Leaked AMD Radeon RX 7700 & RX 7800 GPU Benchmarks Emerge

[T0@st]

A set of intriguing 3DMark Time Spy benchmark results have been released by hardware leaker All_The_Watts!!—these are alleged to have been produced by prototype Radeon RX 7700 and Radeon RX 7800 graphics cards (rumored to be based on variants of the Navi 32 GPU). The current RDNA 3 lineup of mainstream GPUs is severely lacking in middle ground representation, but Team Red is reported to be working on a number of models to fill in the gap. We expect a number of leaks to emerge as we get closer to a rumored product reveal scheduled for late August (to coincide with Gamescon).

The recently released 3DMark Time Spy scores reveal that the alleged Radeon RX 7700 candidate scored 15,465 points, while the RX 7800 achieved 18,197 points—both running on an unspecified test system. The results (refer to the Tom's Hardware-produced chart placed below) are not going to generate a lot of excitement at this stage when compared to predecessors and some of the competition—evaluation samples are not really expected to be optimized to a great degree. We hope to see finalized products with decent drivers putting in a good appearance and performing better later on this year.



View at TechPowerUp Main Site | Source

 

[wNotyarD]

The results (refer to the charts placed below) are not going to generate a lot of excitement at this stage when compared to predecessors and some of the competition—evaluation samples are not really expected to be optimized to a great degree.

If it isn't too much to ask, how those fare so we know what numbers are going on?

EDIT: didn't see the last image before asking

 

[docnorth]

The results seem acceptable. Pricing will make them look good or bad

 

[fancucker]

basically the same performance for the same price, after 2 years. and an insufficient RT upgrade, middling FSR. total disappointment. AMD needs to step it up.

 

[3x0]

basically the same performance for the same price, after 2 years. and an insufficient RT upgrade, middling FSR. total disappointment. AMD needs to step it up.

Everyone needs to step it up

 

[john_]

are not going to generate a lot of excitement at this stage when compared to predecessors and some of the competition

I think we all knew it by now. The fact that AMD hasn't released an RX 7700 and an RX 7800 card was pretty clear indications that the new RDNA3 cards where underperforming in gaming. And even if 7900XT and XTX where based on GPUs with more stream processors and Raytracing accelarators than the 6900X/6950X, to hide the fact that they where NOT an improvement over 6900X/6950X models, at least in gaming, the RX 7600 made it pretty clear that in gaming RDNA3 = RDNA2.
I guess pricing will be about the same as the current price of the equivalent in performance RDNA2 GPU.

 

[ZoneDymo]

Everyone needs to step it up

This, we know these will probably perform decent enough....but the price will be way too high for what they offer, which will sadly also not help in the second hand market which also has way too high pricing still.

 

[Nostras]

basically the same performance for the same price, after 2 years. and an insufficient RT upgrade, middling FSR. total disappointment. AMD needs to step it up.

Do you have info on pricing then? How can you say this without knowing that?
We can bitch and moan about naming scheme and what performance they actually put out, but without knowing the price calling this a total disappointment is whack to me.

 

[sethmatrix7]

basically the same performance for the same price, after 2 years. and an insufficient RT upgrade, middling FSR. total disappointment. AMD needs to step it up.

My guy, they aren’t going to change the hardware of the card to accommodate another manufacturers technology that only applies to like 15 games- especially when enabling said feature nerfs performance even on their competitors cards.

 

[Icon Charlie]

basically the same performance for the same price, after 2 years. and an insufficient RT upgrade, middling FSR. total disappointment. AMD needs to step it up.

I agree with this comment and will add.

1. The AMD TAX is still enforced. This is one of the reasons why people are not buying into the product. Double the price for a X670 motherboard over the X570 series is insane. And for the record they will NOT drop the price on charging the AIB's the cost of their chipset.

2. The cost is still too damned for the overall upgrade. Again you MUST have DDR 5. You MUST have a new motherboard. Because...

