Files

153 lines
7.1 KiB
TypeScript

import React from 'react';
const hardware = [
{
category: 'LATEST: Blackwell Ultra',
name: 'NVIDIA B300 GPU',
description: 'Blackwell Ultra B300 GPU delivers the highest single-GPU AI performance with 288GB HBM3e memory, 12 TB/s bandwidth, and 2nd-gen Transformer Engine.',
imageUrl: '/nvidia-blackwell-ultra.jpg',
stats: [
{ value: '288 GB', label: 'HBM3e Memory' },
{ value: '12 TB/s', label: 'Memory Bandwidth' },
{ value: '~$5.99', label: '/GPU-hr Cloud' },
{ value: 'Blackwell Ultra', label: 'Architecture' },
{ value: '1800 GB/s', label: 'NVLink 5.0' },
{ value: '2nd Gen', label: 'Transformer Engine' },
]
},
{
category: 'Rack-Scale: Blackwell Architecture',
name: 'NVIDIA GB200 NVL72',
description: 'Rack-scale system with 8x B200 GPUs and 2x Grace CPUs. 1.4 EXAFLOPS FP4 performance with 1,536GB total HBM3e memory and direct liquid cooling.',
imageUrl: '/nvidia-blackwell-ultra.jpg',
stats: [
{ value: '1,536 GB', label: 'Total HBM3e' },
{ value: '1.4 EX', label: 'FLOPS FP4' },
{ value: '8x B200', label: '+ 2x Grace' },
{ value: 'From $399K', label: 'Purchase' },
{ value: '$29,999', label: '/mo Lease' },
{ value: 'Liquid', label: 'Cooled' },
]
},
{
category: 'Blackwell Architecture',
name: 'NVIDIA B200 GPU',
description: 'The Blackwell B200 GPU delivers exceptional AI performance with 192GB HBM3e memory and 2nd-gen Transformer Engine for next-generation AI workloads.',
imageUrl: '/nvidia-blackwell-ultra.jpg',
stats: [
{ value: '192 GB', label: 'HBM3e Memory' },
{ value: '8 TB/s', label: 'Memory Bandwidth' },
{ value: '9 PFLOPS', label: 'FP4 AI Performance' },
{ value: 'Blackwell', label: 'Architecture' },
{ value: '1800 GB/s', label: 'NVLink 5.0' },
{ value: '2nd Gen', label: 'Transformer Engine' },
]
},
{
category: 'Blackwell Architecture',
name: 'NVIDIA B100 GPU',
description: 'Blackwell B100 GPU combines exceptional AI performance with 192GB memory, perfect for training and inference of the largest AI models.',
imageUrl: '/nvidia-dgx-spark-and-dgx-station.jpg',
stats: [
{ value: '192 GB', label: 'HBM3e Memory' },
{ value: '8 TB/s', label: 'Memory Bandwidth' },
{ value: '7 PFLOPS', label: 'FP4 AI Performance' },
{ value: 'Blackwell', label: 'Architecture' },
{ value: '1800 GB/s', label: 'NVLink 5.0' },
{ value: 'PCIe Gen6', label: 'System Interface' },
]
},
{
category: 'Hopper Architecture',
name: 'NVIDIA H200 GPU',
description: 'The H200 offers HBM3e with 141GB of memory at 4.8 terabytes per second to handle massive datasets for generative AI and HPC.',
imageUrl: '/h200-nvl.jpg',
stats: [
{ value: '141 GB', label: 'HBM3e Memory' },
{ value: '4.8 TB/s', label: 'Memory Bandwidth' },
{ value: '97 TFLOPS', label: 'FP64 Performance' },
{ value: '4 PFLOPS', label: 'FP8 AI Performance' },
{ value: 'Hopper', label: 'Architecture' },
{ value: '900 GB/s', label: 'NVLink C2C' },
]
},
{
category: 'Hopper Architecture',
name: 'NVIDIA H100 GPU',
description: 'The NVIDIA H100 Tensor Core GPU delivers exceptional performance, scalability, and security for data centers, accelerating workloads from enterprise AI to HPC.',
imageUrl: '/nvidia-dgx-spark-and-dgx-station.jpg',
stats: [
{ value: '80 GB', label: 'HBM3 Memory' },
{ value: '3.35 TB/s', label: 'Memory Bandwidth' },
{ value: '67 TFLOPS', label: 'FP64 Performance' },
{ value: '2 PFLOPS', label: 'FP8 AI Performance' },
{ value: 'Hopper', label: 'Architecture' },
{ value: 'PCIe Gen5', label: 'System Interface' },
]
},
{
category: 'COMING 2027: Vera Rubin Architecture',
name: 'NVIDIA R-Series (Vera Rubin)',
description: 'Next-generation Vera Rubin architecture GPUs are expected in 2027. Contact us to join the waitlist for early access when available.',
imageUrl: '/nvidia-dgx-spark-and-dgx-station.jpg',
stats: [
{ value: 'R-Series', label: 'Architecture' },
{ value: '2027', label: 'Expected' },
{ value: 'HBM4', label: 'Memory Type' },
{ value: 'Next Gen', label: 'NVLink' },
{ value: 'TBD', label: 'Performance' },
{ value: 'Waitlist', label: 'Status' },
]
},
];
const Stat: React.FC<{ value: string; label: string }> = ({ value, label }) => (
<div className="text-center">
<p className="text-2xl md:text-3xl font-bold text-primary">{value}</p>
<p className="text-gray-400 text-sm mt-1">{label}</p>
</div>
);
const HardwareCard: React.FC<{ item: typeof hardware[0], onSelectProduct: (name: string) => void }> = ({ item, onSelectProduct }) => (
<div className="bg-dark-card border border-dark-border rounded-xl overflow-hidden flex flex-col">
<div className="bg-black p-4">
<img src={item.imageUrl} alt={item.name} className="w-full h-80 object-contain aspect-video" />
</div>
<div className="p-8 flex-grow flex flex-col">
<span className="text-primary font-bold uppercase tracking-widest text-sm">{item.category}</span>
<h3 className="text-3xl font-bold text-white mt-2 mb-4">{item.name}</h3>
<p className="text-gray-400 mb-6 flex-grow">{item.description}</p>
<div className="grid grid-cols-2 sm:grid-cols-3 gap-6 pt-6 border-t border-dark-border">
{item.stats.map(stat => <Stat key={stat.label} {...stat} />)}
</div>
<div className="mt-8">
<button
onClick={() => onSelectProduct(item.name)}
className="w-full bg-primary text-black font-bold py-3 px-6 rounded-lg hover:bg-primary-dark transition-all duration-300 transform hover:scale-105"
>
Request Access
</button>
</div>
</div>
</div>
);
const HardwareSpec: React.FC<{ onSelectProduct: (name: string) => void }> = ({ onSelectProduct }) => {
return (
<section id="hardware" className="py-20 md:py-28 bg-dark-bg">
<div className="container mx-auto px-6">
<div className="text-center mb-16">
<h2 className="text-3xl md:text-5xl font-bold text-white">Explore Our Hardware Fleet</h2>
<p className="mt-4 text-lg text-gray-400 max-w-2xl mx-auto">Access the most powerful and sought-after AI accelerators on the market.</p>
</div>
<div className="grid grid-cols-1 lg:grid-cols-1 gap-8 max-w-6xl mx-auto">
{hardware.map(item => <HardwareCard key={item.name} item={item} onSelectProduct={onSelectProduct} />)}
</div>
</div>
</section>
);
};
export default HardwareSpec;