AMD 7800 / 7700 reviews

[Shmee]

Would be nice if I got a free Starfield code for the 7900 XTX I bought in January.

 

[RnR_au]

You can only wonder what would have happened if they had ported N21 to N5...

I like AMD's experimentation. Hopefully they have learned a heap from this 'failure'.

 

[igor_kavinski]

Would be nice if I got a free Starfield code for the 7900 XTX I bought in January.

Ask your retailer for the coupon code.

 

[GodisanAtheist]

Ask your retailer for the coupon code.

-Yep, usually just a stack of them sitting on the manager's desk.

 

[igor_kavinski]

-Yep, usually just a stack of them sitting on the manager's desk.

If not, he should directly contact AMD. Would let us know how much AMD values their customers.

 

[Saylick]

Alright @uzzi38, let's see how high she clocks

https://twitter.com/x/status/1699778239412236714

 

[Brainonska511]

If not, he should directly contact AMD. Would let us know how much AMD values their customers.

Why would AMD give out a code? They had a different promotion going on earlier in the year that likely paired with that purchase (depending on the timing, either the Last of Us Part 1, or Dead Island 2 + Callisto Protocol).

 

[DAPUNISHER]

If not, he should directly contact AMD. Would let us know how much AMD values their customers.

Are you sucking the nitrous oxide out of the whip cream can? Why would a company with shareholders give a $100 game code to previous customers? GTFO.

The 7700XT did its job in upselling the 7800XT. Shipped and sold by Newegg are all OOS. OOS on Amazon, or shipping dates over 3 weeks out. Can get it at B&H, but I don't deal with them because they make you sign for the package. Homie don't play that. Not paying the markup from Platinum Micro or another Ferengi dealer.

 

[VirtualLarry]

tried nerdgearz.com, or aaawave.com ?

XFX RX 7800 XT SPEEDSTER MERC319 RADEON BLACK Gaming Graphics Card with 16GB GDDR6 HDMI 3xDP, AMD RDNA 3 RX-78TMERCB9

Shop our widest range of computer parts & components; graphic cards, motherboards, and software. GPU Mining at the best prices! Free Shipping!

www.aaawave.com

 

[DAPUNISHER]

tried nerdgearz.com, or aaawave.com ?

Zero confidence in either vendor. I expect they are Ferengi too.

EDIT: AAAwave GPU pricing is surprisingly good. No 7800XT however. I prefer Amazon since I get cash back, fast shipping, and don't have to deal with fed ex or UPS. Both of which suck compared to the Amazon sub contractor.

 

[AtenRa]

Even after this video , RTX4070 will sell 2-3x times the volume of 7800XT.

 

[GodisanAtheist]

Nvidia outselling AMD is a foregone conclusion, not only because mindshare but also because they simply put out way more cards into the channel and in markets AMD is basically absent from.

If the 4070 only outsells the 7800XT by 2-3x though, that would be a massive failure on the part of NV.

It will outsell by a factor of 5-6x to meet standard NV volumes over AMD.

 

[Heartbreaker]

Nvidia outselling AMD is a foregone conclusion, not only because mindshare but also because they simply put out way more cards into the channel and in markets AMD is basically absent from.

If the 4070 only outsells the 7800XT by 2-3x though, that would be a massive failure on the part of NV.

It will outsell by a factor of 5-6x to meet standard NV volumes over AMD.

That seems to check out with last gen ov 3080 over 6800xt. In many ways this is just this fight over again.

I think if AMD starts making that kind of dent, you would see NVidia cut 4070 prices, and then it would return to 5-6x...

 

[igor_kavinski]

Are you sucking the nitrous oxide out of the whip cream can? Why would a company with shareholders give a $100 game code to previous customers? GTFO.

Let him try. If it fails, we get some people here very happy who will expose their Green team allegiance even more while boobooing AMD

 

[Tuna-Fish]

This is a direct quote from TSMC.

Compared to N5, N4P will also deliver a 22% improvement in power efficiency as well as a 6% improvement in transistor density. In addition, N4P lowers process complexity and improves wafer cycle time by reducing the number of masks.

4N is not N4P. nVidia is not using N4P. They are using "4N", which is their minor tweaks on top of N5. Not in any way meaningfully different from the minor tweaks AMD has also applied to N5.

I will add a wikipedia link showing all the enhanced variants of 5nm TSMC that you insist do not exist. Or that all TSMC 5nm silicon is the same throughout the product cycle.

Why do you refuse to read the posts you are replying to, and instead keep arguing against things that are not said in them?

The Nvidia silicon is at least 22% more power efficient than the N5 that AMD uses. How Nvidia manages that efficiency gain shows in the reviews of power draw.

You do understand that architectural differences can account for the difference in power draw?

This is almost certainly where the difference is coming from. Ampere unquestionably used a much inferior process when compared to RDNA2, yet managed to be similar if not better in power efficiency. Now NV has the same process, of course they are more efficient.

 

[Abwx]

yet managed to be similar if not better in power efficiency. Now NV has the same process, of course they are more efficient.

Better certainly not, they had lower perf/Watt, and this despite much lower frequencies, they relied on big chips clocked moderatly to have acceptable perf/watt.

That being said it seems that when it s Nvidia that has worse perf/watt it doesnt count, it matters only if it s AMD...

AMD Radeon RX 7800 XT & RX 7700 XT im Test: Lautstärke, Temperatur, Leistungsaufnahme und OC

Radeon RX 7800 XT & 7700 XT im Test: Lautstärke, Temperatur, Leistungsaufnahme und OC / Lautstärke & Kühlung

www.computerbase.de

 

[Timorous]

Alright @uzzi38, let's see how high she clocks

https://twitter.com/x/status/1699778239412236714

TPU got nearly 3Ghz out of their nitro+

Method seems to be set Max to 5000, Min to 2800, max out the power slider and then keep undervolting until you are unstable then add a few more mv. Once the core is dialed you can clock the ram up. If core clocks seem to stick up the minimum clock.

With that they saw a 15% gain in timespy. Would love to see more games tested to see if that holds up in actual games.

 

[Hans Gruber]

4N is not N4P. nVidia is not using N4P. They are using "4N", which is their minor tweaks on top of N5. Not in any way meaningfully different from the minor tweaks AMD has also applied to N5.


Why do you refuse to read the posts you are replying to, and instead keep arguing against things that are not said in them?

You do understand that architectural differences can account for the difference in power draw?

This is almost certainly where the difference is coming from. Ampere unquestionably used a much inferior process when compared to RDNA2, yet managed to be similar if not better in power efficiency. Now NV has the same process, of course they are more efficient.

I do read the posts before I reply.

I am tired of arguing with you. I provided literature from TSMC indicating power efficiency gains of N4P. The Nvidia silicon 4N is the most advanced process on the 5nm node. There are probably more power efficiency gains in 4N over N4P. AMD is using less efficient/standard (N5) 5nm TSMC than Nvidia. The power use across the Nvidia 40 series cards is clear. I am certain the price of the silicon is much higher for 4N.

The TSMC 5nm process is 4 years old now. Throughout the node they released enhanced processes on that node that yield more power efficiency vs. the standard N5 silicon. Tuna Fish denies that there are different processes and variants on the 5nm node that are offered throughout the life of the TSMC 5nm node. The TSMC 7nm had 4 different processes on the node throughout it's lifecycle.

 

[Tuna-Fish]

The Nvidia silicon 4N is the most advanced process on the 5nm node.

It is not. It is very similar to the original N5.

There are probably more power efficiency gains in 4N over N4P.

There are not, because the 4N predates the N4P.

I will add a wikipedia link showing all the enhanced variants of 5nm TSMC that you insist do not exist. Or that all TSMC 5nm silicon is the same throughout the product cycle.

I am very aware that there are many variants of the TSMC N5. My point from the beginning has been that 4N is a very mild, almost indistinguishable variant. It is absolutely not the most advanced 5nm process, and I do not understand why you keep claiming it is.

 

[Hans Gruber]

It is not. It is very similar to the original N5.

There are not, because the 4N predates the N4P.


I am very aware that there are many variants of the TSMC N5. My point from the beginning has been that 4N is a very mild, almost indistinguishable variant. It is absolutely not the most advanced 5nm process, and I do not understand why you keep claiming it is.

