Cloud Mercato tested CPU performance using a range of encryption speed tests:
Cloud Mercato's tested the I/O performance of this instance using a 100GB General Purpose SSD. Below are the results:
I/O rate testing is conducted with local and block storages attached to the instance. Cloud Mercato uses the well-known open-source tool FIO. To express IOPS the following parametersare used: 4K block, random access, no filesystem (except for write access with root volume and avoidance of cache and buffer.
.png)


Thanks, I tried "io1" and "st1", but both of them showed the same error no sufficient capacity.

I'm running IPv6 EKS cluster with p4de.24xlarge EC2 instance. P4de instances are used for workloads like large language models where we need large clusters and IPv6 helps with IP exhaustion. I noticed that `aws-node` pod is failing on p4de.24xlarge instances, but other instances from C and M instance families ran without any issues.

I also have the same issue, there is no capacity available for p4d.24xlarge instance-type in N. Virginia and Oregon regions. I have retried many times since 2 days ago but have had no luck. Is there any solution other than "try again later"?

Thanks, but I tried US-WEST-2a, US-WEST-2b, US-WEST-2c (these are the options listed in the drop down list), same error. Also, I tried GP3, also showing same error.

Hashcat is telling me that it cannot find any CUDA-compatible platforms. Since the exact same setup works with all other AWS GPU instances it might be a problem with the weird way their p4d instances are connected to the A100s.

Thank you so much for your assistance! I’ve switched to the Israel (Tel Aviv) – il-central-1 region, but the P4de option still doesn’t appear in the dropdown. It seems that it might not be available in Israel at the moment. However, I did find the P5 option in many regions, so I may go with that instead. Thanks again for all your help!

I try to launch one p4d.24xlarge instance in the region of "Oregon" (US-WEST-2). I applied for 2200 GB EBS for this instance.but I get this error: "We currently do not have sufficient p4d.24xlarge capacity in zones with support for 'gp2' volumes. Our system will be working on provisioning additional capacity."

Thanks for the information, but I did not put preference on the AZ. I assume it should choose AZ automatically. But the instance still launch failed.

The p4d instance type is listed at but p4de is missing note p4d != p4de. However the specs of p4de were added 1 year ago

Thank you for your response! I’ve reviewed the N. Virginia and Oregon regions and tested both Amazon Linux and Ubuntu (64-bit x86 architecture). However, I only see the P4d option in the dropdown, not the P4de. Is there any special access I need to request for the p4de option to appear?

I would like to run experiments with 8 A100 (80GB) GPUs. It seems like only p4de-instance supports that set-up. However, I only see p4d in the list but not p4de when creating an instance. Could you please tell me where I can get the access to p4de?

I also have the same issue, there is no capacity available for p4d.24xlarge instance-type in N. Virginia and Oregon regions. I have retried many times since 2 days ago but have had no luck. Is there any solution other than "try again later"?

Thanks, but I tried US-WEST-2a, US-WEST-2b, US-WEST-2c (these are the options listed in the drop down list), same error. Also, I tried GP3, also showing same error.

The Amazon AWS EC2 P4d instances scale up to over 4000 NVIDIA A100 GPUs for those customers who want to run AI or HPC workloads in the cloud.

AWS has announced the general availability of Amazon Elastic Compute Cloud (Amazon EC2) P4d instances with EC2 UltraClusters capability. The cloud giant has touted the offering as the next generation of GPU-powered instances for providing the best performance in machine learning training and high-performance computing (HPC) in the cloud.

Thanks, I tried "io1" and "st1", but both of them showed the same error no sufficient capacity.

It seems that there is currently a capacity constraint for the p4d.24xlarge instance type in the "Oregon" (US-WEST-2) region with "gp2" volumes.

I try to launch one p4d.24xlarge instance in the region of "Oregon" (US-WEST-2). I applied for 2200 GB EBS for this instance.but I get this error: "We currently do not have sufficient p4d.24xlarge capacity in zones with support for 'gp2' volumes. Our system will be working on provisioning additional capacity."

Specify a subnet that resides in one of those AZ above. You also may want to try with 'gp3' EBS volumes.

I tried to launch p4d.24xlarge for some more time in different regions, and eventually, it worked. It doesn't seem that the choice of the storage type (gp2, gp3, or io1) affects the ability to launch.

