Thursday, October 10th 2019

AMD TRX40 Chipset Not Compatible with 1st and 2nd Gen Threadrippers

AMD is giving finishing touches to its 3rd generation Ryzen Threadripper HEDT processor lineup, and the first wave of these chips, starting with a 24-core model, will launch alongside the AMD TRX40 chipset. It turns out that the chipset won't be compatible with 1st and 2nd generation Ryzen Threadripper processors. The upcoming 3rd generation Threadripper chips won't be backwards-compatible with the AMD X399 chipset, either. We've been hearing from reliable sources rumors of this segmentation from AMD for a few days now, and tech journalist ReHWolution just tweeted its confirmation having obtained info on upcoming motherboards from a leading brand.

The underlying reason between this restriction remains a mystery. We know that the EPYC "Rome" MCM is pin-compatible with first-generation EPYC "Naples" chips due to the fact that the newer chips are drop-in compatible with older servers via a BIOS update. The TR4 socket, too, is nearly identical to SP3r2, but for four out of eight memory channels being blanked out. It remains to be seen if for TRX40 motherboards, AMD re-purposed these unused pins for something else, such as additional PCIe connectivity or more electrical pins. We'll find out in November, when AMD is expected to launch these chips.
Source: ReHWolution (Twitter)
Add your own comment

66 Comments on AMD TRX40 Chipset Not Compatible with 1st and 2nd Gen Threadrippers

#1
FeelinFroggy
I was hoping these rumors would turn out to be false. I would love to upgrade my 1950x and drop it into my x399 Zenith Extreme. It was the most I ever paid for a motherboard.
Posted on Reply
#2
Valantar
Well that's a damn shame. Seems like AMD is taking some cues from Intel here, more or less. I was really hoping to snag a 3rd-gen TR as an upgrade for the 1920X in my partner's workstation some time down the line. If that also requires a $300 motherboard swap, that suddenly becomes a lot less enticing.
Posted on Reply
#3
kapone32
This is interesting I thought that with Intel's announcement that X299 boards will support their new Cascade Lake CPUs that AMD would not do this but I think I might know why (just guessing). The TR40 and the next chips will all be PCI_E 4.0. This could mean unfortunately that the TR40 boards may be more expensive across the stack than X399 from every vendor.
Posted on Reply
#4
Aldain
FeelinFroggy, post: 4130755, member: 174277"
I was hoping these rumors would turn out to be false. I would love to upgrade my 1950x and drop it into my x399 Zenith Extreme. It was the most I ever paid for a motherboard.
You still have zen +

Valantar, post: 4130757, member: 171585"
Well that's a damn shame. Seems like AMD is taking some cues from Intel here, more or less. I was really hoping to snag a 3rd-gen TR as an upgrade for the 1920X in my partner's workstation some time down the line. If that also requires a $300 motherboard swap, that suddenly becomes a lot less enticing.
AMD never ever promised HEDT longevity , only MA4
Posted on Reply
#5
aktpu
Perhaps AMD is worried about power delivery, 16 cores was king for first gen but I doubt that they do less than 24 cores now so maybe they are playing it safe? Also, is this AMD-thing or a Asrock-thing?
Posted on Reply
#6
ZoneDymo
honestly, this is pretty sad stuff from AMD, even IF there is a good reason for this it would be nice if they would at least make a 3rd gen Threadripper chip specifically for X399 so those adopters can at least upgrade to that.
Posted on Reply
#7
xkm1948
@JohnWal
See? Told ya it wont be.

At the same time. Only 2 gen support for HEDT on AMD is bad. Man even X370 can use RyZen 3xxx. I was hoping to put in quote for updating the TR builds in lab as well as helping other labs to upgrade their TR based system. The lab has one 1950X build, 3 2950X builds and 1 2990WX builds. Now that upgrade path is gone. Bummer for sure. I mean the 2990WX runs on Linux and is only used for comparative genomic analysis, NUMA issue is not too bad for that one.

Shame AMD, shame. I had higher hope for you. :( :(

aktpu, post: 4130761, member: 172709"
Perhaps AMD is worried about power delivery, 16 cores was king for first gen but I doubt that they do less than 24 cores now so maybe they are playing it safe? Also, is this AMD-thing or a Asrock-thing?
Zen+ based TR 2xxx series had 24 and 32 cores, can be overclocked over 4GHz. Power delievery is a non issue for the X399 boards released last year.
Posted on Reply
#8
Metroid
There we go, amd starting to behave like intel.
Posted on Reply
#9
Chrispy_
Damn. We have a bunch of x399 boards that would appreciate some Zen2 love.

