Intel's ambitious plan to challenge AMD and NVIDIA in the AI accelerator market may still be a little questionable, according to recent comments from interim co-CEO Michelle Johnston Holthaus at the Barclays 22nd Annual Global Technology Conference. The company's "Falcon Shores" project, which aims to merge Gaudi AI capabilities with Intel's data center GPU technology for HPC workloads, received surprising commentary from Holthaus. "We really need to think about how we go from Gaudi to our first generation of Falcon Shores, which is a GPU," she stated, before acknowledging potential limitations. "And I'll tell you right now, is it going to be wonderful? No, but it is a good first step."

Intel's pragmatic approach to AI hardware development was further highlighted when Holthaus addressed the company's product strategy. Rather than completely overhauling their development pipeline, she emphasized the value of iterative progress: "If you just stop everything and you go back to doing like all new product, products take a really long time to come to market. And so, you know, you're two years to three years out from having something." The co-CEO advocated for a more agile approach, stating, "I'd rather have something that I can do in smaller volume, learn, iterate, and get better so that we can get there." She acknowledged the enduring nature of AI market opportunities, particularly noting the current focus on training while highlighting the potential in other areas: "Obviously, AI is not going away. Obviously training is, you know, the focus today, but there's inference opportunities in other places where there will be different needs from a hardware perspective."

Intel has prepared "Jaguar Shores," its "next-next" generation AI and HPC accelerator, successor to its upcoming "Falcon Shores" GPU. Revealed during a technical workshop at the SC2024 conference, the chip was unveiled by Intel's Habana Labs division, albeit unintentionally. This announcement positions Jaguar Shores as the successor to Falcon Shores, which is scheduled to launch next year. While details about Jaguar Shores remain sparse, its designation suggests it could be a general-purpose GPU (GPGPU) aimed at both AI training, inferencing, and HPC tasks. Intel's strategy aligns with its push to incorporate advanced manufacturing nodes, such as the 18A process featuring RibbonFET and backside power delivery, which promise significant efficiency gains, so we can expect to see upcoming AI accelerators incorporating these technologies.

Intel's AI chip lineup has faced numerous challenges, including shifting plans for Falcon Shores, which has transitioned from a CPU-GPU hybrid to a standalone GPU, and cancellation of Ponte Vecchio. Despite financial constraints and job cuts, Intel has maintained its focus on developing cutting-edge AI solutions. "We continuously evaluate our roadmap to ensure it aligns with the evolving needs of our customers. While we don't have any new updates to share, we are committed to providing superior enterprise AI solutions across our CPU and accelerator/GPU portfolio." an Intel spokesperson stated. The announcement of Jaguar Shores shows Intel's determination to remain competitive. However, the company faces steep competition. NVIDIA and AMD continue to set benchmarks with performant designs, while Intel has struggled to capture a significant share of the AI training market. The company's Gaudi lineup ends with third generation, and Gaudi IP will get integrated into Falcon Shores.

According to Taiwanese media Commercial Times, Intel is significantly increasing its outsourcing of "Arrow Lake" CPU production to TSMC, a strategic move as it grapples with persistent issues in its own foundry division. This decision to outsource a substantial portion of Arrow Lake's production is a significant shift in Intel's strategy, showing the company's rising reliance on external partners to meet quality and performance demands. The Arrow Lake Core Ultra 200 series is Intel's first major outsourcing initiative, in which Intel gave its core IP to third-party foundries, more specifically for a 3 nm node at TSMC. However, it clearly indicates the performance gaps in Intel's own Intel Foundry and the high demand expectations for the new CPUs. Originally intended to use Intel 20A node, Intel shifted focus of 18A node for its products and upcoming foundry customers.

Intel's recent orders with TSMC extend to its upcoming Lunar Lake chips and next-generation Falcon Shores AI GPUs, both of which will use TSMC's 3 nm process. Although Intel's 18A node remains promising, the company relies on current products to sustain its revenue streams, making TSMC's support crucial in ensuring timely shipments. This increased outsourcing reflects Intel's need to maintain competitive performance in the short term. Once its Foundry division meets performance and capacity targets, Intel aims to bring more CPU manufacturing back in-house. However, if anything goes wrong, Intel could face challenges securing sufficient volume from TSMC, as the foundry has longstanding commitments with major clients like Apple, NVIDIA, Qualcomm, and AMD.

Intel's upcoming Falcon Shores GPU is shaping up to be a powerhouse for AI and high-performance computing (HPC) workloads, but it will also be an extreme power hog. The processor, combining Gaudi and Ponte Vecchio successors into a single GPU, is expected to consume an astonishing 1500 W of power - more than even Nvidia's beefy B200 accelerator, which draws 1000 W. This immense power consumption will require advanced cooling solutions to ensure the Falcon Shores GPU operates efficiently and safely. Intel's partners may turn to liquid cooling or even full immersion liquid cooling, a technology Intel has been promoting for power-hungry data center hardware. The high power draw is the cost of the Falcon Shores GPU's formidable performance promises. Intel claims it will deliver 5x higher performance per watt and 5x more memory capacity and bandwidth compared to its Ponte Vecchio products.