3. Buy their own admission. AMD performance, with their own marketing gimmick to promote their 8000 series CPU's. THEY STATED the performance increases from Zen3 to Zen 4 was a paltry +13% using 22 different types of programs for their testing. +13 PERCENT. From Zen 4 to Zen 5 they stated around +20% performance increase. Meaning that from Zen 3 to Zen 5 we are talking maybe around 30%.

4. My own rig is a 5900 OEM and on stress testing using Cinebench with temps lower than 76 degrees C., (all day long depending on the temps, under high loads, 34 to 48c normal loads, air cooled with a hyper T2 Cooler ) its running close to what a 7700X @ 58 watts average.

Why am I going to spend 6 to 8 hundred dollars to get a + 13% performance average to go to a Zen 4?? We already know when you compare apples to apples with Intel and their use of DDR 4/5 Motherboards we only saw a small increase in performance.

And I don't even trust their talking heads... Dr. Su and all the way down to marketing are guilty in plausible deniability is selling their product. Just like the current trend in the industry... Talk nonsense to the gerbils because they know that they will always buy because...

They always do.

And I believe that if you compare a DDR 4 4800 to a DDR 5 4800 (YES they do exist) with a comparable CPU from both Zen 3 nd Zen 4, You are not going to get that OW WOW factor that you just paid over 600 dollars for a....

Upgrade.

Zen 4 is mediocre at best. A clusterFluke at worst when you add the long times to get into windows. The excessive heat issues. Quality control issues and the cost...

I am not spending $600-800 upgrading my rig, because of the cost of ownership when originally created is only $40 dollars a month.

Only the smart people that did their research know when to and how to spend their money are the winners here. Zen 4 was a hard pass for me. The prices are just too high for what you get overall.

As proven by AMD themselves.

 

[Lew Zealand]

Do you have info on pricing then? How can you say this without knowing that?

We don't even know if these are ES or final samples so these could be taken as minimum performance expectations.

 

[TheinsanegamerN]

Do you have info on pricing then?

It's going to be screwed, like everything else this generation.

How can you say this without knowing that?
We can bitch and moan about naming scheme and what performance they actually put out, but without knowing the price calling this a total disappointment is whack to me.

Well, we have this thing called "past precedence". We know that AMD is not interested in providing good perf/$, and even with disappointing cards they still under-deliver on price. AMD is not a budget brand, the CEO herself has stated this more then once.

We have watched AMD release a flat line in perf/$ ratios compared to rDNA2. So why would that change now?

 

[MarsM4N]

So in short: RX7700 = RX6800 & RX7800 = RX6800XT. The performance stagnation continues.

And how the performance aligns with Nvidia's offerings, it almost smells like collution. Nvidia at least made some major power consumption improvements, AMD on the other hand made a step backwards with their broken power management. If you're into AMD then better get a 6000 Series card before they are gone. Way better deal. Or just skip this doomed GPU generation.

 

[TheoneandonlyMrK]

I'll sit sharpening my pitchfork a while yet, at least until a review deternes the shitnados strength.

Given the shitstorm Nvidia just dumped on the low to now mid slash high end, the Bars Low, f£#@£#n very low so low efficiency is on lips pre review, lolz
.

 

[ZoneDymo]

So in short: RX7700 = RX6800 & RX7800 = RX6800XT. The performance stagnation continues.

And how the performance aligns with Nvidia's offerings, it almost smells like collution. Nvidia at least made some major power consumption improvements, AMD on the other hand made a step backwards with their broken power management. If you're into AMD then better get a 6000 Series card before they are gone. Way better deal. Or just skip this doomed GPU generation.

I mean, then why does the RX7900XT(X) sit so high for efficiency on TPU's results?
I mean yeah Nvidia RTX4080 is top dog, but AMD is not far behind and beating their own older generation in that department.

 

[Zunexxx]

I mean, then why does the RX7900XT(X) sit so high for efficiency on TPU's results?
I mean yeah Nvidia RTX4080 is top dog, but AMD is not far behind and beating their own older generation in that department.