Still, I can't complain; Even if you add the cost of those X399 boards, 1950x + new board ran circles around the Intel CPU-only costs for those nodes added to the farm, both in actual performance and also in performance/$
Posted on Reply
#10
Mats
Metroid, post: 4130769, member: 178915"
There we go, amd starting to behave like intel.
You think this is enough proof for that?
Posted on Reply
#11
Valantar
Aldain, post: 4130759, member: 170164"
AMD never ever promised HEDT longevity , only MA4
I know - TR4 was always an off-roadmap project. But it's still disappointing.
Posted on Reply
#12
xkm1948
Chrispy_, post: 4130779, member: 185623"
Damn. We have a bunch of x399 boards that would appreciate some Zen2 love.

Still, I can't complain; Even if you add the cost of those X399 boards, 1950x + new board ran circles around the Intel CPU-only costs for those nodes added to the farm, both in actual performance and also in performance/$
If the workloads utilize any form of AVX512, Intel nodes runs circles around TR system with same core config. Not sure what field you work in. In my area we are starting to see wide spread adoption of AVX512 in codes which gives AVX512 capable Intel system a huge boost in performance. The same amount of microbiome data pushed through using the same pipeline would finish almost 30% faster on Intel system comparing to the AMD TR system with just AVX2.

Example of a DNA sequence alignment


A lot of bioinformatics tools benefit a huge amount from implementation of AVX512. See this paper over here actually discussing the benefit of AVX512 in research

https://www.ncbi.nlm.nih.gov/pubmed/29726911


I wish AMD will have AVX512 support on their Zen arc soon.
Posted on Reply
#13
theoneandonlymrk
xkm1948, post: 4130788, member: 50521"
If the workloads utilize any form of AVX512, Intel nodes runs circles around TR system with same core config. Not sure what field you work in. In my area we are starting to see wide spread adoption of AVX512 in codes which gives AVX512 capable Intel system a huge boost in performance. The same amount of microbiome data pushed through using the same pipeline would finish almost 30% faster on Intel system comparing to the AMD TR system with just AVX2.

Example of a DNA sequence alignment


A lot of bioinformatics tools benefit a huge amount from implementation of AVX512. See this paper over here actually discussing the benefit of AVX512 in research

https://www.ncbi.nlm.nih.gov/pubmed/29726911


I wish AMD will have AVX512 support on their Zen arc soon.
Fair enough and important to note , just as important though is the fact that not everyone is doing the same things as you.
Bioinformatics is a niche of computer use not the sole use or AMD would be in trouble.
Posted on Reply
#14
HTC
Something doesn't seem right: Epyc CPUs will re-utilize the same socked but TR gen 3 won't???

Had Epyc Rome force a change in socket, one would expect TR gen 3 to do the same, but since it didn't ... something doesn't add up here!
Posted on Reply
#15
xkm1948
theoneandonlymrk, post: 4130793, member: 82332"
Fair enough and important to note , just as important though is the fact that not everyone is doing the same things as you.
Bioinformatics is a niche of computer use not the sole use or AMD would be in trouble.
I would not say it is niche in HPC. At my university more than half of the nodes are saturated with varies form of bioinformatics job: biomedical imaging data analysis and intergration; DNA/RNA/Protein sequencing; Protein modeling; Population genetics; neural network based multiomics cancer research.

It is HEDT we are talking about here. With its higher core count and good pricing AMD has seen some good adoption in research labs. What I was saying is the now widely used AVX512 and soon-to-be implemented Deep Learning Boost will slowly claw back the marketshare and mindshare from Threadripper again.

As a matter of fact one of my friend at Oregon State University working for HPC maintenance is already seeing some "buyer's remorse" Oregon State spent a crap ton of money on EPYC processors for bioinformatics applications.
https://www.advancedhpc.com/wp-content/uploads/2018/09/18165553-AMD-OSU-AHPC-Case-Study.pdf

Mainstream desktop nobody cares about AVX512 that much. That I agree RyZen 3xxx is killing it.
Posted on Reply
#16
dicktracy
We already knew that AMD wants to milk ever penny from enthusiasts since they announced the R9 series and disable SMT for lowerend models.
Posted on Reply
#17
MazeFrame
Looking back at the "oh, whoops" many mainboard makers had when 2nd Gen TR happened made AMD proceed with caution.
IIRC Asus had a VRM cooling kit for one of their boards, MSI launched some updated ones.

Putting out new specifications is a safe route. Mobo makers will have proper power delivery, buyers will not have burning mobos.

