Latest news with #Exynos2600
Phone Arena
3 hours ago
- Business
- Phone Arena
Samsung faces a deadline to choose which processor will power the Galaxy S26 and Galaxy S26+
There is a lot of pressure on Samsung to design a more than merely capable Exynos 2600 application processor (AP). After all, the company is looking to deploy a homegrown Exynos SoC inside next year's Galaxy S26 and Galaxy S26 + flagship models in the vast majority of markets. Due to yield issues at Samsung Foundry, during the last two years Samsung had to dig deep into its pockets and pay more to equip all of its flagship series phones with the latest Snapdragon chipsets instead of using its own Exynos APs. Samsung had to pay $400 million to use the Snapdragon 8 Elite AP across the entire Galaxy S25 lineup. | Image credit-Qualcomm The reason that yield is so important is that the figure is a percentage showing how many usable dies have been created against the maximum number that can be diced from a silicon wafer. Samsung Foundry was rumored to have a 30% yield for production using the same process node earmarked for the Exynos 2500. A low yield increases the number of defective dies and decreases the number of usable ones. As a result, not only does the price for working chips rise, at a yield of 30% it was possible that Samsung would not be able to build enough chips to cover the amount it needed. Samsung reportedly started manufacturing a prototype of the Exynos 2600 AP and the results of the chipset's first Geekbench run through was impressive with a single-core score of 2400 and a multi-core tally of 10200. The impressive part was that the multi-core score topped that of the A18 Pro. We should point out that the latter chipset, used to power the iPhone 16 Pro and iPhone 16 Pro Max, is built by TSMC using its second-generation 3nm node; the Exynos 2600 is produced using Samsung Foundry's 2nm node which uses Gate-All-Around (GAA) transistors that reduce current leaks and improve the drive current producing improved performance and energy efficiency. Also, to make this a fair comparison, the Exynos 2600 AP's benchmark scores should be compared with those of the yet to be released A19 Pro. That chipset will be built by TSMC using its third-generation 3nm process node and will debut later this year on the iPhone 17 Pro and iPhone 17 Pro Max. Another point to consider is that tweet from tipster @OreXda doesn't mention the efficiency of the Exynos 2600. As a result, it is a possibility that in order to achieve the high mjlti-core benchmark results, the chipset had to use an unrealistic amount of power, certainly more than the component will be fed if used to drive the Galaxy S26 line when the phones are released during Q1 next year. An interesting change in the configuration of the Exynos 2600 AP was mentioned by @OreXda. Originally designed with 10 cores (giving it the "decacore" designation), the tipster says that the Exynos 2600 will feature two Cortex Prime "X" CPU cores along with six Cortex-A cores. That would make the Exynos 2600 an octa-core AP. If Samsung Foundry doesn't improve its yield, Samsung would have to consider using the Snapdragon 8 Elite 2 for Galaxy to power all of the models in the flagship series with Qualcomm's flagship processor. The phones won't be released until next January or February but because of lead times, Samsung will have to decide by the end of this year whether yields are sufficient (60%-70% and higher is required) to allow Samsung Foundry to manufacture enough Exynos 2600 SoCs to power the Galaxy S26 and Galaxy S26 + units in all markets other than the U.S., China, and Canada. In those countries, the Galaxy S26 and Galaxy S26 + will be powered by the Qualcomm Snapdragon 8 Elite 2 for Galaxy AP. The top-of-the-line Galaxy S26 Ultra will sport the Qualcomm SoC in all markets. Secure your connection now at a bargain price! We may earn a commission if you make a purchase This offer is not available in your area.