It sounds like you're trying to launch a p4d.24xlarge instance type, which utilizes NVIDIA A100 GPUs, but you're encountering capacity constraints in the regions you're attempting to launch in. This can happen when there's high demand for those specific instance types and not enough physical hardware available to fulfill all requests.

I tried starting p4.24xlarge in several regions. Every time, I get an error saying, "We currently do not have sufficient p4d.24xlarge capacity in zones with support for 'gp2' volumes. Our system will be working on provisioning additional capacity." I've been trying for several days. I also tried using gp3 and io1 storages to the same effect. I need a machine with A100 GPU.

Thank you for your answer. Do I have to subscribe to a paid support plan to be able to contact AWS Support? Right now, when I try to create an issue, I see "Technical support is unavailable under the Basic Support plan".

It seems like you're encountering capacity constraints for the p4d.24xlarge instance type with GPU storage options (gp2, gp3, io1) across multiple regions. This can happen when there is high demand for specific instance types or when AWS is experiencing limitations in provisioning hardware resources

Gartner Peer Insights content consists of the opinions of individual end users based on their own experiences, and should not be construed as statements of fact, nor do they represent the views of Gartner or its affiliates. Gartner does not endorse any vendor, product or service depicted in this content nor makes any warranties, expressed or implied, with respect to this content, about its accuracy or completeness, including any warranties of merchantability or fitness for a particular purpose. This site is protected by hCaptcha and its [Privacy Policy](https://hcaptcha.com/privacy) and [Terms of Service](https://hcaptcha.com/terms) apply.

Gartner Peer Insights content consists of the opinions of individual end users based on their own experiences, and should not be construed as statements of fact, nor do they represent the views of Gartner or its affiliates. Gartner does not endorse any vendor, product or service depicted in this content nor makes any warranties, expressed or implied, with respect to this content, about its accuracy or completeness, including any warranties of merchantability or fitness for a particular purpose. This site is protected by hCaptcha and its [Privacy Policy](https://hcaptcha.com/privacy) and [Terms of Service](https://hcaptcha.com/terms) apply.

Gartner Peer Insights content consists of the opinions of individual end users based on their own experiences, and should not be construed as statements of fact, nor do they represent the views of Gartner or its affiliates. Gartner does not endorse any vendor, product or service depicted in this content nor makes any warranties, expressed or implied, with respect to this content, about its accuracy or completeness, including any warranties of merchantability or fitness for a particular purpose. This site is protected by hCaptcha and its [Privacy Policy](https://hcaptcha.com/privacy) and [Terms of Service](https://hcaptcha.com/terms) apply.

Gartner Peer Insights content consists of the opinions of individual end users based on their own experiences, and should not be construed as statements of fact, nor do they represent the views of Gartner or its affiliates. Gartner does not endorse any vendor, product or service depicted in this content nor makes any warranties, expressed or implied, with respect to this content, about its accuracy or completeness, including any warranties of merchantability or fitness for a particular purpose. This site is protected by hCaptcha and its [Privacy Policy](https://hcaptcha.com/privacy) and [Terms of Service](https://hcaptcha.com/terms) apply.

Gartner Peer Insights content consists of the opinions of individual end users based on their own experiences, and should not be construed as statements of fact, nor do they represent the views of Gartner or its affiliates. Gartner does not endorse any vendor, product or service depicted in this content nor makes any warranties, expressed or implied, with respect to this content, about its accuracy or completeness, including any warranties of merchantability or fitness for a particular purpose. This site is protected by hCaptcha and its [Privacy Policy](https://hcaptcha.com/privacy) and [Terms of Service](https://hcaptcha.com/terms) apply.

Gartner Peer Insights content consists of the opinions of individual end users based on their own experiences, and should not be construed as statements of fact, nor do they represent the views of Gartner or its affiliates. Gartner does not endorse any vendor, product or service depicted in this content nor makes any warranties, expressed or implied, with respect to this content, about its accuracy or completeness, including any warranties of merchantability or fitness for a particular purpose. This site is protected by hCaptcha and its [Privacy Policy](https://hcaptcha.com/privacy) and [Terms of Service](https://hcaptcha.com/terms) apply.
Consider these: