Dive Brief:
- Chipmaker Advanced Micro Devices saw revenues plummet 18% YoY in Q2, down to $5.4 billion.
- The fall was driven by a steep drop in client segment demand, which includes products like computer processing units and PC chipsets. Revenue in the segment fell 54% YoY to just under $1 billion.
- AMD is hoping to harness the growing demand for AI chips to drive revenue growth in coming quarters, including in its datacenter segment, which it expects to ramp in the second half of the year, CEO and Chair Lisa Su said on a Q2 earnings call last week.
Dive Insight:
AMD's business has been struggling in recent quarters amid stagnant PC demand, with revenue flat from Q1 to Q2 this year.
The chipmaker's client segment actually performed better this past quarter, up from a 65% YoY revenue drop in Q1. The company blamed the decline in Q1 to its decision to ship orders at levels "significantly below consumption to reduce downstream inventory,” Su said on a May 2 earnings call.
While the company waits for a hopeful rebound in the PC market, AMD is turning its attention to manufacturing chips for AI needs. The company is increasing its AI-related R&D investments and has been working rapidly to launch new AI chips to rival competitor Nvidia.
AMD announced its new generative AI chip, the MI300X, earlier this summer, and is now sampling the product with some customers now. It plans to launch and begin ramping production of the chip in Q4, Su said on the Aug. 1 call.
The CEO noted that the company has been focused on building out the MI300X's supply chain "for quite some time," to combat constrained supplies in the industry.
"There's no question that it's tight in the industry," Su said. "However, we have sort of commitments for significant capacity across the entire supply chain."
Like AMD, Nvidia is also prioritizing its AI chip manufacturing. Nvidia, which will release its Q2 results later this month, has announced its own new generative AI chip offerings in recent weeks. The chips are particularly for use in data centers, which Nvidia EVP and CFO Colette Kress said are seeing "exponential growth" during a Q1 earnings call.
The rise of AI-focused chips are the beginning of a "10-year transition to basically recycle or reclaim the world's data centers and build it out as accelerated computing," Co-founder, CEO and President Jensen Huang said on the call. "You'll have a pretty dramatic shift in the spend of the data center from traditional computing, and to accelerated computing with smart NICs, smart switches, of course, GPUs, and the workload is going to be predominantly generative AI.”