Phone Arena
a day ago
- Phone Arena
Exynos 2600 may debut another in-house Samsung replacement
Samsung has been making moves to become more self-reliant for quite a while, especially focusing on its in-house processors like the upcoming Exynos 2600. A recent rumor corroborates an older report that Samsung is apparently also looking to introduce an in-house GPU on its Exynos 2600 chipset. Last year, it was reported that the Exynos 2600 would have an in-house Samsung GPU. A new leak ( translated source ) gives further credence to this claim. If true, it could be a contentious shift on Samsung's part, as many tech enthusiasts still prefer Snapdragon chips over Exynos. Another in-house Samsung component will be subject to bias, regardless of whether it is warranted or not. Nevertheless, this is just part of Samsung and Apple becoming more self-reliant to decrease costs. The iPhone 16e debuted with Apple's in-house C1 cellular modem, and the Galaxy S26 may come with Samsung's own GPU if the Exynos 2600 is ready on time. Samsung was unable to prepare the Exynos 2500 for the launch of the Galaxy S25 line, and is determined to not repeat that mistake. The entire Samsung Galaxy S25 series uses the Snapdragon 8 Elite processor. | Image credit — Samsung The Exynos 2600 will be manufactured using Samsung Foundry's 2 nm process. Exynos processors have historically lagged behind their Snapdragon counterparts, both in performance and heat management. However, I'm still hopeful that Samsung can perfect Exynos and achieve some semblance of what the Mac did with Apple silicon: an unprecedented performance jump. If the Exynos 2600 debuts in the Galaxy S26 , then Samsung will have beat Apple to a 2 nm chip. The iPhone 18 will use the 3 nm A19 chips manufactured by TSMC ( Taiwan Semiconductor Manufacturing Company ). However, if the Exynos 2600 isn't ready on time, then Samsung will likely use the more expensive 3 nm Snapdragon 8 Elite Gen 2. A former Huawei GPU lead is reportedly working with Samsung to design its in-house GPU for the Exynos 2600. Prior to this, Exynos chips have used GPUs designed after AMD's architecture. As this will be a first-generation design, it may not be as powerful as the competition, but I don't think it should take Samsung too long to improve upon this. Lastly, some consumers may not even get to use Samsung's in-house GPU on their Galaxy S26 . Reports indicate that the company is moving back to separate variants for different regions. So, while some countries will get the Exynos 2600, others may still get a Snapdragon phone instead. Secure your connection now at a bargain price! We may earn a commission if you make a purchase Check Out The Offer
Yahoo
2 days ago
- Yahoo
Exynos 2600 曝光，GPU 提高 62%，性能可觀
文章來源： Samsung Exynos 2600 處理器細節數據被曝光，這款由採用 2nm 製程工藝的處理器在架構和性能上將會有非常明顯的變動。 🚨ERROR REPORT🚨 The E2600 expected score has been miscalculated with the condition of 10% mass production yield. I apologize for false reporting, and fixed to 40% (current Samsung 2nm yield) Geekbench 6 Single-Core: ~2950 Multi-Core: ~10,200 3DMark / GFX : Same as result — Connor/코너/コナー (@OreXda) June 17, 2025 爆料顯示，為了更好的平衡性能與能效表現，Exynos 2600 放棄採用現款 Exynos 2400 的十核架構，改為「2顆Cortex-X大核+6顆Cortex-A能效核」的八核設計，更貼近 Snapdragon 8 Elite 版的配置方案。 從該處理器的性能來看，其 Geekbench 6 單核跑分有望達到2400左右，多核成績約9400，較前代分別提升 20%和 15%，多任務處理與單核運算能力有明顯的提升。 Xclipse 960 在 GPU 性能方面表現不錯，3DMark Wild Life Extreme 測試成績有望達5800分，較前代 Xclipse 940 提升62%（NotebookCheck 數據），GFXBench Aztec Ruins 測試幀率有望達 85FPS（測試分辨率還未知），圖形渲染效率顯著提升，能更好地滿足高畫質遊戲等場景對性能的需求。 儘管 Samsung 2nm 代工工藝目前試產良率約30%，但 Exynos 2600 已啓動原型處理器批量生產。Samsung LSI 處理器設計部門與代工團隊正通過工藝優化全力提升良率，為該處理器後續大規模商用上市積極做準備。
India Gazette
5 days ago
- Business
- India Gazette
Race for advanced 2nm semiconductor chips heats up in H2 2025 as TSMC, Samsung, Intel gear up for production
New Delhi [India], June 16 (ANI): The global race to lead in next-generation 2-nanometer (nm) semiconductors is expected to intensify in the second half of 2025, with top foundries TSMC and Samsung Electronics preparing to begin mass production. At the same time, Intel is also looking to outpace its rivals with the launch of its more advanced 1.8nm process technology. According to a report by The Korea Herald, Taiwan's TSMC has already started receiving client orders for its 2nm process node. The chips are expected to be produced at its Baoshan and Kaohsiung factories in Taiwan in the latter half of the year. This marks a major milestone for TSMC, as it is adopting gate-all-around (GAA) transistor architecture for the first time in its 2nm chips. The new node promises 10-15 per cent better performance, 25-30 per cent lower power consumption, and a 15 per cent boost in transistor density compared to the current 3nm technology. Meanwhile, Samsung Electronics, the second-largest foundry player, is also targeting mass production of 2nm chips in the second half of 2025. The company confirmed in its latest earnings report that it will begin producing mobile chips using its 2nm node this year. Although it didn't mention the specific product, it is widely believed to be the Exynos 2600 chip, expected to power the upcoming Galaxy S26 series in early 2026. Samsung was the first to adopt GAA technology with its 3nm chips, but initially struggled with low yield rates. The company now aims to improve its 2nm production efficiency using its earlier experience. As per The Korea Herald, TSMC currently dominates the global foundry market, holding a 67.6 per cent share in Q1 2025. It added