AMD Execution Thread [2022]

  • Thread starter Deleted member 2197
  • Start date
Status
Not open for further replies.
The prices of current AMD GPUs are crashing way harder than NVIDIA GPUs, seems the demand for AMD GPUs is much lower than the demand of RTX GPUs.

Earlier this month, AMD stated their Graphics devision suffered a decline in Q2, and will sustain another decline in Q3.




Should be interesting to see the Q2 marketshare numbers this quarter, will AMD fall below 17% (as per their Q1 share?) or manage to sustain/exceed it?

Probably exceed. Their quarters don't line up perfectly but NVs GeForce division crashed 44% (expected)last Q

Overall numbers are in. Shipments down 14.9 % compared last Q. AMD's dropped 7.6 %, Intels 9.8 % and NVIDIA whooping 25.7 %.
Marketshares now AMD 20 % (+1.1), Intel 62 % (+2.2), NVIDIA 18 % (-3.15)
 
Irrelevant numbers, wait for dGPU numbers, AMD's decline is cushioned by their iGPUs sold in Ryzen CPUs.

NV dropped 25.7%
Overall dGPU market dropped only 22.6%

That would imply that someone dropped less than 22.6% in order for the overall market to only drop 22.6% when NV as the dominant dGPU maker dropped 25.7%, unless NV makes a significant number of integrated GPUs.

Does NV make any integrated GPUs? I'm actually curious as I don't pay as much attention to all of the market sectors that they sell product into.

Regards,
SB
 
NV dropped 25.7%
Overall dGPU market dropped only 22.6%

That would imply that someone dropped less than 22.6% in order for the overall market to only drop 22.6% when NV as the dominant dGPU maker dropped 25.7%, unless NV makes a significant number of integrated GPUs.

Does NV make any integrated GPUs? I'm actually curious as I don't pay as much attention to all of the market sectors that they sell product into.

Regards,
SB
Nope, not to the markets JPR counted, for NVIDIA they're all discrete.
 
Irrelevant numbers, wait for dGPU numbers, AMD's decline is cushioned by their iGPUs sold in Ryzen CPUs.
Discrete GPU numbers should be interesting. As far as Total GPU market share both AMD and Nvidia have increased share (yoy) with no ranking changes.
As for market share, Intel remained the leading GPU supplier in Q2 2022 with 62% of the market, AMD came second with 20% (a significant increase from 16% in Q2 2021), whereas Nvidia was No. 3 with an 18% market share (up from 15% a year ago).
 
Something doesn't add up there. NVIDIAs shipments dropped 25.7% according to the other JPR report, and they literally make nothing but discrete products for the categories JPR counts. How could it be less now?
Again, AMD numbers are cushioned by their iGPUs, however dGPUs reveal the whole story: AMD dGPUs numbers dropped by 35% QtQ and was actually down 5% compared to last year, while NVIDIA dropped only by 19%, and is still up 3% compared to last year. So NVIDIA's drop is much lesser compared to AMD, which is reflected on the street price of their GPUs, which crashed harder than NVIDIA.
 
Again, AMD numbers are cushioned by their iGPUs, however dGPUs reveal the whole story: AMD dGPUs numbers dropped by 35% QtQ and was actually down 5% compared to last year, while NVIDIA dropped only by 19%, and is still up 3% compared to last year. So NVIDIA's drop is much lesser compared to AMD, which is reflected on the street price of their GPUs, which crashed harder than NVIDIA.
Again, if NVIDIA only produces discrete chips (in categories JPR counts) and their shipments drop 25.7%, how can it drop only 19% in discrete GPUs category which was the only one counted in the first place?
AMD numbers aren't relevant for that.
 
, if NVIDIA only produces discrete chips (in categories JPR counts) and their shipments drop 25.7%, how can it drop only 19% in discrete GPUs category which was the only one counted in the first place?
AMD numbers aren't relevant for that.
Who gives a damn about reports counting iGPUs numbers in the mix? these were always a sham to trump up Intel's non existing presence in the GPU market. These reports were always stacked full of irregularities and problems, with highly ambiguous methodoloy of counting the iGPUs inside, surplus of illogical swings, and basically no meaningful trends, .. the actual meat and bones has always been the report counting AIB numbers.

Even the AIB report is not 100% accurate, as it misses data center GPUs.
 
Last edited:
Who gives a damn about reports counting iGPUs numbers in the mix? these were always a sham to trump up Intel's non existing presence in the GPU market. These reports were always stacked full of irregularities and problems, with highly ambiguous methodoloy of counting the iGPUs inside, surplus of illogical swings, and basically no meaningful trends, .. the actual meat and bones has always been the report counting AIB numbers.