Tpu efficiency charts are not totally accurate. There are multiple reports navi31 drawing 2-3 timesore power with similar fps compared to the 4080. Pretty sure there were couple YouTubers talking about that too.

https://www.reddit.com/r/nvidia/comments/14vxbb8
Heres a link.

 

[tabascosauz]

Tpu efficiency charts are not totally accurate. There are multiple reports navi31 drawing 2-3 timesore power with similar fps compared to the 4080. Pretty sure there were couple YouTubers talking about that too.

https://www.reddit.com/r/nvidia/comments/14vxbb8
Heres a link.

But different reviewers focus on different things - that doesn't make either of them less accurate. w1zz tests generally AAA and otherwise demanding games - in that case, why would you framecap if you're trying to find maximum performance at 4K max quality? Ali is an avid player of less demanding but higher framerate esports titles at "optimal" quality settings - in that case, why would you NOT framecap, when it's part of generally accepted good practice for most of those games? You can't get the whole picture from just one side's data.

TPU already has a chart for 60Hz, it's been telling this story since launch. Navi31 is stuck at the 110W best case scenario due to a variety of factors (ie. the way Radeon handles VRAM, and inherent memory rails' draw). AD102/AD103/AD104 all scale down well below 100W, enabling completely fanless gaming under the right circumstances.

I'll sit sharpening my pitchfork a while yet, at least until a review deternes the shitnados strength.

Given the shitstorm Nvidia just dumped on the low to now mid slash high end, the Bars Low, f£#@£#n very low so low efficiency is on lips pre review, lolz
.

Of course I agree that speculating on leaks is stupid. But in your heart of hearts, do you really believe that launch prices will be attractive for a midrange-high end GPU? AMD is gonna launch a product right where you want it to be, and give themselves 0 zero room for hefty price drops?

A little myopic to focus on the "shitstorm". It takes two to tango, and by golly, AMD has been tangoing.

 

[Minus Infinity]

Highly underwhelming but given how pathetic the 7600 is, it as highly expected. RDNA3 has been a waste of effort from AMD. Poor power efficiency, performance way weaker than hyped, launch prices to high. Honestly, if they repeat this crap with RDNA4 I don't see there being an RDNA5, even Intel will outsell them in a few years if it can get Battlemage and Celestial out on time.

 

[AnotherReader]

Highly underwhelming but given how pathetic the 7600 is, it as highly expected. RDNA3 has been a waste of effort from AMD. Poor power efficiency, performance way weaker than hyped, launch prices to high. Honestly, if they repeat this crap with RDNA4 I don't see there being an RDNA5, even Intel will outsell them in a few years if it can get Battlemage and Celestial out on time.

I think your prognostication of the future is unlikely. So far, Intel is even further behind AMD than AMD is behind Nvidia. The A770 is twice the size of the RX 7600 and consumes 50% more power. Despite all that, the RX 7600 is 10 percent faster at 1080p. Intel will do well because of their strong partnerships with OEMs, but they have a long way to go, and given how Pat Gelsinger has been closing product lines, the GPU division's future is uncertain in the long term.

 

[Pumper]

Seems like the goal here is to replace the dwindling 6000 series supply rather than releasing something that could be considered a new gen.

 

[Broken Processor]

I read about AMD testing a vapour chamber for CPU heatsink and deciding it wasn't worth it for the small performance improvement maybe they should apply that logic to everything they do.

I don't get it if they are releasing the 7800 and it performs basically the same as the 6800xt why not just keep making the 6800xt it uses older node so can keep price down and still profit not to mention saving on design costs.

At this point they are trolling themselves.

 

[TheoneandonlyMrK]

But different reviewers focus on different things - that doesn't make either of them less accurate. w1zz tests generally AAA and otherwise demanding games - in that case, why would you framecap if you're trying to find maximum performance at 4K max quality? Ali is an avid player of less demanding but higher framerate esports titles at "optimal" quality settings - in that case, why would you NOT framecap, when it's part of generally accepted good practice for most of those games? You can't get the whole picture from just one side's data.