that TSMC's 2nm yield rate has surpassed 60 per cent, a key level for stable mass production. In comparison, Samsung's yield rate reportedly stands at around 40 per cent, with a market share of 7.7 per cent. Despite the tough competition, the 2nm market outlook remains strong. TSMC Chairman C.C. Wei said that demand for 2nm chips has already surpassed that of the 3nm generation, driven by demand from smartphones and high-performance computing. He also noted that the number of 2nm chip designs, or 'tape-outs,' in the first two years is expected to exceed those of previous nodes like 3nm, 4nm, and 5nm. Counterpoint Research has predicted that TSMC will reach full utilization of its 2nm production capacity by the fourth quarter of 2025, faster than for any previous node. Key clients are expected to include Apple, Qualcomm, MediaTek, AMD, and even Intel. To stay competitive, Samsung is making strategic moves to strengthen its foundry business. The company recently hired Margaret Han, a former TSMC executive, to lead its US foundry operations. Meanwhile, US chipmaker Intel is placing its bets on the 1.8nm process, known as 18A, to turn around its foundry business. Kevin O'Buckley, general manager of Intel Foundry Services, admitted that the company had missed some of its earlier deadlines but confirmed that 18A is now on track for high-volume manufacturing in the second half of 2025. He stated, 'I'll be very direct to admit we did not deliver all of our schedules for 18A.' Intel hopes to challenge both TSMC and Samsung in the advanced node race with this new process. (ANI)
Korea Herald
6 days ago
- Business
- Korea Herald
Samsung, TSMC set stage for fierce race in 2nm chip tech
The global race to lead the market for next-generation 2-nanometer semiconductors is set to intensify in the second half of this year, with leading foundries TSMC and Samsung Electronics preparing to begin mass production. Meanwhile, Intel is looking to leapfrog the competition with the more advanced 1.8nm process, aiming to revive its struggling foundry business. According to industry sources Sunday, TSMC has already begun receiving client orders for its 2nm process node, which is expected to be produced at its Baoshan and Kaohsiung fabs in Taiwan in the second half of this year. For the first time, TSMC is adopting gate-all-around transistor architecture in its 2nm chips, marking a major technological shift. The new node is expected to offer 10-15 percent better performance, 25-30 percent power reduction and a 15 percent increase in transistor density compared to the current 3nm process. Samsung Electronics, the second-largest player in the foundry market, is also targeting the second half of 2025 for the start of its 2nm chip production. In its latest earnings report for the January-March period, Samsung confirmed that it will start mass production of mobile chips using its 2nm process node within the year. While the company did not specify what the product would be, it is widely expected to be the Exynos 2600, the upcoming application process chip for its flagship Galaxy S26 series, set for release in early 2026. Samsung was the industry's first chipmaker to adopt the advanced GAA architecture for its earlier 3nm node process, but it struggled with low yield rates in the initial stages. Leveraging its prior experience with the cutting-edge architecture, the company now aims to improve its manufacturing stability and output at the 2nm level. TSMC, which is dominating the global foundry market, securing 67.6 percent share in the first quarter this year, appears to be on track. Sources say the company has achieved yield rates above 60 percent, crossing the threshold for stable volume production. Samsung, holding a 7.7 percent market share, is tasked with raising its yield rate, which reportedly stands at around 40 percent. Amid the intense competition, the outlook for the 2nm node is optimistic, with demand projected to surpass that of the previous 3nm generation. TSMC Chairman C.C. Wei said during a recent earnings call that demand for 2nm has already outpaced that for 3nm, driven by both smartphone and high-performance computing applications. 'We expect the number of new tape-outs for 2nm technology in the first two years to exceed those of 3nm and even 5nm or 4nm,' Wei said, highlighting robust interest from key customers. Market intelligence firm Counterpoint Research echoed this projection, forecasting that TSMC will achieve full utilization of its 2nm production capacity by the fourth quarter of this year -- faster than any previous node. Major customers are expected to include Apple, Qualcomm, MediaTek, AMD, and even Intel. Samsung's challenge will be to attract top tech clients to remain competitive in the advanced node market. For the mission, the chipmaker has recently brought on board Margaret Han, a former TSMC executive, to lead its US foundry business at Samsung Device Solutions Americas. Meanwhile, US chip giant Intel is betting on its 1.8nm process, branded as 18A, to reassert itself in the foundry market. At a recent Direct Connect event, Kevin O'Buckley, general manager of Intel Foundry Services, admitted the company was behind schedule. 'I'll be very direct to admit we did not deliver all of our schedules for 18A,' he said. Still, he emphasized that 18A is now on track for high-volume manufacturing in the second half of 2025, and that Intel believes it will soon be ready to challenge TSMC and Samsung in advanced nodes.