Even the AIB report is not 100% accurate, as it misses data center GPUs.
NVIDIA has no iGPUs (in categories counted by JPR), they can't affect their numbers.
 
Again, if NVIDIA only produces discrete chips (in categories JPR counts) and their shipments drop 25.7%, how can it drop only 19% in discrete GPUs category which was the only one counted in the first place?
AMD numbers aren't relevant for that.
Are you sure both articles are talking about the same thing? The second article talks about AIB shipments and “desktop discrete”. How do laptop GPUs fit in?
 
Are you sure both articles are talking about the same thing? The second article talks about AIB shipments and “desktop discrete”. How do laptop GPUs fit in?
Very first sentence of the article, bolding by me.
The Add-in Board report is a quarterly report that focuses on the market activity of PC graphics controllers for mobile and desktop computing.
 
Very first sentence of the article, bolding by me.

In that case the JPR articles are inconsistent or counting different things. Nvidia's shipments can't drop by both 26% and 19% at the same time. It must mean JPR is excluding some percentage of Nvidia's shipments from the AIB report.
 
NVIDIA has no iGPUs (in categories counted by JPR), they can't affect their numbers.
I just told you, iGPUs reports are never accurate.


Ignoring the fact that NVIDIA actually increased their market share this quarter, the breakdown of GPUs sold is more interesting.

Over a 10 months avg, AMD sold 33% of their GPUs from the 6000 series, 17% from the 5000 series (so not RT capable), while selling a whopping 50% from Vega and Polaris.

Comparatively, NVIDIA sold 43% from RTX 3000 series, 20% from RTX 2000, and 9% from GTX 1600, with the remaining 27% consisitng of Pascal and Maxwell (laptop GPUs I assume).

So more than 63% of NVIDIA GPUs sold are RT capable vs AMD which stands at only 33%, and more than 72% of NVIDIA GPUs sold are modern architectures (Ampere + Turing), vs AMD which stands at only 50%.
 
Nvidia's shipments can't drop by both 26% and 19% at the same time
They can, if the number of iGPUs sold were significantly higher this quarter.

This is what actually happened.

MW_PR_Q222.png



Let me give you a theoretical example, lets assume that 500k iGPUs + dGPUs were sold in Q1, 380k of them are iGPUs, 100k NVIDIA dGPUs and 20k AMD dGPUs (120k in total).

NVIDIA sets at 20% of the total pack, while maintaining a dominant 83% of the dGPUs market.

In Q2, it becomes 400k iGPUs + dGPUs, of them are 315k iGPUs, 72k are NVIDIA dGPUs and 13k are AMD dGPUs (85k in total).

So NVIDIA dropped in the total pack by 2%, with their shares dropping to 18%, while actually managing to increase their dGPUs shares to 85%, despite the overall drop.

You get the gest of it.
 
They can, if the number of iGPUs sold were significantly higher this quarter.

This is what actually happened.

MW_PR_Q222.png



Let me give you a theoretical example, lets assume that 500k iGPUs + dGPUs were sold in Q1, 380k of them are iGPUs, 100k NVIDIA dGPUs and 20k AMD dGPUs (120k in total).

NVIDIA sets at 20% of the total pack, while maintaining a dominant 83% of the dGPUs market.

In Q2, it becomes 400k iGPUs + dGPUs, of them are 315k iGPUs, 72k are NVIDIA dGPUs and 13k are AMD dGPUs (85k in total).

So NVIDIA dropped in the total pack by 2%, with their shares dropping to 18%, while actually managing to increase their dGPUs shares to 85%, despite the overall drop.

You get the gest of it.
No, they can't, because JPR doesn't count any category where NVIDIA makes anything resembling iGPU.
It's NVIDIAs shipments, that according to first report, dropped 25.7 % and according to second 19 %. When both only count discrete GPUs for NVIDIA, one has to be wrong.
 
They can, if the number of iGPUs sold were significantly higher this quarter.

This is what actually happened.

MW_PR_Q222.png



Let me give you a theoretical example, lets assume that 500k iGPUs + dGPUs were sold in Q1, 380k of them are iGPUs, 100k NVIDIA dGPUs and 20k AMD dGPUs (120k in total).

NVIDIA sets at 20% of the total pack, while maintaining a dominant 83% of the dGPUs market.

In Q2, it becomes 400k iGPUs + dGPUs, of them are 315k iGPUs, 72k are NVIDIA dGPUs and 13k are AMD dGPUs (85k in total).

So NVIDIA dropped in the total pack by 2%, with their shares dropping to 18%, while actually managing to increase their dGPUs shares to 85%, despite the overall drop.

You get the gest of it.

Yes they can drop total share while increasing discrete share. But it’s not possible for them to have two different numbers for the percentage drop in discrete shipments.
 
Status
Not open for further replies.
Back
Top