Friday, June 27th 2025
Intel's Server Share Slips to 67% as AMD and Arm Widen the Gap
In just a few years, AMD has gone from the underdog to Intel's most serious challenger in the server world. Thanks to its EPYC processors, AMD now captures about a third of every dollar spent on server CPUs, up from essentially zero in 2017. Over that same period, Intel's share has slipped from nearly 100% to roughly 63%, signaling a significant shift in what companies choose to power their data centers. The real inflection point came with AMD's Zen architecture: by mid-2020, EPYC had already claimed more than 10% of server-CPU revenues. Meanwhile, Intel's rollout of Sapphire Rapids Xeons encountered delays and manufacturing issues, leaving customers to look elsewhere. By late 2022, AMD was over the 20% mark, and Intel found itself under 75% for the first time in years.
Looking ahead, analysts at IDC and Mercury Research, with data compiled by Bank of America, expect AMD's slice of the revenue pie to grow to about 36% by 2025, while Intel drops to around 55%. Arm-based server chips are also starting to make real inroads, forecast to account for roughly 9% of CPU revenue next year as major cloud providers seek more energy- and cost-efficient options. By 2027, AMD could approach a 40% revenue share, Intel may fall below half the market, and Arm designs could capture 10-12%. Remember that these figures track revenue rather than unit sales: AMD's gains come primarily from high-end, high-core-count processors, whereas Intel still shifts plenty of lower-priced models. With AMD poised to launch its Genoa and Bergamo EPYCs and Intel banking on the upcoming E-core Xeon 6 series to regain its footing, the fight for server-CPU supremacy is far from over. Still, Intel's once-unbeatable lead is clearly under threat.
Sources:
Jukan Choi on X, via HardwareLuxx
Looking ahead, analysts at IDC and Mercury Research, with data compiled by Bank of America, expect AMD's slice of the revenue pie to grow to about 36% by 2025, while Intel drops to around 55%. Arm-based server chips are also starting to make real inroads, forecast to account for roughly 9% of CPU revenue next year as major cloud providers seek more energy- and cost-efficient options. By 2027, AMD could approach a 40% revenue share, Intel may fall below half the market, and Arm designs could capture 10-12%. Remember that these figures track revenue rather than unit sales: AMD's gains come primarily from high-end, high-core-count processors, whereas Intel still shifts plenty of lower-priced models. With AMD poised to launch its Genoa and Bergamo EPYCs and Intel banking on the upcoming E-core Xeon 6 series to regain its footing, the fight for server-CPU supremacy is far from over. Still, Intel's once-unbeatable lead is clearly under threat.
36 Comments on Intel's Server Share Slips to 67% as AMD and Arm Widen the Gap
One really clear case where Epyc is not even an option is for 4S and 8S systems, AMD only scales up to 2S systems.
Intel finally released the 4S/8S Granite Ridge products up to 86 Cores recently. Here is the list of products on the wiki page:
en.wikipedia.org/wiki/Granite_Rapids
There wasn't much news regarding the remaining SKU launch but I believe they came out in February 2025. The 8S SKUs only go up to 86 cores unlike the 128 core 2S stuff that were released late last year.
I am not really sure how they will come back - they will need to bring in new people to turn it around.
They may not be as popular, and I never said those were the majority, but they are still relevant nonetheless.You can get prices way bellow that for any reasonably sized quote, fwiw. Same applies to AMD's public prices, ofc, but Intel is being way more aggressive with discounts (and the reduced revenue pointed in OP demonstrates that).
It's a testimony to how good the AMD chips are that they've gotten this far, I remember when intel had 95% of the market and nobody ever thought AMD would make headroom.
The company as a whole is still laying people off, left and right.
www.techpowerup.com/338319/intel-winds-down-automotive-division-prepares-for-major-layoffsNiche use case, at this point. Ain't gonna get Intel any % of market share.
But I'm not sure what you're trying to get at. AMD doesn't cover all use cases, but I've already said that their CPUs do make sense for most cases nonetheless.
Ironically, it was with Threadripper that AMD dropped the ball hard on, what with their indecision on supporting TR as thoroughly as Ryzen or EPYC until relatively recently. Now if only they'd offer some 12 and 16 core options in the premium consumer end, instead of only the prosumer end. Or heck, a new 8 core for those who want more I/O than cores.
256,255 views Jul 26, 2017
Intel is the world's most abusive, monopolistic tech company.
Now, AMD, apply this to your
GCNRDNAUDNA releases k?I remember them (Intel) predicting they might lose a total of 18% enterprise share to AMD. Oh boy how they miscalculated"Your telling me we can get an Intel powered offering from SI X for 75% of the cost for an equivalent AMD setup from SI Y? Oh and they are offering their 5 year support package at a discount? SI X is where we go." We all know that the EPYC options have been the superior option in most use cases for multiple generations but bean counters always have the final say on any purchase in companies. This is why there has been a slow decline for Intel over the last few years but its only been getting worse recently as the cost/performance arguments have been getting less and less competitive even with their heavy discounts.
Regarding 4 and 8S systems, they are very niche use case now and are mainly aimed at people who are limited by U space. So either the hyper dense (talking thousands of nodes in a single datacenter) or those being charged by U space being utilised by their hosting company. Otherwise it makes more sense to utilise more 2S systems for a multitude of reasons.
The B#50 series of boards could stand to have the same increased lane counts, just a PCIe generation behind save for the first PCIe slot and the first M.2 slot. Still would provide a good number of I/O regardless, and would be ideal for cheaper home servers that only need a 6 or 8 core and mostly I/O for NVMe or SATA raid cards.
Just gotta hope there was something already in the pipeline that is promising.