Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallApollo Global Management President Jim Zelter has warned that compute and energy could constrain the pace of AI expansion. That is a reported executive view, not a measured forecast: the available account of his October 6, 2026 Bloomberg interview does not include the full interview or a detailed explanation of his reasoning. Independent evidence does show rapidly rising data-center electricity use—and constraints that go beyond how much power is generated.
What Zelter’s reported warning does—and does not—establish
A short synopsis of Bloomberg’s October 6, 2026 item says Zelter discussed compute and energy bottlenecks as AI investment accelerates. It also identifies topics including Apollo’s AI and data-center exposure, the US economy, private capital in defense, and European investment. Without the full Bloomberg article or transcript, those descriptions do not establish a detailed Apollo investment thesis, specific portfolio positions, or a quantified forecast.
As an Amazon Associate I earn from qualifying purchases.
The underlying question is practical: can the supply of computing equipment, data-center capacity, and dependable electricity grow quickly enough to serve expanding AI workloads? Evidence from the International Energy Agency (IEA) indicates that demand is rising, but it does not show that every region will face the same constraint or that a shortage is inevitable.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsWhat the electricity figures say
The IEA estimated that data centers consumed about 415 terawatt-hours (TWh) of electricity in 2024—around 1.5% of global electricity use. Its 2025 report projected about 945 TWh in 2030 under its base case. That is a scenario, not a guaranteed outcome; the agency discusses uncertainty and uses multiple scenarios.
#1 Best Overall
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
More recently, the IEA reported that electricity demand from data centers grew 17% in 2025, faster than overall global electricity demand. Demand from AI-focused data centers climbed faster still. These figures establish a strong growth trend, not a uniform local grid shortage: the impact depends on where facilities are built and what generation, transmission, and connection capacity is available there.
Where a bottleneck can form
| Potential constraint | What it means |
|---|---|
| Compute equipment | Accelerators and servers must be available and installed. Having hardware on order is not the same as having operating capacity. |
| Electricity generation | New data centers need enough power to run their equipment. More generation alone may not solve a site-level delivery problem. |
| Grid connection and delivery | Transmission capacity and interconnection determine whether electricity can reach a facility. The IEA identifies delayed grid connections as a constraint on near-term expansion. |
| Equipment and approvals | Tight supply chains for items such as transformers and gas turbines, alongside planning and regulatory approvals, can slow projects even when demand and financing are present. |
| Data-center construction | Servers need facilities that can be built, connected, and operated. The IEA’s projected growth should therefore be read alongside infrastructure and delivery timelines, not as a simple count of machines. |
These dependencies explain why “energy” is not just a question of how much electricity a country produces. A facility needs power delivered reliably at the right place and time, and its equipment and connection must be ready as well.
Rank #2
- Delivers 25 TFLOPS of AI inference performance using four MemryX MX3 accelerators in a compact M.2 2280 M-key module.
- Connects through PCI Express Gen 3 and above for straightforward integration into embedded and edge systems with an M.2 M-key slot.
- Includes 42 MB of distributed on-chip weight storage to support low-latency, batch-1 pipelined inference.
- Designed for low-power deployment with a 1 W–12 W operating range and host-managed airflow requirements.
- Supports ONNX, TensorFlow, Keras, and TFLite models through the MemryX software stack on Linux and Windows hosts.
Why efficiency does not settle the demand question
The IEA says electricity use per AI task is declining rapidly as efficiency improves. But lower energy use for an individual task does not by itself determine total consumption: wider AI adoption and more energy-intensive applications can increase the number and type of tasks being run. The relevant comparison is energy per task against the total volume and mix of workloads.
AI also creates unusually demanding operating conditions for power systems. The IEA’s 2026 executive summary describes rising AI-server rack power density and large, rapid swings in power during AI training and use. Those characteristics make reliable supply and storage relevant; a projected peak for a rack should not be mistaken for typical consumption across all server racks.
Rank #3
- Professional AI & Creator Workstation: AMD Radeon AI PRO R9700 GPU with 32GB GDDR6 is engineered for AI development, professional content creation, and compute-intensive workloads.
- Massive 32GB Memory Capacity: 32GB of GDDR6 memory on a 256-bit bus provides ample bandwidth for large AI models, 8K video editing, and complex 3D rendering.
- Advanced RDNA 4 with AI Accelerators: 64 Compute Units with 3rd Gen Ray Tracing and dedicated 2nd Gen AI Accelerators for groundbreaking AI performance and visual computing.
- Professional Blower Cooling: Efficient single blower design exhausts heat directly out of the chassis, ideal for multi-GPU workstation and server configurations.
- Enterprise-Grade Thermal Solution: Vapor chamber heatsink with industrial Honeywell PTM7950 thermal interface material ensures reliable cooling under sustained professional loads.
How Apollo has framed the issue
In a June 2026 mid-year outlook, Apollo framed compute around three variables: demand, price, and supply. The discussion asks whether energy and data-center capacity can keep pace with demand, and notes that supply and demand each have multiple dimensions. That provides context for why infrastructure is part of the conversation, but it is separate from the October Bloomberg report and should not be treated as confirmation of Zelter’s specific reasoning in that interview.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What to watch to judge whether constraints are worsening
- Equipment delivery: whether accelerators, servers, transformers, and other critical equipment can be delivered on project schedules.
- Time to connect: whether proposed data centers can secure grid interconnections and dependable power at their chosen sites.
- Local capacity: whether generation and transmission can serve concentrated new loads, rather than relying only on global electricity totals.
- Workload growth: whether rising AI use offsets reductions in energy per task.
- Project execution: whether planning, permitting, construction, and infrastructure upgrades keep pace with investment plans.
The IEA’s data support treating compute and energy as credible constraints to monitor, not as proof that AI growth will hit a fixed ceiling. The outcome will depend on equipment supply, local infrastructure, project timing, and how quickly AI demand and efficiency evolve.
Quick Recap
Best Value
- Memory Size: 16 GB GDDR6 ECC.
- Memory Bus Width: 128-bit.
- Memory Bandwidth: 200 GB/s.
- CUDA Cores: 1280.
- Peak Single Precision floating point performance: 18 Tflops (GPU Boost Clocks).
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
