• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Intel has a Pretty Big Problem..


"After Ryzen 7 9800X3D with 6000+ Units Shipments, Ryzen 5 5500 is at Number Two, Outselling the Entire Core Ultra 200S Stack"

Several CPU models missing from their data like the 14700K which significantly out sells the 14900K (or was). I was always curious how poor the 14600K was doing though - a segment that normally sells well but AMD seems to have stolen a lot of that market and/or people still buying 12th gen instead.

EDIT: For the previous month they had 14700 sales at less than 5% of 14900 sales which is not reflected elsewhere, hence why you need a range of sources for these things.

Mindfactory numbers (keep in mind they sell heavily into an AMD market and tend to cater for that market):

14700K 6,680
14900K 4,300
14600K 4,260
12th/13th gen mid-range - 10,000-20,000 i.e. 13700K 11,240.

7800X3D 82,030
9800X3D 21,150
7700X 20,030
7950X 7,240
7600(X) 33,880

Average across several retailers (of varying sales volume) who sell into more "neutral" markets:

14600K 4,600
14700K 5,600
14900K 2,900
12th/13th gen mid range average ~7,000

7800X3D 11,000
9800X3D 3,200
7700X 3,200
AMD 7000/9000 mid range average ~5,000 (Going back to the 5000 series sales volumes are quite high - one retailer sold nearly 60,000 Ryzen 5600).

These numbers are a little deceptive though as across the entire product stack including pre-builds, etc. Intel still outsell AMD by quite a margin.
 
Last edited:
Overall market Intel is still massively outselling AMD though AMD is selling well in certain segments.

Actual failure rates at retail so far the 7000/9000 series have nearly twice the failure rate of Intel 13th/14th gen ironically - this "50%" failure BS isn't showing up at retail so far.
Where's these supposed numbers posted. Has Intel/AMD actually posted any for the respective gens?

As unless they have all this talk of either party having high returns is just pure speculation ! Any that say one or the other is worse is just guessing.
 
Where's these supposed numbers posted. Has Intel/AMD actually posted any for the respective gens?

As unless they have all this talk of either party having high returns is just pure speculation ! Any that say one or the other is worse is just guessing.

Neither Intel or AMD have reported returns, but there are a number of retailers (for example Puget Systems) who've released numbers and quite a few retailers like Mindfactory, alza.sk, etc. have return rates on the product pages. Personally I have access to a certain amount of returns data through work for several companies like VIP Computers, Zoostorm and so on but I'm not at liberty to reproduce that data.
 
Last edited:
On the subject of official numbers I can't find it from a quick Google and I refuse to make an account on the **** formerly known as Twitter but at some point Jaykihn posted screenshots supposedly from an Intel failure analysis report showing something like an expected 1 in 478 failure rate for the 14th gen though I might have been misunderstanding the data.

EDIT: One of the documents can't find the other one without a Twitter account.
 
Last edited:
As discussed 10x previously, it’s unlikely many large customers of Intel could provide failure data without breaching contract. We’ll probably never know the exact number, but it’s significant enough to have forced firms to switch vendor mid cycle, left Intel struggling to ship RMA replacements in any reasonable time and introduce a specific warranty outside of its usual supply chain.

Probably the best indicator would be look at Intels share prices a quarter or three, either side of the 13/14 and Cultra releases.
 
Probably the best indicator would be look at Intels share prices a quarter or three, either side of the 13/14 and Cultra releases.

fDhgA55.png


The share price is mostly reactive to news, they'd have to be bleeding billions due to failures for it to make a dent in terms of returns.

The major change in early 2024 is due to the publishing of Intel's financials, concerns over Intel misleading shareholders over their foundry business and layoffs, the next major change in July 2024 is reactive to the news about failures hitting the media along with another round of layoffs and poor results with its quarterly report.