Intel may need to develop proprietary hardware modules or a new Open Accelerator Module (OAM) spec to support such extreme power levels, as the current OAM 2.0 tops out around 1000 W. Slated for release in 2025, the Falcon Shores GPU will be Intel's GPU IP based on its next-gen Xe graphics architecture. It aims to be a major player in the AI accelerator market, backed by Intel's robust oneAPI software development ecosystem. While the 1500 W power consumption is sure to raise eyebrows, Intel is betting that the Falcon Shores GPU's supposedly impressive performance will make it an enticing option for AI and HPC customers willing to invest in robust cooling infrastructure. The ultra-high-end accelerator market is heating up, and the HPC accelerator market needs a Ponte Vecchio successor.

According to ServeTheHome, Intel has decided to discontinue its high-performance computing (HPC) product line, Ponte Vecchio, and shift its focus towards developing its next-generation data center GPU, codenamed Falcon Shores. This decision comes as Intel aims to streamline its operations and concentrate its resources on the most promising and competitive offerings. The Ponte Vecchio GPU, released in January of 2023, was intended to be Intel's flagship product for the HPC market, competing against the likes of NVIDIA's H100 and AMD's Instinct MI series. However, despite its impressive specifications and features, Ponte Vecchio faced significant delays and challenges in its development and production cycle. Intel's decision to abandon Ponte Vecchio is pragmatic, recognizing the intense competition and rapidly evolving landscape of the data center GPU market.

By pivoting its attention to Falcon Shores, Intel aims to deliver a more competitive and cutting-edge solution that can effectively challenge the dominance of its rivals. Falcon Shores, slated for release in 2025, is expected to leverage Intel's latest process node and architectural innovations. Currently, Intel has Gaudi 2 and Gaudi 3 accelerators for AI. However, the HPC segment is left without a clear leader in the company's product offerings. Intel's Ponte Vecchio is powering Aurora exascale supercomputer, which is the latest submission to the TOP500 supercomputer lists. This is also coming after the Rialto Bridge cancellation, which was supposed to be an HPC-focused card. In the future, the company will focus only on the Falcon Shores accelerator, which will unify HPC and AI needs for high-precision FP64 and lower-precision FP16/INT8.

NVIDIA is anticipating supply issues for its upcoming Blackwell GPUs, which are expected to significantly improve artificial intelligence compute performance. "We expect our next-generation products to be supply constrained as demand far exceeds supply," said Colette Kress, NVIDIA's chief financial officer, during a recent earnings call. This prediction of scarcity comes just days after an analyst noted much shorter lead times for NVIDIA's current flagship Hopper-based H100 GPUs tailored to AI and high-performance computing. The eagerly anticipated Blackwell architecture and B100 GPUs built on it promise major leaps in capability—likely spurring NVIDIA's existing customers to place pre-orders already. With skyrocketing demand in the red-hot AI compute market, NVIDIA appears poised to capitalize on the insatiable appetite for ever-greater processing power.

However, the scarcity of NVIDIA's products may present an excellent opportunity for significant rivals like AMD and Intel. If both companies can offer a product that could beat NVIDIA's current H100 and provide a suitable software stack, customers would be willing to jump to their offerings and not wait many months for the anticipated high lead times. Intel is preparing the next-generation Gaudi 3 and working on the Falcon Shores accelerator for AI and HPC. AMD is shipping its Instinct MI300 accelerator, a highly competitive product, while already working on the MI400 generation. It remains to be seen if AI companies will begin the adoption of non-NVIDIA hardware or if they will remain a loyal customer and agree to the higher lead times of the new Blackwell generation. However, capacity constrain should only be a problem at launch, where the availability should improve from quarter to quarter. As TSMC improves CoWoS packaging capacity and 3 nm production, NVIDIA's allocation of the 3 nm wafers will likely improve over time as the company moves its priority from H100 to B100.

During the ISC High Performance 2023 international conference, Intel announced interesting roadmap updates to its high-performance computing (HPC) and artificial intelligence (AI). With the scrapping of Rialto Bridge and Lancaster Sound, Intel merged these accelerator lines into Falcon Shores processor for HPC and AI, initially claiming to be a CPU+GPU solution on a single package. However, during the ISC 2023 talk, the company forced a change of plans, and now, Falcon Shores is GPU only solution destined for a 2025 launch. Originally, Intel wanted to combine x86-64 cores with Xe GPU to form an "XPU" module that powers HPC and AI workloads. However, Intel did not see a point in forcing customers to choose between specific CPU-to-GPU core ratios that would need to be in an XPU accelerator. Instead, a regular GPU solution paired with a separate CPU is the choice of Intel for now. In the future, as workloads get more defined, XPU solutions are still a possibility, just delayed from what was originally intended.

Regarding Intel's Gaudi accelerators, the story is about to end. The company originally paid two billion US Dollars for Habana Labs and its Gaudi hardware. However, Intel now plans to stop the Gaudi development as a standalone accelerator and instead use the IP to integrate it into its Falcon Shores GPU. Using modular, tile-based architecture, the Falcon Shores GPU features standard ethernet switching, up to 288 GB of HBM3 running at 9.8 TB/s throughput, I/O optimized for scaling, and support for FP8 and FP16 floating point precision needed for AI and other workloads. As noted, the creation of XPU was premature, and now, the initial Falcon Shores GPU will become an accelerator for HPC, AI, and a mix of both, depending on a specific application. You can see the roadmap below for more information.