• 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.

*** The AMD RDNA 4 Rumour Mill ***

humbug said:
Why would it have 800 Watts power draw? The 7900 XTX is less than half of that. If the 5090 is 50% faster than the 4090 no one is going to say that's impossible it would have 700 watts power draw.

The rumour is it got cancelled because it would be too expensive.
Click to expand...


Too expensive? Sounds like bs. People have been paying $2k for a 4090
 
Last edited:
"Unfortunately, though, we’re not expecting to see such a GPU or card ever appear, as it’s rumored that AMD hasn’t been able to make its new chiplet technology scale to this level."

So its not being released because AMD have tried to make it, but haven't been able to get it to work yet; so price not the reason (or at least not the sole one if it could have been made to work by spending more on it than they would get back).
 
i reckon amd too will be eventually embracing a superscalar architecture like nvidia, all basic functional units bunched together, its been a great success for nvidia, not only have they utterly destroyed amd in raw performance but are currently 1.5 generations ahead of them in efficiency
 
Last edited:
brostradamus said:
i reckon amd too will be eventually embracing a superscalar architecture like nvidia, all basic functional units bunched together, its been a great success for nvidia, not only have they utterly destroyed amd in raw performance but are currently 1.5 generations ahead of them in efficiency
Click to expand...

Supescaler is simply executing multiple instructions per clock cycle, or what we now call and measure as IPC, this was cracked in the 1970's along with the move away from RISC to programable logic.
For decades everything has been Superscaler.

Also, the 7900 XTX despite being on 5nm vs the 4090 4nm is 85% the performance at 85% the power consumption.
 
Last edited:
humbug said:
Supescaler is simply executing multiple instructions per clock cycle, or what we now call and measure as IPC, this was cracked in the 1970's along with the move away from RISC to programable logic.
For decades everything has been Superscaler.
Click to expand...

amd has vector units, nvidia has scalar bunched up functional units called cuda/tensor/rt cores etc... amd has to package instructions using a more complicated scheduler to maintain high utilisation of their vector units, nvidia just has to segregate instructions by the type of operation and then push it in the pipeline..

a very simplistic way to describe both architectures could be in terms of lets say visa application:
amd: runs a single window to process a single visa application. to process an application documents A, B, C will be processed by a single window per applicant
nvidia: maintains separate dedicated windows for processing documents A, B, C.. everyone submits their documents in a common queue, the documents are then separated and reorganized in smaller downstream queues specific to a given document type (separate queues for A, B, C and so on)


humbug said:
Also, the 7900 XTX despite being on 5nm vs the 4090 4nm is 85% the performance at 85% the power consumption.
Click to expand...

perf numbers below for timespy extreme
rtx 4090: 19261
7900 xtx: 14688
and rtx 4090 doesnt touch max rated TDPs

and the performance gulf will be even greater for RT apps
 
Last edited:
brostradamus said:
amd has vector units, nvidia has scalar bunched up functional units called cuda/tensor/rt cores etc... amd has to package instructions using a more complicated scheduler to maintain high utilisation of their vector units, nvidia just has to segregate instructions by the type of operation and then push it in the pipeline..

a very simplistic way to describe both architectures could be in terms of lets say visa application:
amd: runs a single window to process a single visa application. to process an application documents A, B, C will be processed by a single window per applicant
nvidia: maintains separate dedicated windows for processing documents A, B, C.. everybody, submits their documents in a common queue, the documents are then separated and processed by dedicated windows which specialise in processing a specific document




perf numbers below for timespy extreme
rtx 4090: 19261
7900 xtx: 14688
and rtx 4090 doesnt touch max rated TDPs

and the performance gulf will be even greater for RT apps
Click to expand...

None of this has anything to do with Superscaling.

8FP.

H100: 3.96 TeraFlops
MI 300X: 5.23 TeraFlops

You're cherry picking a synthetic benchmark.

Multiple Gamers overall performance 4K.

RX 7900 XTX: 100%
RTX 4090 : 122%

Power consumption in gaming.