I think it is probably a best indicator that Level1 Techs got played by disgruntled ex-Intel employees blowing up some small but not insignificant issues into a "big issue" (most of the failure mechanism details could only have come from people within Intel as there is no way to diagnose them to that extent from external testing) trying to get revenge on Intel. Meanwhile the rest of the tech media have mostly backed off the story after doing investigations - which either means Intel warned them off which is unlikely in the case of some like Gamer's Nexus or more likely they realised they were being played. One of the best indicators is that RAD Game Tools are still saying the issues affect a small number of Intel CPUs mostly 13900s and 14900s as they through the nature of their business are one of the most exposed to the issues.
 
Last edited:
fDhgA55.png


The share price is mostly reactive to news, they'd have to be bleeding billions due to failures for it to make a dent.

The major change in early 2024 is due to the publishing of Intel's financials, concerns over Intel misleading shareholders over their foundry business and layoffs, the next major change in July 2024 is reactive to the news about failures hitting the media along with another round of layoffs and poor results with its quarterly report.

Unlikely that’s the case. There are some non typical interesting features across that timeline. Overlaying AMDs results and then looking at its results 6-12 after Intel offending releases may shed some light.

Off the record claims are the failure rates are around 40-50% after 1500-3000 hours. The question is compounded because there seems to have two separate issues with one remaining constant and the other possibly being more prevalent in batches. I think we’ve been over this before.
 
Unlikely that’s the case. There are some non typical interesting features across that timeline. Overlaying AMDs results and then looking at its results 6-12 after Intel offending releases may shed some light.

There is absolutely nothing relevant there, AMD's results over the 2022-2024 time frame like Intel's mostly reflect overall market conditions, with 2024 buoyed by strong FY23 results and data centre performance - outside of market conditions AI/DC performance has been the driver of its share results both up and down.

Off the record claims are the failure rates are around 40-50% after 1500-3000 hours.

I'd say that is BS there are 100s companies using large numbers of these CPUs for things like 24x7 render farms (I know someone who works in this area and only seeing 3% failure rate) who will have 10+K hours on these CPUs by now and loads of small studios doing stuff like Alderon games are doing and there is a complete absence of reports of anything particularly abnormal failure wise - if these companies were seeing double digit failures generally it would be all over the tech press not just a couple of unsubstantiated instances reported by Level1 Techs.
 
Last edited:
There is absolutely nothing relevant there, AMD's results over the 2022-2024 time frame like Intel's mostly reflect overall market conditions, with 2024 buoyed by strong FY23 results and data centre performance - outside of market conditions AI/DC performance has been the driver of its share results both up and down.



I'd say that is BS there are 100s companies using large numbers of these CPUs for things like 24x7 render farms (I know someone who works in this area and only seeing 3% failure rate) who will have 10+K hours on these CPUs by now and loads of small studios doing stuff like Alderon games are doing and there is a complete absence of reports of anything particularly abnormal failure wise - if these companies were seeing double digit failures generally it would be all over the tech press not just a couple of unsubstantiated instances reported by Level1 Techs.


There is no way your friend has run these for 10,000 hours doing small studio stuff* 2000 hours a year would be a huge stretch.
 
There is no way your friend has run these for 10,000 hours doing small studio stuff* 2000 hours a year would be a huge stretch.

I have a friend who works in IT for one of the bigger advertising companies who uses 100s of 14th gen CPUs to render previews (as while for final renders Xeons and/or GPUs are faster, in production previews build quicker on 14900s) - these are in pretty much constant use.
 
Last edited:
I have a friend who works in IT for one of the bigger advertising companies who uses 100s of 14th gen CPUs to render previews (as while for final renders Xeons and/or GPUs are faster, in production previews build quicker on 14900s) - these are in pretty much constant use.

That’s mental. He’s should probably consider leaving.
 
As of the 13th of May he'll have been doing the job 20 years I don't think he is leaving any time soon :s

Well if his IT job is maintaining hundreds of 14900K systems he definitely has a few years of work ahead of him. Assuming the business lasts that long.

For context, 400 14900 servers would be around 2.5 megawatt hours per day, plus cooling, ancillary power and networking etc. The capex and opex numbers are horrendous.
 
Intel needs another new socket, meanwhile Zen6 is still AM5

 
Last edited:
Back
Top Bottom