AMD's new Zen 5 flagship gets benchmarked — Ryzen 9 9950X Engineering Sample isn't as impressive in Blender at maximum power settings
Pushing the limits on the Ryzen 9 9950X doesn't massively improve throughput, unfortunately.
Since our last AMD Ryzen 9 9950X ES leaked benchmarks story, AnandTech forum member Igor_Kavinsky has continued posting new Engineering Sample benchmarks in his original thread. The enthusiast has now undertaken 253W PPT and "Unlimited" Package Power Tracking (PPT) testing. This is in addition to the previously-covered 90W, 120W, 160W, and 230W PPT results we've covered. We've also included our own Ryzen 9 7950X benchmarking results for a quick comparison with the newer chip.
In our last article on these benchmarks, we noted that the Ryzen 9 9950X seems to boast significant efficiency improvements over the Ryzen 9 7950X, not just higher performance in general. In particular, it was noted that the Ryzen 9 9950X seems capable of outperforming the Ryzen 9 7950X even when operating at a lower maximum wattage. It also remained fairly competitive with the 170W Ryzen 7950X at wattages as low as 120W.
AMD Ryzen 9 9950X ES Blender Benchmark Scores
Row 0 - Cell 0 | Ryzen 9 9950X "Unlimited PPT" | Ryzen 9 9950X 253W PPT | Ryzen 9 9950X 230W PPT | Ryzen 9 9950X 160W PPT | Ryzen 9 7950X 170W PPT | Ryzen 9 9950X 120W PPT | Ryzen 9 9950X 90W PPT | Ryzen 9 9950X 60W PPT |
Blender "Monster" Benchmark Score | 367.6 | 366 | 353.4 | 319.7 | 289.7 | 268.7 | 227.5 | 153.2 |
Blender "Junkshop" Benchmark Score | 231.4 | 230 | 226.1 | 205.8 | 172.8 | 177.5 | 150.6 | 101.8 |
Blender "Classroom" Benchmark Score | 180.1 | 179 | 171.3 | 152.5 | 136.7 | 129.8 | 108.8 | 72.7 |
Blender Overall Benchmark Score | 779.1 | 775 | 750.8 | 678 | 599.2 | 576 | 486.9 | 327.7 |
*Author's Note: My prior article on this topic referred to these power targets as "TDP" instead of "PPT". These are...mostly the same thing, but whereas TDP (Thermal Design Power) refers to the CPU's power target, PPT (Package Power Tracking) refers to all power being directed to the CPU socket, and adjustments change the maximum wattage to the socket, and thus real TDP is lower. "Unlimited PPT" allows for as much wattage as the CPU and socket can support.
Isolating the two new benchmark results, one has to immediately note that only minor improvements have been gained by pushing the PPT power limits to their absolute maximum. The 253W result with a 5.5 GHz overclock still maintains impressive temperatures of 61C or less thanks to the liquid cooling setup used with this ES. However, fully removing the power limits kicks up temps to 80C under liquid cooling while achieving only the most marginal of performance improvements.
In other words, the most impressive results here... still start at around 170W, compared to the preceding CPU. It is fully within expectations for a successor to outperform its predecessor at the same or higher power targets, but the efficiency gains remain the most impressive aspect of this story.
That said, it's still nice that the Ryzen 9 9950X could be pushed this far with (apparently) a standard liquid cooling setup, though we don't know if it was done with an AIO or a custom loop. Apparently, no CPU delidding was needed to achieve these results, and in fact doing so would have likely upset AMD, since this is an Engineering Sample that must eventually be returned, per Igor_Kavinski's secondhand reports. (Note that while Igor posts these benchmarks, an unnamed source is actually running them and sending them to him.)
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Christopher Harper has been a successful freelance tech writer specializing in PC hardware and gaming since 2015, and ghostwrote for various B2B clients in High School before that. Outside of work, Christopher is best known to friends and rivals as an active competitive player in various eSports (particularly fighting games and arena shooters) and a purveyor of music ranging from Jimi Hendrix to Killer Mike to the Sonic Adventure 2 soundtrack.
It’s been 30 years since Intel’s infamous Pentium FDIV bug reared its ugly head – a math bug caused Intel’s first CPU recall
AMD crowns the Ryzen 7 9800X3D a ‘gaming legend’ in a surprise announcement — chipmaker claims $479 Zen 5 3D V-Cache chip is up to an average 20% faster than Intel Core Ultra 9 flagship
-
TerryLaze Admin said:An AMD Ryzen 9 9950X ES gets a full suite of Blender benchmarks at different PPT/TDP targets.