You lie Tuna fish. you said all 5nm variants were the same. Even worse, you wrongly assumed customized processes were named for the companies who used them. They were created by TSMC and customers could choose which process they wanted to purchase on the 5nm silicon node. As the 5nm matured, they release newer more efficient processes on the node.

Nvidia GPU's are ultra efficient because of the 5nm process they are on. If AMD had RDNA3 on 4N, you would see power consumption numbers dramatically drop and probably increased performance as well. It's the cost for 4N that keeps AMD on the standard N5.

 

[Heartbreaker]

You lie Tuna fish. you said all 5nm variants were the same. Even worse, you wrongly assumed customized processes were named for the companies who used them. They were created by TSMC and customers could choose which process they wanted to purchase on the 5nm silicon node. As the 5nm matured, they release newer more efficient processes on the node.

Nvidia GPU's are ultra efficient because of the 5nm process they are on. If AMD had RDNA3 on 4N, you would see power consumption numbers dramatically drop and probably increased performance as well. It's the cost for 4N that keeps AMD on the standard N5.

AMD couldn't be on 4N, because that N literally stands for NVidia.

It's just like Turing on 12FFN.

Pascal was on TSMC 16FF.

For Turing NVidia named it 12FFN, with the N for NVidia. It's really just a slightly tweaked TSMC 16FF. The same kind of tweak any big company would have had working with TSMC, but only NVidia seems arrogant enough to name their tweaks and pretend it's an all new node.

If you check the transistor density. Both 16FF and 12FFN are about 25MT/mm2, and no one else got 12FFN - N being for NVidia.

 

[gdansk]

If AMD had RDNA3 on 4N, you would see power consumption numbers dramatically drop and probably increased performance as well

Your only basis for that claim is a wiki article specifying "4N" characteristics as "Unspecified" and only citation a Nvidia marketing announcement.
If "4N" is more advanced than N5 why does AD103 only have 120 million transistors per mm² while AMD's N31 GCD has 150 million transistors per mm²? Kinda weird that somehow AMD is acheiving better density on a worse process, isn't it Hans? The process is not primary difference between AD103/AD104 and N31/N32. It's the architecture, multi-chip design and specific implementation. AMD's not suffering from being cheap on process (for the GCDs anyway).

 

[Hans Gruber]

AMD couldn't be on 4N, because that N literally stands for NVidia.

It's just like Turing on 12FFN.

Pascal was on TSMC 16FF.

For Turing NVidia named it 12FFN, with the N for NVidia. It's really just a slightly tweaked TSMC 16FF. The same kind of tweak any big company would have had working with TSMC, but only NVidia seems arrogant enough to name their tweaks and pretend it's an all new node.

If you check the transistor density. Both 16FF and 12FFN are about 25MT/mm2, and no one else got 12FFN - N being for NVidia.

Ok, Heartbreaker. Riddle me this. Why were the TSMC 7nm processes all named N7, N7P ,N7+ and N6? Were those all featuring an "N" when Nvidia was not a customer? Or the 5N process on the 5nm silicon? Is that for Nvidia? What about N4, is that named for Nvidia as well?

 

[Heartbreaker]

Ok, Heartbreaker. Riddle me this. Why were the TSMC 7nm processes all named N7, N7P ,N7+ and N6? Were those all featuring an "N" when Nvidia was not a customer? Or the 5N process on the 5nm silicon? Is that for Nvidia? What about N4, is that named for Nvidia as well?

N at the beginning is TSMC nomenclature.
N at the end is NVidia specific.

Anand mention of the 12nm FFN:

NVIDIA Volta Unveiled: GV100 GPU and Tesla V100 Accelerator Announced

www.anandtech.com

"built on TSMC’s still green 12nm “FFN” process (the ‘n’ stands for NVIDIA; it’s a customized higher perf version of 12nm for NVIDIA)."

Though turns out it's just tweaked 16nm.

 

[gdansk]

Ok, Heartbreaker. Riddle me this. Why were the TSMC 7nm processes all named N7, N7P ,N7+ and N6? Were those all featuring an "N" when Nvidia was not a customer? Or the 5N process on the 5nm silicon? Is that for Nvidia? What about N4, is that named for Nvidia as well?

"4N" is not N4... maybe some source of the weird argument here.