7900 XTX: 356 watts
RTX 4090: 411 watts (116%)

They are near enough the same performance per watt, again despite the 4090 being on a better lithography node, TSMC 4nm vs TSNC 5nm.

www.techpowerup.com

AMD Radeon RX 7900 XTX Review - Disrupting the GeForce RTX 4080

Navi 31 is here! The new $999 Radeon RX 7900 XTX in this review is AMD's new flagship card based on the wonderful chiplet technology that made the Ryzen Effect possible. In our testing we can confirm that the new RX 7900 XTX is indeed faster than the GeForce RTX 4080, but only with RT disabled.
www.techpowerup.com www.techpowerup.com
 
Last edited:
humbug said:
None of this has anything to do with Superscaling.
Click to expand...
thats how scalar-ness is interpreted, nvidia has a scalar architecture, amd uses a vector architecture, in a relative sense
humbug said:
H100: 3.96 TeraFlops
MI 300X: 5.23 TeraFlops
Click to expand...
this is misleading because:
a. these are theoretical max values, only works for ideal load
b. you havent mentioned the data type which makes those numbers meaningless (fp64/fp32/fp16/tf32/int8.. what?)
c. tdp's are missing: mi300x 750 watt, h100 350w
www.techpowerup.com

NVIDIA H100 PCIe 80 GB Specs

NVIDIA GH100, 1755 MHz, 14592 Cores, 456 TMUs, 24 ROPs, 81920 MB HBM2e, 1593 MHz, 5120 bit
www.techpowerup.com www.techpowerup.com
www.techpowerup.com

AMD Radeon Instinct MI300X Specs

AMD Aqua Vanjaram, 2100 MHz, 19456 Cores, 1216 TMUs, 0 ROPs, 196608 MB HBM3, 2525 MHz, 8192 bit
www.techpowerup.com www.techpowerup.com
oh yeah and they are not from the same generation either going by launch dates, maybe you are unaware but nvidia has already launched b200

humbug said:
You're cherry picking a synthetic benchmark.
Click to expand...
its the defacto dx12 benchmark
and its hypocritical at the same time, because you have conveniently skipped RT benchmarks
we are talking about the full extent of capabilities of both chips, so it doesnt make much sense to skip RT performance

here data on port royal:
7900 xtx: 15793
rtx 4090: 25692 (+63%)
www.eteknix.com

AMD Radeon RX 7900 XT & 7900 XTX Review - Page 8

AMD is finally ready to throw some fresh hardware into the endless battle between Nvidia and AMD for gaming dominance, sure Intel is part of the rumble these days too, but it’s fair to say they’ve got…
www.eteknix.com

humbug said:
TSMC 4nm vs TSNC 5nm.
Click to expand...
and tsmc 4n is just a custom 5n node for nvidia as has been reported by popular press outlets
 
Last edited:
brostradamus said:
thats how scalar-ness is interpreted, nvidia has a scalar architecture, amd uses a vector architecture, in a relative sense

this is misleading because:
a. these are theoretical max values, only works for ideal load
b. you havent mentioned the data type which makes those numbers meaningless (fp64/fp32/fp16/tf32/int8.. what?)
c. tdp's are missing: mi300x 750 watt, h100 350w
www.techpowerup.com

NVIDIA H100 PCIe 80 GB Specs

NVIDIA GH100, 1755 MHz, 14592 Cores, 456 TMUs, 24 ROPs, 81920 MB HBM2e, 1593 MHz, 5120 bit
www.techpowerup.com www.techpowerup.com
www.techpowerup.com

AMD Radeon Instinct MI300X Specs

AMD Aqua Vanjaram, 2100 MHz, 19456 Cores, 1216 TMUs, 0 ROPs, 196608 MB HBM3, 2525 MHz, 8192 bit
www.techpowerup.com www.techpowerup.com
oh yeah and they are not from the same generation either going by launch dates, maybe you are unaware but nvidia has already launched b200


its the defacto dx12 benchmark
and its hypocritical at the same time, because you have conveniently skipped RT benchmarks
we are talking about the full extent of capabilities of both chips, so it doesnt make much sense to skip RT performance

here data on port royal:
7900 xtx: 15793
rtx 4090: 25692 (+63%)
www.eteknix.com

AMD Radeon RX 7900 XT & 7900 XTX Review - Page 8

AMD is finally ready to throw some fresh hardware into the endless battle between Nvidia and AMD for gaming dominance, sure Intel is part of the rumble these days too, but it’s fair to say they’ve got…
www.eteknix.com


and tsmc 4n is just a custom 5n node for nvidia as has been reported by popular press outlets
Click to expand...

Yeah but listen... pineapple belongs on pizza.
 
You must log in or register to reply here.
Back
Top Bottom