AMD's new Zen 5 flagship gets benchmarked — Ryzen 9 9950X Engineering Sample isn't as impressive in Blender at maximum power settings : Read moreIt also remained fairly competitive with the 170W Ryzen 7950X at wattages as low as 120W.
Not just the prior article...the 7950x has a TDP of 170W and a PPT of 230W.
...
*Author's Note: My prior article on this topic referred to these power targets as "TDP" instead of "PPT". These are...mostly the same thing, but whereas TDP (Thermal Design Power) refers to the CPU's power target, PPT (Package Power Tracking) refers to all power being directed to the CPU socket, and adjustments change the maximum wattage to the socket, and thus real TDP is lower. "Unlimited PPT" allows for as much wattage as the CPU and socket can support.
So roughly a 30% increase in performance for the same power, not bad at all.
But apparently also no overclocking headroom at all since AMD maxed it out as much as possible already. -
usertests Blender was supposed to be one of the best-case scenarios for Zen 5, something to keep in mind.Reply -
thisisaname Wonder what is the bottle neck in the system if more power does little to increase performance?Reply
Is it as TerryLaze says and they do not have any overclocking head room left? -
SunMaster thisisaname said:Is it as @TerryLaze says and they do not have any overclocking head room left?
Given what is happening to the higher end 13k/14k series CPUs, perhaps Intel didn't have headroom either. -
NinoPino Apparently <6 Ghz is becoming a hard limit for actual technology, very interesting to see.Reply
They need to work hard on architectures and IPC. -
Thunder64 NinoPino said:Apparently <6 Ghz is becoming a hard limit for actual technology, very interesting to see.
They need to work hard on architectures and IPC.
Similar things had been said about 4GHz. We shall see what the future holds. -
NinoPino
Looks like we'll stay glued to high 5 Ghz for 3 Intels generations (13, 14, 15th) and 2 AMD generations (7000 / 9000).Thunder64 said:Similar things had been said about 4GHz. We shall see what the future holds.
Not happened very often. -
abufrejoval I don't know why the 61°C should impress you, because I believe it's a programmed threshold.Reply
If I remember correctly, Lisa Su said that Ryzens will only ever go beyond their PPT limits if your CPU won't exceed such a relatively low temperature threshold.
They'll keep clocking up to 90°C within PPT limits, but if you want to go outside, a much lower temperature needs to be maintained.
So what's impressive is perhaps more the cooler that is able to keep it at around 60°C with 250 Watts of heat.
I'd love to know if the CCDs have been maintained at similar size even with the process shrink, just to maintain enough cooling surface. I guess they've invested more transistors into IPC improvements, but they might have also added more dark silicon for the purpose of maintaining cooling capabilities at these high clocks.
I guess by doing the opposite of what they do for the compact cores, they could theoretically even reach 6GHz at the cost of a much larger CCD (which they can't fit, either). But the economic optimum is much nearer the compact cores, even if the desktop enthusiast market probably wouldn't buy that, even if it meant a max of 32 cores instead of 16.
I'd love to see a dual 12 core (4P+8C) variant benchmarked, to see where that would wind up...
And another with V-cache on the P cores...
So many possibilities, but AMD needs to keep the variants as low as possible for economy! -
thestryker
TSMC N4 is a refinement of N5 so even if the chips were roughly the same transistor wise they aren't likely much smaller if at all. If I'm remembering the numbers correctly it's a single digit percentage density improvement over N5.abufrejoval said:I'd love to know if the CCDs have been maintained at similar size even with the process shrink, just to maintain enough cooling surface. -
TerryLaze
I believe the issues are purely due to mobos pushing way too much Vcore into the CPUs in an attempts to make auto clocks more stable at higher clocks.SunMaster said:Given what is happening to the higher end 13k/14k series CPUs, perhaps Intel didn't have headroom either.
If you tune it yourself or get a mobo that doesn't push everything to 11 from the get go you would have no issues.
Also if AMD doesn't come out pre release and forces mobo makers to use safe settings the same will happen for them, now that the node has become good enough to use high values.
Also unlimited PPT = 318W pretty much in line with what the 14th gen draws unlimited, and it shows that it's PBO which is overclocking just like all core enhancement (and all the equivalents) on intel, I wonder how the reviewers are going to spin this into a positive for amd while keeping it a negative for intel.