TPU already has a chart for 60Hz, it's been telling this story since launch. Navi31 is stuck at the 110W best case scenario due to a variety of factors (ie. the way Radeon handles VRAM, and inherent memory rails' draw). AD102/AD103/AD104 all scale down well below 100W, enabling completely fanless gaming under the right circumstances.



Of course I agree that speculating on leaks is stupid. But in your heart of hearts, do you really believe that launch prices will be attractive for a midrange-high end GPU? AMD is gonna launch a product right where you want it to be, and give themselves 0 zero room for hefty price drops?

A little myopic to focus on the "shitstorm". It takes two to tango, and by golly, AMD has been tangoing.

No your right of course, I should be going mental about stats I don't know?!?

Myopic, check that mirror.

Price's, I can't comment on, I'm no mind-reader, you do you.

And no it didn't take two to tango Huang sets Nvidia's agenda not Lisa Su.

 

[gdp77]

I do not expect anything interesting from RDNA3 GPUs. They failed miserably in performance, they failed even more in power consumption (If you live in Europe, the electricity cost is outrageous) and AMD failed miserably to at least price the GPUs competitively. Seems that they know that there is no other option and price their GPUs a little lower than Nvidia. You could argue that they are price fixing the market in coordination with Nvidia. I see reviewers/youtubers saying all the time that the "market is down" etc. Well, that's wrong. The majority of customers want to buy, but the products suck in so many ways that it is difficult to comprehend. AMD tries to fool its customer with 7000 series, offering almost 3-year old performance, with no obvious improvements on power consumption or at least cheaper prices. On the other hand Nvidia offers the "beast" 4090 which costs as much as 2 or more salaries on most European countries. Lower Nvidia tiers are overpriced (as usual). At least Nvidia got the power efficiency right and their cards are very very efficient. AMD did nothing right on 7000 series.

 

[Minus Infinity]

I think your prognostication of the future is unlikely. So far, Intel is even further behind AMD than AMD is behind Nvidia. The A770 is twice the size of the RX 7600 and consumes 50% more power. Despite all that, the RX 7600 is 10 percent faster at 1080p. Intel will do well because of their strong partnerships with OEMs, but they have a long way to go, and given how Pat Gelsinger has been closing product lines, the GPU division's future is uncertain in the long term.

AMD is actually losing market share to Intel, not gaining share from Nvidia. Latest GPU figures are sad for all but terrible for AMD. I'm skipping this gen entirely. I thought Turing was bad.

 

[MarsM4N]

But different reviewers focus on different things - that doesn't make either of them less accurate. w1zz tests generally AAA and otherwise demanding games - in that case, why would you framecap if you're trying to find maximum performance at 4K max quality? Ali is an avid player of less demanding but higher framerate esports titles at "optimal" quality settings - in that case, why would you NOT framecap, when it's part of generally accepted good practice for most of those games? You can't get the whole picture from just one side's data.

TPU already has a chart for 60Hz, it's been telling this story since launch. Navi31 is stuck at the 110W best case scenario due to a variety of factors (ie. the way Radeon handles VRAM, and inherent memory rails' draw). AD102/AD103/AD104 all scale down well below 100W, enabling completely fanless gaming under the right circumstances.

Wizzard's power consumption measurements for gaming aren't very conclusive since he's using only 1(!) game (Cyberpunk 2077) as a reference. Optimus Tech used at least 6 games for his tests where the RTX 4080 is drawing around 100W to 200W less power than the 7900 XTX. In CS:GO the 4080 is even drawing ~60% less power, that's just crazy! And he didn't even use a lower power target which would bring the numbers down even more, lol.

Power Consumption Testing Details:

• Gaming: Cyberpunk 2077 is running at 3840x2160 with Ultra settings and ray tracing disabled. Textures are set to "low", to ensure that cards with 4 GB and 6 GB aren't handicapped due to their VRAM size. We ensure the card is heated up properly, which ensures a steady-state result instead of short-term numbers that won't hold up in long-term usage.

Also if you're looking at the 100 most played games on STEAM you will find maybe a hand full of heavy demanding AAA games. Folks mostly play older or indie games, and that's what matters in the end.