-gen server CPUs arrive in 2026
AMD launches Instinct MI350 series, confirms MI400 in 2026 with 432GB HBM4 memory
AMD Instinct MI350: With up to 1,400 watts and new package against Goliath nVidia
AMD supercharges AI Data Centers with powerful MI350 GPU at Advancing AI 2025
From MI350 to MI500: AMD’s bold AI accelerator roadmap through 2027

Citiraj:

AMD manages to shift the dynamics of data center markets into its favour, server CPU market share now poised to match Intel by 2026. AMD manages to scale up to a 50% market share in the server CPU segment in less than a decade, credited to competitive EPYC CPUs. Interestingly, AMD's share in the server CPU market back in 2017 was zero percent, since the firm hadn't come up with competitive offerings at all. When the firm's CEO took over the office, AMD prioritized the workstation segment, coming up with the EPYC "Naples" series, which brought Zen architecture to the platform for the first time. Fast-forward a few years, and we see AMD gaining adoption at a much larger stage, mainly catalyzed by the EPYC lineups.

07.05.2025., 16:45	#10324
Dottore Extrema Thule Moj komp Datum registracije: Feb 2005 Lokacija: 「 ✖ ✖ ✖ 」 Postovi: 7,582	Trenutačno je jako slaba ponuda matičnih ploča sa 10Gbe LAN i škrtoj ponudi pcie x8. Uglavnom više manje odnosilo na B850m i x870 između 2,5 i 5Gbe. Prije sa AM4 x570 ATX verziji bili dostupni 10Gbe __________________

07.05.2025., 16:58	#10325
The Exiled McG Moj komp Datum registracije: Feb 2014 Lokacija: Varaždin Postovi: 7,988	Modeli poput ProArt X670E-Creator Wi-Fi nude nativno 10GbE + 2.5GbE, ali ta ploča košta skoro 200€ skuplje, nego da kupiš npr. MSI Tomahawk X670E Wi-Fi i zasebnu PCI-E 10GbE karticu. __________________ AMD Ryzen 9 9950X \| Noctua NH-U12A chromax.black \| MSI MAG B650 Tomahawk Wi-Fi \| 128GB Kingston FURY Beast DDR5-5200 \| 256GB AData SX8200 Pro NVMe \| 2x4TB WD Red Plus \| Fractal Define 7 Compact \| Seasonic GX-750 AMD Ryzen 5 7600 \| Noctua NH-U12A chromax.black \| MSI MAG B650 Tomahawk Wi-Fi \| 128GB Kingston FURY Beast DDR5-5200 \| 256GB AData SX8200 Pro NVMe \| 2x12TB WD Red Plus \| Fractal Define 7 Compact \| eVGA 650 B5