To all the "they killed muh X399"-criers: Where were you when Intel launched a new Socket and chipset nearly every generation?
Posted on Reply
#18
CheapMeat
I really don't think they're doing it as a money grab. It's very likely an actual true technical limitation issue that would cost more to do and quality assurance issue.
Posted on Reply
#19
Mamya3084
Lets wait and see. Could you imagine if AMD repeats the intel mistake, and artificially makes the boards incompatible. Aka, skylake/coffeelake refresh.
Posted on Reply
#20
Lorec
Perfect time to get a 1950X with a X399 board for cheapo ;) already have my eye on a used threadripper, just waiting for boards to get cheaper.
Posted on Reply
#21
Vya Domus
HTC, post: 4130796, member: 51238"
Something doesn't seem right: Epyc CPUs will re-utilize the same socked but TR gen 3 won't???

Had Epyc Rome force a change in socket, one would expect TR gen 3 to do the same, but since it didn't ... something doesn't add up here!
Until we hear this officially from AMD I am unsure what to believe. If TRX40 isn't compatible with Zen 1 and Zen 1+ that wouldn't be that big of a deal, there is little reason to buy these boards and pair them with the older TRs but if you are required to have TRX40 for Zen 2 that would be strange indeed. Again, gotta have AMD's official statement.

theoneandonlymrk, post: 4130793, member: 82332"
Fair enough and important to note , just as important though is the fact that not everyone is doing the same things as you.
Bioinformatics is a niche of computer use not the sole use or AMD would be in trouble.
The performance you get is unrelated to the field, there is no such thing as a AMD, Nivida or Intel having an advantage in a niche field. Computation is computation, if something scales well with vectorization or multithreading it should do so on any CPU, of course the limiting factor becomes the hardware.

And with that been said Zen 2 now has double the FP throughput so 24 cores should nicely match Intel's top of the line 18 core part even with AVX 512.
Posted on Reply
#22
kapone32
aktpu, post: 4130761, member: 172709"
Perhaps AMD is worried about power delivery, 16 cores was king for first gen but I doubt that they do less than 24 cores now so maybe they are playing it safe? Also, is this AMD-thing or a Asrock-thing?
Lorec, post: 4130820, member: 183267"
Perfect time to get a 1950X with a X399 board for cheapo ;) already have my eye on a used threadripper, just waiting for boards to get cheaper.
Look for the As rock X399 Phantom Gaming 6 the MSRP is $249.99 US
Posted on Reply
#23
efikkan
I have to say I disagree with most people in this thread about the criticism of AMD breaking backwards compatibility. While everyone would love to have hardware with endless upgradability, having this only would become useful after ~4-5 years when there is a significant upgrade, but then everything embedded on the motherboard would be outdated. If we were to have real compatibility across several generations we would need to make motherboards more simple, modular and barebone like back in the 286/386/486 days, where motherboards were basically more or less just expansion ports. This is an idea that I like, at least in theory.

We have to remember that this platform is meant for workstations, which makes reliability the most important trait. Motherboard makers are already barely maintaining support for any motherboard beyond 2 years, and they certainly don't test each model with enough hardware before shipping a BIOS update. As we've seen with AM4, the compatibility with older hardware is questionable at best. And while AMD motherboards are getting the same premium prices as some "premium" Intel boards, certain makers still fail to deliver the same quality (*ahem* MSI, Gigabyte…). If anything AMD should focus their energy on two things; 1) Firmware testing ahead of product launch 2) Put pressure on motherboard makers to do their best when making the BIOSes. A few hiccups after a product launch is excusable, repeated problems three months later is a deal breaker for workstation use.
Posted on Reply
#24
Zubasa
xkm1948, post: 4130788, member: 50521"
If the workloads utilize any form of AVX512, Intel nodes runs circles around TR system with same core config. Not sure what field you work in. In my area we are starting to see wide spread adoption of AVX512 in codes which gives AVX512 capable Intel system a huge boost in performance. The same amount of microbiome data pushed through using the same pipeline would finish almost 30% faster on Intel system comparing to the AMD TR system with just AVX2.

Example of a DNA sequence alignment


A lot of bioinformatics tools benefit a huge amount from implementation of AVX512. See this paper over here actually discussing the benefit of AVX512 in research

https://www.ncbi.nlm.nih.gov/pubmed/29726911


I wish AMD will have AVX512 support on their Zen arc soon.
I am curious, is that 30% faster performance than TR despite the AVX clocks / off-set?
Or is that a Xeon only thing?
Posted on Reply
#25
Vya Domus
Zubasa, post: 4130832, member: 30988"
Or is that a Xeon only thing?
All of Intel's CPU scale their frequency according to AVX load.
Posted on Reply
Add your own comment