OpenAI's release of GPT-4.5 has been delayed due to a shortage of GPUs. CEO Sam Altman mentioned on X that the rollout will be staggered and will involve adding tens of thousands of GPUs next week. This shortage is prompting OpenAI to develop its own AI silicon in partnership with Broadcom.

The high cost of GPT-4.5, at $75 per million input tokens and $150 per million output tokens, is a concern, but Altman emphasizes its unique intelligence.

NVIDIA's GPUs are in high demand, with Blackwell GPUs sold out until October. The expansion of AI infrastructure, including massive AI supercomputers and data centers, is driving this demand.

➀ CyberPowerPC is addressing the missing ROPs issue in NVIDIA's GeForce RTX 50 series GPUs by testing each unit before shipping.

➁ The company is ensuring every RTX 50 series powered desktop gaming PC is tested for the correct amount of ROPs before they ship out to customers.

➂ NVIDIA has confirmed the issue, advising RTX 50 owners to contact their AIB manufacturer for a replacement.

➀ NVIDIA确认Blackwell Ultra和Vera Rubin AI架构进展顺利，预计在GTC 2025上将有重大宣布；

➁ 尽管初期有延误，但生产已加速，保持了年度发布节奏；

➂ 新一代Rubin AI GPU将采用先进的HBM4内存，SK hynix、三星和美光等合作伙伴参与其中。

➀ Nvidia reported a revenue of $39.3 billion for Q4 FY, up 12% q-o-q and 78% y-o-y;

➁ Revenue for FY 2025 reached $130.5 billion, marking a 114% y-o-y increase;

➂ The CEO, Jensen Huang, emphasized the growth of Blackwell AI supercomputers and the impact of AI on various industries.

➀ SK hynix has achieved a 70% yield rate on its HBM4 12-Hi memory, which is set to be used in NVIDIA's upcoming Rubin R100 AI GPUs.

➁ The test yield is an indicator that can be used to make estimations on the actual future yield rate, with SK hynix aiming for a yield upwards and into the late 90% range.

➂ TSMC, as a key partner, is expected to expand its CoWoS advanced packaging capacity to handle the large Rubin chip demand.

➀ Nvidia's RTX 50 series GPUs have been experiencing black screens and Blue Screens of Death (BSODs) after an update to their driver;

➁ Nvidia representatives suggest updating the GPU's vBIOS or installing new drivers to fix the issue;

➂ Some users have reported that the latest vBIOS update for their MSI RTX 5090 Gaming Trio OC resolved the black screen issue.

➀ This paper discusses the post-silicon validation of the IBM POWER9 processor, highlighting its significance in the context of increasing multi-core processor adoption by hyperscalers.

➁ IBM's Threadmill, a tool for processor verification, has been enhanced since the POWER7, including optimized template allocation and debugging strategies for multi-core write drops.

➂ The validation methodology for POWER9 has significantly improved bug detection rates compared to POWER8, with new techniques like AI prioritization and hardware irritators to induce bugs.

➀ NVIDIA has announced the integration of Cisco Silicon One into its Spectrum-X Ethernet networking platform, making it the only partner silicon in Spectrum-X. This move validates Cisco's technology and its importance in the AI technology landscape.

➁ Cisco will develop systems combining NVIDIA Spectrum silicon with Cisco OS software, allowing customers to standardize on both Cisco networking and NVIDIA technology in the data center.

➂ The collaboration aims to enable a new level of interoperability in the industry, leveraging both companies' strengths.

➀ This article provides a comprehensive hierarchy of GPU performance, ranking nearly every GPU released in the past seven years based on benchmarks.

➁ The article discusses the performance of GPUs across different resolutions and the impact of technologies like ray tracing and DLSS.

➂ It highlights the latest GPUs from Nvidia's Ada Lovelace and AMD's RDNA 3 architectures, as well as Intel's Arc GPUs.

➀ NVIDIA is reportedly using scarcity marketing for its GeForce RTX 50 series GPUs, creating an impression of shortage despite sufficient inventory.

➁ The real reason for the low availability of the GeForce RTX 50 series GPUs is that NVIDIA is engaging in 'scarcity marketing', controlling the supply and not releasing the GPUs.

➂ The GPUs are expected to be released in large quantities by April or May.

➀ A Reddit user has reported a GeForce RTX 5080 FE with fewer ROPs than expected;

➁ This issue is similar to the one previously reported on Zotac GeForce RTX 5090 Solid and GeForce RTX 5070 Ti;

➂ NVIDIA confirms a 'production anomaly' affecting a small number of cards, promising to replace affected units.

➀ Leaked benchmarks of the RTX 5070 show mixed results against the RTX 4070 Super, being up to 20% slower than the RTX 5070 Ti.

➁ The RTX 5070 is powered by a slightly cut-down GB205 die with 48 SMs and 12GB of GDDR7 memory.

➂ Despite its lower performance, the RTX 5070 still offers a 20% lead over the base RTX 4070.

➀ Nvidia's GeForce RTX 5080 graphics cards may have only 104 ROPs instead of the advertised 112;

➁ The issue is similar to the one found in the RTX 5090 and RTX 5070 Ti;

➂ Nvidia has confirmed a 0.5% defect rate in the affected GPUs.

➀ Nvidia has confirmed a manufacturing defect in the RTX 5070 Ti that reduces performance by up to 10%;

➁ The affected GPUs have 88 ROPs instead of the specified 96;

➂ Nvidia has asked customers to contact their board partners for a replacement.

➀ DeepSeek's R1 model sparked a major response in the tech industry, causing NVIDIA's stock to fall 16.79% and its market value to lose $90 billion, a record in American financial history.

➁ Huang Renxun, NVIDIA's CEO, praised DeepSeek as a major advancement in artificial intelligence but had not publicly responded to the issue until Thursday.

➂ Huang Renxun addressed the misunderstanding among investors about DeepSeek's progress in the AI field, which led to an incorrect reaction in the market to NVIDIA's stock.

➀ In February 2006, Nvidia experienced a significant financial surge with a 24% year-over-year revenue increase and nearly tripled income.

➁ The company reported revenues of $583.8m for the first quarter of fiscal 2006, surpassing the $471.9m from the previous year.

➂ Nvidia's gross margin improved significantly, positioning the company at the center of the digital media revolution.

➀ AMD's RX 9070 XT is up to 28% faster than the RX 7800 XT, but falls short of the RX 7900 XTX.

➁ Benchmark scores from Geekbench suggest it matches Nvidia's RTX 4070 Super in synthetic tests.

➂ While the RX 9070 XT shows a decent improvement over its predecessor, real-world performance remains to be seen.

➀ Silex Technology America, Inc. has launched the SX-PCEBE, the world's first embedded radio module supporting Wi-Fi 7 (802.11be) and Bluetooth 5.3.

➁ The module is equipped with Qualcomm’s QCC2076 SoC and supports dual-band simultaneous operation across 2.4GHz, 5GHz, and 6GHz.

➂ It is designed for AI-driven devices and works seamlessly on platforms like NVIDIA, Qualcomm, and NXP.