Intel is in a better position financially than AMD was, and could catch up again. I tend to think it depends a lot on the organization and the people.
IC: A few people consider you 'The Father of Zen', do you think you’d scribe to that position? Or should that go to somebody else?
JK: Perhaps one of the uncles. There were a lot of really great people on Zen. There was a methodology team that was worldwide, the SoC team was partly in Austin and partly in India, the floating-point cache was done in Colorado, the core execution front end was in Austin, the Arm front end was in Sunnyvale, and we had good technical leaders. I was in daily communication for a while with Suzanne Plummer and Steve Hale, who kind of built the front end of the Zen core, and the Colorado team. It was really good people. Mike Clark's a great architect, so we had a lot of fun, and success. Success has a lot of authors - failure has one. So that was a success. Then some teams stepped up - we moved Excavator to the Boston team, where they took over finishing the design and the physical stuff, Harry Fair and his guys did a great job on that. So there were some fairly stressful organizational changes that we did, going through that. The team all came together, so I think there was a lot of camaraderie in it. So I won't claim to be the ‘father’ - I was brought in, you know, as the instigator and the chief nudge, but part architect part transformational leader. That was fun.
If Intel only had a competitive advantage during that era because they traded performance for security, was there ever really an advantage?
I personally don't think Gelsinger is going to be able to fix their culture issues. He's been more intent on pumping stock and asking for handouts than innovating, which is part of what got them where they are in the first place, for good or ill.
Security is important, I would always prefer a secure product. But given design habits that elide security for performance, as well as a compromised supply chain, the only choice we have is to side with the devil(s) we know.
Even with longer fulfillment timelines, availability isn't necessary worth the surprise cost of competitiveness.
Lifecycling early and having to buy additional hardware to meet capacity was passed onto our customers.
After being ahead of Intel in the Pentium 4 era. The problem back then was marketing and giant amount of (probably illegal now) bundling that was rampant back then.
AMD had to make a much better product to get manufacturers to consider it, not just being price competitive. It took the P4 train wreck to get the market to shift but Intel had enough lock-in to make it to the Core generation without losing too much market share because so many of the vendors had those contracts.
Thunderbird was released in 1999 and from there to the Core 2 Duo release in 2006, AMD was the performance leader (certainly at a huge discount compared to intel offerings).
Their top performance was lower than Intel in many cases, but it was certainly not "absolute garbage". For low- to mid-range they offered similar performance with usually a (much) lower price point. For normal "average person" usage they were often the better choice in terms of "bang for your buck".
The main reason we were selling Intels is because we got a rebate from Intel for every CPU we sold. Financially it didn't really matter if we sold a €400 computer, €500, or €600 computer: the profit margin for us was roughly identical (a little bit more, but not much), but with Intel you got a rebate back-channelled, so then it did matter.
Intel can build ARMs and RISC-Vs as well as anybody. Harder will be adapting their salespeople to non-monopoly conditions.
The CPU is great. It’s power envelope and performance feel amazing.
My biggest complaint is the there is a very limited selection of Laptops with AMD’s chips, especially the top tier. The one I bought required me to replace the Wifi/Bluetooth card (it came with a barely “supported” MediaTek one) for Linux.
I have my gripes with MacOS after years of linux/windows usage, but responsiveness on M1 is not one of them.
I have turned off the genie effect and perhaps other animations, but my Ubuntu on my Lenovo still feels faster than the Mac.
It’s a more traditional style than a Unibody, sure. But it’s fine.
Lenovo's is the only one I find with 16" 3840x2400 OLED, but just a 6850, $2800. Or with 6950 but just 1920x1200, $2635. None with both. Advantage, Radeon R6500M GPU. Doesn't say which wifi it has.
https://www.lenovo.com/us/en/search?fq=&text=21D4000HUS
Dell has a 17" 6900HX, 4k, 32G RAM, RX6850 GPU, $2800. Or 6800H and RX6700, $2550. Mediatek MT7921 wifi...
https://www.dell.com/en-us/work/shop/dell-laptops/alienware-...
But I swore never to buy another Lenovo. Or Dell. Hoping I won't have to swear off ASUS too.
Build quality is excellent, but the first one died after 2 days: charge light wouldn't even go on. Waiting for #2. Wish me luck!
If they had tried to force the P4 mobiles instead of rethinking and retooling the PIII core with some of the P4's FSB innovations and other things, they probably wouldn't of had as competitive mobile processors and maybe wouldn't have dethroned AMD a few years later.
The Ryzen chips are already clocking over 3 GHz, so there isn't much more scaling from power left. That's why 5nm Zen4 probably won't move the needle too much.
Surprising results. Definitely not in the "very far" range.
Keep in mind the 6850U you're comparing against has a fan too. Notebookcheck says it has a 40W turbo on the gen 2 model, and the gen 3 Intel version has a 28W sustained boost. An M2 uses around 20W.
Also, a lot of those benchmarks are showing all x86 CPUs with a 3x+ lead, which indicates software optimization problems. Geekbench 5 is optimized for both arm64 and x86 and shows the M2 ahead of the 6800U. There are also just broken results like the LC0 test where the M1 is substantially faster than the M2. Overall, your results don't seem very valid.
> Overall, your results don't seem very valid.
They're not "my" results, and thermals were not part of the discussion.
Regarding performance, if one wants to be rigorous (and that's why I didn't state "68xx is faster than Mx"), one can't make absolute claims either way, as Torvalds himself has criticized Geekbench.
I still think that describing the performance difference as "not very far" is a valid assessment.